1 2011-07-24 00:00:12 <prof7bit> b4epoche: it was totally home custom made, i even developed a little app framework for it (we had php3 at that time). Parts of it were used by customers and other parts were used by us and another part of it was used by an external supplier. And it was supposed to let them all work together on the customers orders and the customers could track the progress and at some point they had to react and give approval, etc. very funny
   2 2011-07-24 00:00:13 <prof7bit>  system. 100s of kloc.
   3 2011-07-24 00:00:56 <jrmithdobbs> Diablo-D3: btw, you do realise that MS's AD auth/sso has been done via kerberos for ~10 years now
   4 2011-07-24 00:00:59 <jrmithdobbs> right?
   5 2011-07-24 00:01:22 <denisx> jrmithdobbs: I think your pasties in this pastie are messed up
   6 2011-07-24 00:01:27 <Diablo-D3> jrmithdobbs: I havent used MS products in a serious role for about 15 years
   7 2011-07-24 00:01:32 <Diablo-D3> actually, let me rephrase
   8 2011-07-24 00:01:38 <Diablo-D3> I havent used them in a serious role _ever_
   9 2011-07-24 00:02:19 <jrmithdobbs> Diablo-D3: oh so because the group of incompetent admins you dislike more than the others is capable of setting up kerberos on a platform you dislike more than others it's too hard to setup for incompetent admins.
  10 2011-07-24 00:02:23 <jrmithdobbs> i see.
  11 2011-07-24 00:02:37 <prof7bit> during these few years i aged approximately 120 years.
  12 2011-07-24 00:02:48 <jrmithdobbs> denisx: ?
  13 2011-07-24 00:02:53 <Diablo-D3> jrmithdobbs: um, nice strawman
  14 2011-07-24 00:02:56 <prof7bit> at least this is how it felt
  15 2011-07-24 00:03:06 <Diablo-D3> how does a MS product help secure an all linux cluster?
  16 2011-07-24 00:03:07 <jrmithdobbs> Diablo-D3: no that was sarcasm
  17 2011-07-24 00:03:12 <jrmithdobbs> Diablo-D3: yours was the strawman
  18 2011-07-24 00:03:18 <jrmithdobbs> Diablo-D3: linux auths of MS's AD just fine.
  19 2011-07-24 00:03:23 <jrmithdobbs> using kerberos.
  20 2011-07-24 00:03:25 <Diablo-D3> so MS AD runs on linux now?
  21 2011-07-24 00:03:33 <jrmithdobbs> that wasn't a requirement
  22 2011-07-24 00:03:47 <Diablo-D3> all linux cluster implies even the auth server runs linux
  23 2011-07-24 00:03:52 <jrmithdobbs> auth should be seperate from the cluster
  24 2011-07-24 00:04:00 <jrmithdobbs> separate
  25 2011-07-24 00:04:04 <prof7bit> it all boils down to how the user can understand it, everything else is almost meaningless.
  26 2011-07-24 00:04:10 freakazoid has joined
  27 2011-07-24 00:04:13 <Diablo-D3> jrmithdobbs: hows this, no windows in the damned building.
  28 2011-07-24 00:04:33 <jrmithdobbs> hosted AD
  29 2011-07-24 00:04:38 <jrmithdobbs> you suck at this game
  30 2011-07-24 00:04:42 <Diablo-D3> in the entire company
  31 2011-07-24 00:04:49 <jrmithdobbs> 3rd party contractor providing AD
  32 2011-07-24 00:04:54 <jrmithdobbs> you suck at this game
  33 2011-07-24 00:05:01 <Diablo-D3> then I'd use openid and co over kerb.
  34 2011-07-24 00:05:16 <jrmithdobbs> which destroys sso
  35 2011-07-24 00:05:50 <Diablo-D3> not at all
  36 2011-07-24 00:05:50 <denisx> jrmithdobbs: never mind, I thought some of them are in the wrong order...
  37 2011-07-24 00:05:53 <jrmithdobbs> Diablo-D3: basically your argument is *you* don't understand kerberos and *you* can't implement, so it's too hard ;p
  38 2011-07-24 00:06:02 <Diablo-D3> openid and friends are the ultimate sso
  39 2011-07-24 00:06:10 <Diablo-D3> jrmithdobbs: no, I probably could WRITE a kerb impl
  40 2011-07-24 00:06:36 <lolwat`> lolwut openid and krb are totally different
  41 2011-07-24 00:06:43 <jrmithdobbs> denisx: gpgmail hasn't been updated for mail.app 5.0 so some of the indenting accidentally got lost when i signed stuff
  42 2011-07-24 00:06:51 <jrmithdobbs> Diablo-D3: openid doesn't provide sso
  43 2011-07-24 00:07:05 <jrmithdobbs> Diablo-D3: how could you impl kerb if you don't understand what sso is?
  44 2011-07-24 00:07:12 theorb has joined
  45 2011-07-24 00:07:15 <Diablo-D3> jrmithdobbs: openid plus other things DO provide sso
  46 2011-07-24 00:07:21 <jrmithdobbs> what other things?
  47 2011-07-24 00:07:46 <Diablo-D3> oauth, for example, if you need tokens.
  48 2011-07-24 00:07:54 <jrmithdobbs> afaik with openid you have to resend creds for every auth request != sso
  49 2011-07-24 00:08:01 <sacarlson> how do you get listtransactions to provide more than the last 10 transactions?  count=15 returns [] , does mine have a bug?
  50 2011-07-24 00:08:04 <Diablo-D3> jrmithdobbs: no
  51 2011-07-24 00:08:08 <Diablo-D3> openid is a sign in once tech
  52 2011-07-24 00:08:28 <Diablo-D3> as long as your session exists, the sign in sticks
  53 2011-07-24 00:08:38 <jrmithdobbs> oh so no session timeout enforced by the auth layer? sounds like a horrible replacement for krb
  54 2011-07-24 00:08:44 theorbtwo has quit (Ping timeout: 260 seconds)
  55 2011-07-24 00:08:52 <jrmithdobbs> and settable per session by the user
  56 2011-07-24 00:08:55 <Diablo-D3> jrmithdobbs: its used for websites often
  57 2011-07-24 00:09:03 theorb is now known as theorbtwo
  58 2011-07-24 00:09:10 <jrmithdobbs> i know what oath/openid are, and they're awful imho
  59 2011-07-24 00:09:10 <Diablo-D3> now, yes, I do agree that web sign ons in general are screwy
  60 2011-07-24 00:09:21 <sacarlson> jrmithdobbs: you can add tokens in openid sessions that add timeout
  61 2011-07-24 00:09:32 <Diablo-D3> sacarlson: yeah, but the website has to kick you out first
  62 2011-07-24 00:09:49 <Diablo-D3> websites that use openid typically just plop a session cookie in your browser like normal
  63 2011-07-24 00:09:59 <Diablo-D3> the openid provider itself has no control there
  64 2011-07-24 00:10:08 <sipa> devrandom: hmm, no incoming?
  65 2011-07-24 00:10:11 <sacarlson> Diablo-D3: everytime you update the page a new token is provided I think
  66 2011-07-24 00:10:17 <sipa> devrandom: is that on xp or w7?
  67 2011-07-24 00:10:18 <jrmithdobbs> sacarlson: not on most sites
  68 2011-07-24 00:10:18 <Diablo-D3> sacarlson: nope
  69 2011-07-24 00:10:30 <Diablo-D3> there are sites that do that, but it is NOT part of openid
  70 2011-07-24 00:10:30 <jrmithdobbs> sacarlson: that could be done in theory but then you'd have to reauth on every page iirc
  71 2011-07-24 00:10:38 <sacarlson> jrmithdobbs: I"ll have to double check mine then
  72 2011-07-24 00:10:42 <jrmithdobbs> or cache creds
  73 2011-07-24 00:10:45 <jrmithdobbs> which is worse
  74 2011-07-24 00:10:58 <Diablo-D3> websites typically just have a use once session cookie
  75 2011-07-24 00:11:02 <jrmithdobbs> openid wishes it were kerberos basically, but it's not
  76 2011-07-24 00:11:11 <jrmithdobbs> err oauth+openid
  77 2011-07-24 00:11:12 <Diablo-D3> when it expires, you have to log in again, or get a new session cookie that the site gave you
  78 2011-07-24 00:11:34 <cjdelisle> yea if it constantly changed the tabs -> fail
  79 2011-07-24 00:11:59 <Diablo-D3> jrmithdobbs: well, it depends how you define "kerb usage"
  80 2011-07-24 00:12:09 <Diablo-D3> most people using kerb simply arent using everything it can do
  81 2011-07-24 00:12:15 <Diablo-D3> and are using it purely for the sso backend
  82 2011-07-24 00:12:20 <jrmithdobbs> true
  83 2011-07-24 00:12:39 <Diablo-D3> openid's only flaw imo is it requires an actual web browser (ie, its meant for web sites) to login
  84 2011-07-24 00:12:46 <jrmithdobbs> but it can be used for everything from auth->sso->data channel encryption->identity verification of 3rd parties
  85 2011-07-24 00:12:49 <jrmithdobbs> <3 kerberos
  86 2011-07-24 00:13:00 <Diablo-D3> jrmithdobbs: well, theres a small problem with that
  87 2011-07-24 00:13:13 <jrmithdobbs> which is?
  88 2011-07-24 00:13:14 <Diablo-D3> data channel encryption IS handled in openid+oauth land
  89 2011-07-24 00:13:22 <jrmithdobbs> out of band
  90 2011-07-24 00:13:29 <Diablo-D3> its a part of the stack, but not part of either of those specs
  91 2011-07-24 00:13:47 <Diablo-D3> identify verification can be done with oauth, btw
  92 2011-07-24 00:13:56 <Diablo-D3> actually, it can be done with SOME openid providers too
  93 2011-07-24 00:14:21 <jrmithdobbs> ya that's the other problem with openid/oath is every provider is slightly different in spec-breaking ways ;p
  94 2011-07-24 00:14:23 <Diablo-D3> theres at least one that, for example, offer support for shit like finger print readers and crypto dongles
  95 2011-07-24 00:14:30 <Diablo-D3> not spec breaking
  96 2011-07-24 00:14:34 <Diablo-D3> they just offer more choices
  97 2011-07-24 00:14:52 <Diablo-D3> the website using openid doesnt NEED to know whats going on
  98 2011-07-24 00:15:02 <jrmithdobbs> Diablo-D3: google has actually, several times, changed user's ticket id or whatever that is their unique identifier
  99 2011-07-24 00:15:06 <jrmithdobbs> not joking
 100 2011-07-24 00:15:22 <jrmithdobbs> so if you only used google as a provider for some website you'd lose your account because your identifier changed
 101 2011-07-24 00:15:25 <jrmithdobbs> lol
 102 2011-07-24 00:15:32 <Diablo-D3> jrmithdobbs: wtf
 103 2011-07-24 00:15:49 <BlueMatt> damn, sudo via fingerprint swipe is sexy as hell
 104 2011-07-24 00:16:01 <sacarlson> Diablo-D3: you were correct my site doesn't change the token unless the session is broken it would appear
 105 2011-07-24 00:16:04 <Diablo-D3> bluematt: pam has fucking shit for everything
 106 2011-07-24 00:16:10 <BlueMatt> no shit
 107 2011-07-24 00:16:12 <jrmithdobbs> i really wish i was joking, because openid/oauth are better than *nothing* (but not as good as kerberos)
 108 2011-07-24 00:16:25 <Diablo-D3> jrmithdobbs: wtf, that site is somewhat possibly broken
 109 2011-07-24 00:16:32 <Diablo-D3> I mean, wtf google and all
 110 2011-07-24 00:16:40 <Diablo-D3> but Im not using that id
 111 2011-07-24 00:16:43 <jrmithdobbs> Diablo-D3: no google is broken, other providers have done it too
 112 2011-07-24 00:16:55 underscor has quit (Ping timeout: 276 seconds)
 113 2011-07-24 00:17:07 <phantomcircuit> tcatm, am i seriously going to have to use mod_rewrite just to name files trades.json
 114 2011-07-24 00:17:12 <Diablo-D3> what Im saying is, Im not using that for the internal id on my site
 115 2011-07-24 00:17:16 <BlueMatt> now if only dmcrypt would decrypt via fingerprint scan...
 116 2011-07-24 00:17:22 <jrmithdobbs> Diablo-D3: was just an example
 117 2011-07-24 00:17:26 <phantomcircuit> BlueMatt, you can do that
 118 2011-07-24 00:17:31 <phantomcircuit> it's just silly insecure
 119 2011-07-24 00:17:38 <BlueMatt> yea, I would think so
 120 2011-07-24 00:17:39 <Diablo-D3> bluematt: dmcrypt doesnt use pam for auth
 121 2011-07-24 00:17:47 <Diablo-D3> it'd be interesting as hell if it did, though
 122 2011-07-24 00:17:54 <phantomcircuit> BlueMatt, sudo via fingerprint swipe is also insecure as hell
 123 2011-07-24 00:18:04 <BlueMatt> Diablo-D3: I know, I was just wishing if it could use some kind of fingerprint scan hash for key
 124 2011-07-24 00:18:05 <jrmithdobbs> i'm working on using OTP-ish stuff for dm_crypt key storage ;p
 125 2011-07-24 00:18:21 <phantomcircuit> the fingerprint is stored as a hash of some data structure, but it only results in like 12 bits of entropy
 126 2011-07-24 00:18:24 <jrmithdobbs> (technically it's challenge-response, but in a semi-onetime way)
 127 2011-07-24 00:18:25 <Diablo-D3> jrmithdobbs: still, the problem is this
 128 2011-07-24 00:18:29 <BlueMatt> phantomcircuit: well yea, someone can grab a scan of your print from a glass or something...
 129 2011-07-24 00:18:33 <Diablo-D3> jrmithdobbs: I do _not_ want to deploy my own auth scheme
 130 2011-07-24 00:18:36 <BlueMatt> phantomcircuit: thats it?
 131 2011-07-24 00:18:44 <phantomcircuit> BlueMatt, no you can bruteforce it trivially
 132 2011-07-24 00:18:46 <Diablo-D3> jrmithdobbs: its easy to code, its easy to secure from the electronic side
 133 2011-07-24 00:18:49 <jrmithdobbs> Diablo-D3: internally?
 134 2011-07-24 00:18:53 <BlueMatt> phantomcircuit: 2**12 sounds way too small
 135 2011-07-24 00:18:55 <Diablo-D3> jrmithdobbs: yeah
 136 2011-07-24 00:19:02 <jrmithdobbs> why wouldn't you deploy your own auth scheme?
 137 2011-07-24 00:19:02 <phantomcircuit> BlueMatt, lol it is
 138 2011-07-24 00:19:04 <Diablo-D3> jrmithdobbs: but its not actual authentication
 139 2011-07-24 00:19:09 <BlueMatt> phantomcircuit: maybe that was some old model?
 140 2011-07-24 00:19:38 <prof7bit> fingers can be hacked easily.
 141 2011-07-24 00:19:45 <jrmithdobbs> BlueMatt: i don't trust biometric scanners, almost all of them have ended up having HORRIBLEY simple spoofing methods
 142 2011-07-24 00:19:49 <BlueMatt> yea, they can get a print from something you touch...
 143 2011-07-24 00:19:52 <jrmithdobbs> BlueMatt: at least, not as a lone factor
 144 2011-07-24 00:19:54 <phantomcircuit> BlueMatt, my T61's builtin fingerprint swipe would result in a 64 bit "hash" of which the first 42 bits where trivially predicted
 145 2011-07-24 00:20:08 <Diablo-D3> jrmithdobbs: I can have them put an email in, and then even send a click me auth url to the email, it still doesnt tell me who the people really are
 146 2011-07-24 00:20:11 senseles has joined
 147 2011-07-24 00:20:16 <phantomcircuit> not to mention that my fingerprints where all over the thing anyways lol
 148 2011-07-24 00:20:20 <BlueMatt> phantomcircuit: damn...again though, maybe that was an old model...
 149 2011-07-24 00:20:27 <jrmithdobbs> Diablo-D3: if you're talking for external users, sure
 150 2011-07-24 00:20:33 <Diablo-D3> jrmithdobbs: this is for external users
 151 2011-07-24 00:20:35 <phantomcircuit> BlueMatt, uh 2007
 152 2011-07-24 00:20:40 <BlueMatt> yea, the fact that your print is on the laptop does make it fairly insecure if you spend some time...
 153 2011-07-24 00:20:44 <jrmithdobbs> Diablo-D3: i was talking auth within an organization
 154 2011-07-24 00:21:00 <Diablo-D3> jrmithdobbs: people use kerb for sso on public services.
 155 2011-07-24 00:21:05 <Diablo-D3> for a wide range of it, too
 156 2011-07-24 00:21:11 <devrandom> sipa: w7
 157 2011-07-24 00:21:11 <phantomcircuit> BlueMatt, i always thought it a hilarious authentication method
 158 2011-07-24 00:21:12 <jrmithdobbs> Diablo-D3: the reason kerberos (sadly) doesn't work on the internet is because not enough people use it so the WOT backend with realm trust isn't there to make it simple
 159 2011-07-24 00:21:27 <phantomcircuit> "lets take this device covered in your fingerprints and use a fingerprint scanner for authentication"
 160 2011-07-24 00:21:28 <phantomcircuit> ???
 161 2011-07-24 00:21:34 <BlueMatt> phantomcircuit: theoretically its really secure (aside from the stealing prints thing...)
 162 2011-07-24 00:21:51 <sipa> devrandom: hmmm, too bad
 163 2011-07-24 00:21:59 <Diablo-D3> bluematt: fingerprint scanners are not secure and never will be
 164 2011-07-24 00:22:18 <phantomcircuit> BlueMatt, that depends entirely on the accuracy of the scanner and whether you're ok with storing full prints for comparison
 165 2011-07-24 00:22:22 <BlueMatt> Diablo-D3: given the copying print stuff, yes
 166 2011-07-24 00:22:26 <jrmithdobbs> BlueMatt: seriously, never use any biometric scanner as single-factor of auth
 167 2011-07-24 00:22:35 <BlueMatt> jrmithdobbs: absolutely
 168 2011-07-24 00:22:42 <phantomcircuit> also tcatm's new api naming format is unfortunate
 169 2011-07-24 00:22:48 <jrmithdobbs> they all make wild claims and 90%+ of said claims turn out to be blatantly false
 170 2011-07-24 00:23:04 <jrmithdobbs> eg, phantomcircuit's T61 scanner reader (and I'm pretty sure the current gen uses a very similar reader)
 171 2011-07-24 00:23:14 <jrmithdobbs> err s/scanner reader/scanner example/
 172 2011-07-24 00:23:15 <Diablo-D3> you know what I trust?
 173 2011-07-24 00:23:19 <jrmithdobbs> nothing
 174 2011-07-24 00:23:20 <BlueMatt> jrmithdobbs: many in the past have been shit yea...but they do add another 90% that you didnt have before
 175 2011-07-24 00:23:28 <Diablo-D3> if someone auths through ssh and has the right key.
 176 2011-07-24 00:23:29 osmosis has joined
 177 2011-07-24 00:23:32 <BlueMatt> java
 178 2011-07-24 00:23:34 <prof7bit> relying on biometry and promoting it also sends the wrong message to users. politically. the next thing then would be the acceptance of implanted rfid.
 179 2011-07-24 00:23:46 <Diablo-D3> because if they have the right key, if they're NOT them, its already too late and, thus, not my problem.
 180 2011-07-24 00:23:53 <BlueMatt> prof7bit: oh god
 181 2011-07-24 00:23:56 <jrmithdobbs> Diablo-D3: lol, agreed
 182 2011-07-24 00:24:06 <phantomcircuit> lol i went into a club that chips people
 183 2011-07-24 00:24:06 <BlueMatt> so true
 184 2011-07-24 00:24:08 <phantomcircuit> shit was weird
 185 2011-07-24 00:24:14 <phantomcircuit> people scanning their arm to buy drinks
 186 2011-07-24 00:24:16 <jrmithdobbs> Diablo-D3: though i think proper steps should be taken to protect such keys
 187 2011-07-24 00:24:22 <Diablo-D3> jrmithdobbs: of course.
 188 2011-07-24 00:24:24 <Diablo-D3> but honestly
 189 2011-07-24 00:24:29 <Diablo-D3> if you let a fucker walk off with your ssh keys
 190 2011-07-24 00:24:34 <Diablo-D3> that is your own fucking fault
 191 2011-07-24 00:25:01 <jrmithdobbs> btw
 192 2011-07-24 00:25:04 <jrmithdobbs> speaking of dmcrypt
 193 2011-07-24 00:25:08 <jrmithdobbs> LUKS is fuckin retarded
 194 2011-07-24 00:25:28 <jrmithdobbs> (that is all)
 195 2011-07-24 00:25:36 <BlueMatt> how so?
 196 2011-07-24 00:25:46 * BlueMatt knows little about luks other than it wfm
 197 2011-07-24 00:25:57 <jrmithdobbs> it gives end users the illusion of added security that isn't necessarily true
 198 2011-07-24 00:26:08 <BlueMatt> how so?
 199 2011-07-24 00:26:08 <jrmithdobbs> eg, changing your passphrase does not actually change the keying material
 200 2011-07-24 00:26:16 <BlueMatt> well no shit
 201 2011-07-24 00:26:18 <Diablo-D3> jrmithdobbs: yeah
 202 2011-07-24 00:26:26 <jrmithdobbs> so if your passphrase was compromised and you don't know that the keying material hasn't already been compromised ....
 203 2011-07-24 00:26:26 <Diablo-D3> changing the key should be, like, restriping a RAID in place
 204 2011-07-24 00:26:33 <BlueMatt> if a user thinks changing the password changes the keying material, they are fucking retarded
 205 2011-07-24 00:26:38 <Diablo-D3> a slow painful process that burns drives ups
 206 2011-07-24 00:26:40 <Diablo-D3> *up
 207 2011-07-24 00:26:42 <jrmithdobbs> BlueMatt: i'm not done
 208 2011-07-24 00:26:44 genjix has joined
 209 2011-07-24 00:26:47 <jrmithdobbs> BlueMatt: that's just my initial objection ;p
 210 2011-07-24 00:27:09 <Diablo-D3> jrmithdobbs: you know what works GREAT for security?
 211 2011-07-24 00:27:13 <Diablo-D3> an expensive RAID card
 212 2011-07-24 00:27:18 <Diablo-D3> that encrypts each drive individually
 213 2011-07-24 00:27:19 <jrmithdobbs> BlueMatt: their code that DOES the key changing is overly complex and actually does the stupid "rewrite portions multiple times" shit which is fairly succepitble to bugs
 214 2011-07-24 00:27:27 <Diablo-D3> and never tells you what the key is
 215 2011-07-24 00:27:34 <Diablo-D3> remove the drive? dohohoho, the drive is useless.
 216 2011-07-24 00:27:36 mosimo has quit (Read error: Connection reset by peer)
 217 2011-07-24 00:27:42 <BlueMatt> Diablo-D3: until the card fails...
 218 2011-07-24 00:27:48 <Diablo-D3> bluematt: yeah so?
 219 2011-07-24 00:27:49 <jrmithdobbs> BlueMatt: so say there's some bug in the re-passphrasing (because it's not re-keying) that overwrites the wrong portion of the LUKS header ... bam data gone
 220 2011-07-24 00:28:00 <Diablo-D3> why would anyone have only one copy of data
 221 2011-07-24 00:28:21 sipa has left ()
 222 2011-07-24 00:28:21 <jrmithdobbs> because they're users
 223 2011-07-24 00:28:27 <Diablo-D3> if your data storage cluster cant deal with the actual death of machines, you're fucked
 224 2011-07-24 00:28:31 <genjix> because data is money and money is time, and time IS LAW
 225 2011-07-24 00:28:43 <Diablo-D3> hell, I want to, some day, have a data storage cluster that can survive the death of entire countries
 226 2011-07-24 00:28:45 <genjix> first law of timeonomics
 227 2011-07-24 00:29:01 <Diablo-D3> an impervious cloud of impervitude.
 228 2011-07-24 00:29:18 <genjix> would it use stone tablet arrays?
 229 2011-07-24 00:29:27 <Diablo-D3> no
 230 2011-07-24 00:29:37 <jrmithdobbs> BlueMatt: thirdly, LUKS proves that you have encrypted data on the device, destroying any form of plausible deniability
 231 2011-07-24 00:29:42 <Diablo-D3> it'd use the dead flesh of my enemies
 232 2011-07-24 00:30:10 <Diablo-D3> periodically I'd have to get new enemies for more dead flesh, because even at low tempetures, it doesnt last long
 233 2011-07-24 00:30:14 <BlueMatt> jrmithdobbs: thats true, it would be nice to have that
 234 2011-07-24 00:30:26 <jrmithdobbs> BlueMatt: the metadata should never have been implemented as a prefix on the same block device it was protecting
 235 2011-07-24 00:30:32 <jrmithdobbs> that was a HORRIBLE design decision
 236 2011-07-24 00:30:37 <jrmithdobbs> at least, *requiring* it is
 237 2011-07-24 00:30:43 <Diablo-D3> jrmithdobbs: well
 238 2011-07-24 00:30:48 <Diablo-D3> they were trying to make it easy to adopt
 239 2011-07-24 00:30:52 <Diablo-D3> which, honestly, was dumb
 240 2011-07-24 00:30:52 moa7 has left ()
 241 2011-07-24 00:30:55 <jrmithdobbs> ya
 242 2011-07-24 00:31:14 <BlueMatt> does dm-crypt support anything aside from luks?
 243 2011-07-24 00:31:34 <jrmithdobbs> luks sits on top of dm-crypt
 244 2011-07-24 00:31:36 <jrmithdobbs> not the other way around
 245 2011-07-24 00:31:47 <BlueMatt> ah, ok
 246 2011-07-24 00:31:49 Sedra- has joined
 247 2011-07-24 00:32:01 <Diablo-D3> I thought luks was a dm-crypt backend?
 248 2011-07-24 00:32:01 <jrmithdobbs> you can use dm_crypt directly with no luks/on storage metadata, but you have to do your own key management
 249 2011-07-24 00:32:04 agath has quit (Ping timeout: 260 seconds)
 250 2011-07-24 00:32:25 <jrmithdobbs> like i said earlier, i'm working on something like this ;p
 251 2011-07-24 00:32:40 <BlueMatt> well then you should write LUKSv3 with plausible deniability and crap
 252 2011-07-24 00:32:52 <Diablo-D3> jrmithdobbs: you know what I thought was interesting?
 253 2011-07-24 00:32:54 <Diablo-D3> encfs
 254 2011-07-24 00:32:59 <jrmithdobbs> no i think luks is also fundamentally flawed in that it shouldn't live in the kernel anyways
 255 2011-07-24 00:33:12 <jrmithdobbs> key management should be completely sep imho
 256 2011-07-24 00:33:21 <BlueMatt> its really nice to have dm-crypt in kernel
 257 2011-07-24 00:33:24 <Diablo-D3> jrmithdobbs: erm
 258 2011-07-24 00:33:28 <jrmithdobbs> dm_crypt is fine in the kernel
 259 2011-07-24 00:33:31 <Diablo-D3> it has to be either in kernel, or initrd
 260 2011-07-24 00:33:33 <jrmithdobbs> it's luks specifically i don't like
 261 2011-07-24 00:33:47 <Diablo-D3> and stuff in initrd is usually fucking tiny, or partly kernel powered
 262 2011-07-24 00:34:08 <jrmithdobbs> Diablo-D3: bah, i ran my mining cluster off initrd
 263 2011-07-24 00:34:12 <CIA-103> libbitcoin: genjix * rc019840cb36f / (3 files in 3 dirs): Added Elliptic Curve Key class.
 264 2011-07-24 00:34:13 <sacarlson> jrmithdobbs: I came up with this stupid idea using luks https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnxyZW1vdGVrZXllbmNyeXB0fGd4OjU2YmVmY2NlY2Y2ZjhkZjM
 265 2011-07-24 00:34:25 <jrmithdobbs> (seriously, the initrd was pulled down and contained the entire os image)
 266 2011-07-24 00:34:32 <jrmithdobbs> over the network
 267 2011-07-24 00:34:36 Sedra has quit (Ping timeout: 252 seconds)
 268 2011-07-24 00:34:55 <jrmithdobbs> Diablo-D3: initrd is *much* easier to modify than kernel level and *much* easier for 3rd parties to change out as necessary for their needs
 269 2011-07-24 00:35:25 <Diablo-D3> sure.
 270 2011-07-24 00:35:27 <Diablo-D3> Im just saying
 271 2011-07-24 00:35:31 <jrmithdobbs> Diablo-D3: also, luks already requires userland logic in the initrd so taking it out of the kernel actually simplifies things
 272 2011-07-24 00:35:39 <Diablo-D3> jrmithdobbs: but what do you think of encfs?
 273 2011-07-24 00:35:56 <jrmithdobbs> fuse == fail
 274 2011-07-24 00:35:57 <jrmithdobbs> lol
 275 2011-07-24 00:35:58 <Diablo-D3> its a fuse fs that encrypts stuff as normal files
 276 2011-07-24 00:36:04 <jrmithdobbs> fuck fuse
 277 2011-07-24 00:36:13 <Diablo-D3> fuse isnt nearly as bad as people think it is
 278 2011-07-24 00:36:44 <genjix> guys
 279 2011-07-24 00:36:57 <genjix> lets drop the term generating or miners
 280 2011-07-24 00:37:06 <genjix> due to the bad connotation that generating has
 281 2011-07-24 00:37:11 <genjix> of magicking money out of thin air
 282 2011-07-24 00:37:15 <genjix> i propose the term
 283 2011-07-24 00:37:18 <genjix> "processors"
 284 2011-07-24 00:37:44 <lolwat`> https://lkml.org/lkml/2011/6/9/462 lulz were had
 285 2011-07-24 00:38:09 <lolwat`> genjix, but it is magicking money out of thin air.  you are suggesting being dishonest
 286 2011-07-24 00:38:14 <phantomcircuit> Diablo-D3, fuse is REALLY slow
 287 2011-07-24 00:38:26 sacarlson has quit (Ping timeout: 250 seconds)
 288 2011-07-24 00:38:31 <Diablo-D3> phantomcircuit: depends how you define slow
 289 2011-07-24 00:38:36 <Diablo-D3> and it also depends on the fs
 290 2011-07-24 00:38:53 <Diablo-D3> there are some REALLY retarded fuse fses
 291 2011-07-24 00:39:04 <genjix> lolwat`: no, it's an inaccurate term
 292 2011-07-24 00:39:12 <genjix> there is no magicking of anything
 293 2011-07-24 00:39:20 <lolwat`> er but
 294 2011-07-24 00:39:23 <phantomcircuit> Diablo-D3, the onyl fuse fs i would ever use is sshfs
 295 2011-07-24 00:39:25 <Diablo-D3> but fuse itself, generally, your disk IO is magnitudes slower than your cpu performance for any given op
 296 2011-07-24 00:39:33 <lolwat`> the first thing people think when you explain bitcoin is
 297 2011-07-24 00:39:33 <Diablo-D3> phantomcircuit: I use sshfs on a regular basis
 298 2011-07-24 00:39:38 <genjix> i don't give myself 1 million bitcoins
 299 2011-07-24 00:39:39 <lolwat`> "where did the first coins come from"
 300 2011-07-24 00:39:54 <jrmithdobbs> genjix: i like "validators" as jgarzik proposed better
 301 2011-07-24 00:39:58 <genjix> the processors get a reward for validating transactions in the network
 302 2011-07-24 00:40:04 <genjix> validators is good
 303 2011-07-24 00:40:09 <genjix> lets go with that then.
 304 2011-07-24 00:40:14 <Diablo-D3> phantomcircuit: its the only network fs that can saturate gigabit without any tweaking right out of the box
 305 2011-07-24 00:40:20 <jrmithdobbs> Diablo-D3: fuse isn't awful for all things
 306 2011-07-24 00:40:21 <Diablo-D3> phantomcircuit: and while consuming no real cpu time
 307 2011-07-24 00:40:28 <b4epoche_> just say they were 'found'
 308 2011-07-24 00:40:29 <phantomcircuit> Diablo-D3, depends on the cpu ;)
 309 2011-07-24 00:40:30 <Diablo-D3> phantomcircuit: but, for example, I have a 4 drive raid 5
 310 2011-07-24 00:40:31 <jrmithdobbs> Diablo-D3: things like ssh-fs and such i fully endorse
 311 2011-07-24 00:40:41 <jrmithdobbs> Diablo-D3: but it doesn't fit this use case imho
 312 2011-07-24 00:40:43 <Diablo-D3> phantomcircuit: it does about 266mb/sec random writes
 313 2011-07-24 00:40:49 <genjix> i like fuse because i get to feel like a kernel developer by writing my own fs
 314 2011-07-24 00:40:50 <phantomcircuit> my P4 cannot do gbps ssh without tewaks
 315 2011-07-24 00:40:57 <jrmithdobbs> Diablo-D3: eg, you can't do md+lvm on top of a fuse device
 316 2011-07-24 00:41:00 <Diablo-D3> my cpu can do a fucking shitload of encryption in the mean time
 317 2011-07-24 00:41:00 <b4epoche_> the easiest way for me to explain it to people is to say there are 21M bitcoins that will be found
 318 2011-07-24 00:41:01 <jrmithdobbs> and if you do i fear for your data
 319 2011-07-24 00:41:09 <Diablo-D3> jrmithdobbs: no, because its not a device
 320 2011-07-24 00:41:15 <genjix> i made my own network distributed fs using bash scripts and C
 321 2011-07-24 00:41:16 <genjix> XD
 322 2011-07-24 00:41:25 <Diablo-D3> jrmithdobbs: there HAS been talk about fuse-as-a-device
 323 2011-07-24 00:41:26 <genjix> because of fuse
 324 2011-07-24 00:41:34 <jrmithdobbs> Diablo-D3: you could create an image file inside the fuse fs and mount it loopback (completely defeating the purpose)
 325 2011-07-24 00:41:36 zeropointo has joined
 326 2011-07-24 00:41:39 <phantomcircuit> it's there
 327 2011-07-24 00:41:41 <phantomcircuit> let me find it
 328 2011-07-24 00:41:52 <jrmithdobbs> Diablo-D3: but that is an awful idea
 329 2011-07-24 00:41:53 <Diablo-D3> jrmithdobbs: fuse is about file systems, though
 330 2011-07-24 00:41:55 <Diablo-D3> not about block devices
 331 2011-07-24 00:42:10 <Diablo-D3> md and lvm are about block devices
 332 2011-07-24 00:42:15 <jrmithdobbs> Diablo-D3: right, but if i want to encrypt media i want the media level encrypted not just the data layer
 333 2011-07-24 00:42:19 <phantomcircuit> CONFIG_CUSE
 334 2011-07-24 00:42:26 <Diablo-D3> jrmithdobbs: yes, but fuse isnt about encryption
 335 2011-07-24 00:42:30 <phantomcircuit> that's already been implemented
 336 2011-07-24 00:42:36 <jrmithdobbs> Diablo-D3: i'm talking specifically in relation to encfs
 337 2011-07-24 00:42:36 <Diablo-D3> jrmithdobbs: its just that several fuse fs DO encrypt
 338 2011-07-24 00:42:53 <Diablo-D3> jrmithdobbs: encfs is interesting because it encrypts stuff in an interesting way
 339 2011-07-24 00:42:54 <genjix> fuse is to make it easy to build your own fs for userspace devs
 340 2011-07-24 00:43:00 <jrmithdobbs> Diablo-D3: other things fuse should not be used for: zfs, ntfs, etc for block devices
 341 2011-07-24 00:43:10 <Diablo-D3> I use zfs fuse.
 342 2011-07-24 00:43:14 <jrmithdobbs> i'm sorry
 343 2011-07-24 00:43:20 <jrmithdobbs> install freebsd
 344 2011-07-24 00:43:21 <Diablo-D3> Im waiting for btrfs to stop sucking
 345 2011-07-24 00:43:37 <jrmithdobbs> if you really want zfs on something open, freebsd is the way to get it
 346 2011-07-24 00:43:37 <Diablo-D3> but oracle basically doesnt give a fuck anymore now that they own sun
 347 2011-07-24 00:43:44 <Diablo-D3> jrmithdobbs: fbsd can suck my dick
 348 2011-07-24 00:44:02 <jrmithdobbs> then you're stuck with the shittiest zfs implementation in existence ;p
 349 2011-07-24 00:44:03 <Diablo-D3> I once said "oh hey, its not possible for people to be more anal about shit than linux core people"
 350 2011-07-24 00:44:07 <Diablo-D3> and then I ran fbsd.
 351 2011-07-24 00:44:13 <Diablo-D3> the universe proved me wrong.
 352 2011-07-24 00:44:17 <lolwat`> huh
 353 2011-07-24 00:44:26 <jrmithdobbs> Diablo-D3: the freebsd maints aren't that bad
 354 2011-07-24 00:44:26 <lolwat`> linux people aren't anal at all
 355 2011-07-24 00:44:36 <jrmithdobbs> also that, they're not anal enough
 356 2011-07-24 00:44:46 <Diablo-D3> jrmithdobbs: the fbsd community pisses me off so goddamned much that I would rather run hurd.
 357 2011-07-24 00:44:49 <jrmithdobbs> i want to punch all these kids forcing distros to migrate to upstart and similar
 358 2011-07-24 00:44:50 <lolwat`> BSD folks are more academic
 359 2011-07-24 00:44:58 shawn-p has joined
 360 2011-07-24 00:45:01 <Diablo-D3> jrmithdobbs: and no, they're not anal enough in some respects
 361 2011-07-24 00:45:01 <lolwat`> they like to think things through first
 362 2011-07-24 00:45:02 <jrmithdobbs> I DON'T GIVE A FUCK ABOUT BOOT TIMES GODDAMN IT SYSV INIT IS FINE LEAVE IT ALONE
 363 2011-07-24 00:45:13 <Diablo-D3> sysvinit is fine _now_
 364 2011-07-24 00:45:14 <Diablo-D3> you know why?
 365 2011-07-24 00:45:15 <jrmithdobbs> </init rant>
 366 2011-07-24 00:45:22 <Diablo-D3> debian migrated to a sysvinit that does parallel startup.
 367 2011-07-24 00:45:30 <jrmithdobbs> i don't like debian's new stuff either
 368 2011-07-24 00:45:40 <jrmithdobbs> it's not NEARLY as horrible as the others though
 369 2011-07-24 00:45:45 <Diablo-D3> I can go from grub exit to slim ready to type shit in in under 15 seconds
 370 2011-07-24 00:45:56 <Diablo-D3> and xfce instantly starts after Im done with slim
 371 2011-07-24 00:46:00 <jrmithdobbs> i don't give two shits about boot times on linux
 372 2011-07-24 00:46:05 <Diablo-D3> and I have slow drives for my raid1
 373 2011-07-24 00:46:06 <jrmithdobbs> if you do, you should stop rebooting
 374 2011-07-24 00:46:11 <Diablo-D3> I dont reboot
 375 2011-07-24 00:46:17 <Diablo-D3> its just that 15 seconds piss me off when I have to
 376 2011-07-24 00:46:23 <genjix> always love watching you two duke it out, but i gotta work
 377 2011-07-24 00:46:25 <genjix> <3
 378 2011-07-24 00:46:26 genjix has left ()
 379 2011-07-24 00:46:27 <Diablo-D3> FUCKING FGLRX WAAARGH
 380 2011-07-24 00:46:31 <Diablo-D3> etc etc
 381 2011-07-24 00:46:39 <jrmithdobbs> oh see there's your mistake
 382 2011-07-24 00:46:53 <jrmithdobbs> it's 20 years later and guess what? linux? she's not ready for the desktop. ;p
 383 2011-07-24 00:47:00 <Diablo-D3> no
 384 2011-07-24 00:47:07 <Diablo-D3> I just shouldnt develop my own miner
 385 2011-07-24 00:47:12 <Diablo-D3> fglrx has been ultrastable for like 2-3 years
 386 2011-07-24 00:47:18 <jrmithdobbs> and ati should stop writing shit kernel drivers
 387 2011-07-24 00:47:18 <WakiMiko> 2012 will be the year of the linux desktop
 388 2011-07-24 00:47:18 <erus`> do we have wireless drivers yet?
 389 2011-07-24 00:47:19 <jrmithdobbs> lol
 390 2011-07-24 00:47:22 <WakiMiko> xD
 391 2011-07-24 00:47:27 <Diablo-D3> jrmithdobbs: they dont write a kernel driver
 392 2011-07-24 00:47:27 <jrmithdobbs> erus`: lol
 393 2011-07-24 00:47:32 <Diablo-D3> fglrx isnt in the kernel
 394 2011-07-24 00:47:36 <Diablo-D3> even the module is basically blank
 395 2011-07-24 00:47:55 <jrmithdobbs> Diablo-D3: the module lets userspace stuff cause horrible kernel-level things to happen though
 396 2011-07-24 00:48:01 <Diablo-D3> yes
 397 2011-07-24 00:48:04 <Diablo-D3> thats exactly what its for
 398 2011-07-24 00:48:06 <jrmithdobbs> which is worse if anything
 399 2011-07-24 00:48:08 <WakiMiko> gmaxwell: you linked the same pull request twice in your mail
 400 2011-07-24 00:48:11 <Diablo-D3> its not nearly as clever as, say, DRM
 401 2011-07-24 00:48:35 <Diablo-D3> DRM + that new setup code shit in the kernel whos name Im forgetting is how its meant to be done
 402 2011-07-24 00:48:49 <Diablo-D3> btw, as a side note: AMD wants fglrx dead
 403 2011-07-24 00:48:53 gribble has quit (Disconnected by services)
 404 2011-07-24 00:48:53 <Diablo-D3> they want it dead and buried
 405 2011-07-24 00:48:59 <Diablo-D3> and they hate ATI for ever having created it
 406 2011-07-24 00:49:00 <jrmithdobbs> good because it's crap
 407 2011-07-24 00:49:11 <Diablo-D3> AMD's official company motto: "we're not a software company"
 408 2011-07-24 00:49:30 <jrmithdobbs> i've not used linux on the desktop in any kind of semi-serious way since voodoo3 era tbqh, and that had a real drm module ;p
 409 2011-07-24 00:49:37 <Diablo-D3> they've spent millions of dollars on the gallium stack to make it what it is
 410 2011-07-24 00:49:47 <Diablo-D3> and Im glad they have
 411 2011-07-24 00:49:50 <BlueMatt> jrmithdobbs: wait...what do you use then?
 412 2011-07-24 00:49:52 <jrmithdobbs> vt extensions == fuck linux on the desktop
 413 2011-07-24 00:50:04 <jrmithdobbs> BlueMatt: os x
 414 2011-07-24 00:50:06 <Diablo-D3> jrmithdobbs: Ive used linux exclusively as my desktop OS for over a decade
 415 2011-07-24 00:50:13 <Diablo-D3> you know, honestly, osx isnt that bad
 416 2011-07-24 00:50:17 <BlueMatt> jrmithdobbs: well its 1000x better than windows
 417 2011-07-24 00:50:18 <Diablo-D3> but you need a fucking beefy mac
 418 2011-07-24 00:50:25 <Diablo-D3> I own a powerbook g4
 419 2011-07-24 00:50:36 <BlueMatt> thats a damn old mac
 420 2011-07-24 00:50:38 <Diablo-D3> g4 1.25ghz, 2gb of memory, a brand new drive
 421 2011-07-24 00:50:39 <jrmithdobbs> BlueMatt: for work i use a windows 7 box and spend 99% of my time in the debian virtual box vm instance (shell)
 422 2011-07-24 00:50:47 <Diablo-D3> run runs insanely fast under debian + xfce
 423 2011-07-24 00:50:52 <Diablo-D3> I mean, really really really fucking fast
 424 2011-07-24 00:50:54 Shuro__ has joined
 425 2011-07-24 00:51:06 <BlueMatt> anything runs fast on xfce
 426 2011-07-24 00:51:11 <Diablo-D3> as long as its not outright computation (such as, say, decoding a bluray), its fast
 427 2011-07-24 00:51:18 <jrmithdobbs> Diablo-D3: stop being poor and upgrade hardware </typical apple user response>
 428 2011-07-24 00:51:21 <Diablo-D3> under osx? ZOMG, ITS SO FUCKING SLOW
 429 2011-07-24 00:51:21 viggi_ has joined
 430 2011-07-24 00:51:27 <Diablo-D3> jrmithdobbs: heh, I wasnt buying a mac
 431 2011-07-24 00:51:31 lolwat` has quit (Remote host closed the connection)
 432 2011-07-24 00:51:32 <Diablo-D3> I needed a new laptop, and I needed a big endian box
 433 2011-07-24 00:51:36 <Diablo-D3> I killed two birds with one stone
 434 2011-07-24 00:51:37 <BlueMatt> jrmithdobbs: why dont you just run debian then?
 435 2011-07-24 00:51:42 <Diablo-D3> paid $500 for it used like 4-5 years ago
 436 2011-07-24 00:51:55 <Diablo-D3> jrmithdobbs: I tried running osx for an entire year on it
 437 2011-07-24 00:51:59 <Diablo-D3> no debian installed
 438 2011-07-24 00:52:04 <Diablo-D3> I forced myself to do it
 439 2011-07-24 00:52:13 <Diablo-D3> by the end of the year, it was an osx-looking unix box.
 440 2011-07-24 00:52:30 <jrmithdobbs> BlueMatt: because I can't *stand* the abominations foss people come up with for guis
 441 2011-07-24 00:52:38 <BlueMatt> osx is really not bad at all once you get used to it, but i still prefer linux
 442 2011-07-24 00:52:45 <Diablo-D3> the osx gui is great
 443 2011-07-24 00:52:47 <BlueMatt> jrmithdobbs: they are getting rapidly better
 444 2011-07-24 00:52:49 <Diablo-D3> its the underpinnings that suck dick
 445 2011-07-24 00:52:58 <jrmithdobbs> BlueMatt: i've been hearing that for 10 years (not exageration)
 446 2011-07-24 00:53:05 <denisx> i want my two-dimensional spaces back!
 447 2011-07-24 00:53:13 <BlueMatt> I prefer my compiz now
 448 2011-07-24 00:53:18 sacarlson has joined
 449 2011-07-24 00:53:22 <BlueMatt> some of the plugins are ridiculously useful
 450 2011-07-24 00:53:25 WildSoil has joined
 451 2011-07-24 00:53:31 <BlueMatt> (on large-screen or multi-screen setups
 452 2011-07-24 00:53:33 <BlueMatt> )
 453 2011-07-24 00:53:33 <WildSoil> ;;bc,stats
 454 2011-07-24 00:53:35 phantomcircuit_ has joined
 455 2011-07-24 00:53:38 <jrmithdobbs> BlueMatt: also, when i actually did use linux on the desktop, i ran window maker
 456 2011-07-24 00:53:45 <jrmithdobbs> BlueMatt: so os x is basically a wet dream ;p
 457 2011-07-24 00:54:01 <BlueMatt> lol
 458 2011-07-24 00:54:02 fnord0 has quit (Ping timeout: 250 seconds)
 459 2011-07-24 00:54:18 <jrmithdobbs> BlueMatt: i don't care what steve has decided
 460 2011-07-24 00:54:19 uber` has joined
 461 2011-07-24 00:54:27 <jrmithdobbs> BlueMatt: the dock lives on the righthand side and should always be visible goddamn it
 462 2011-07-24 00:54:46 <BlueMatt> I prefer left-hand-side, but whatever
 463 2011-07-24 00:55:00 <Diablo-D3> jrmithdobbs: I think the problem with osx, btw, is the UI acceleration is fucked
 464 2011-07-24 00:55:03 <jrmithdobbs> (though the pretty zoom animations and changable on the fly icon size is an awesome enhancement over NeXT)
 465 2011-07-24 00:55:06 <BlueMatt> when I used mac dock was always lhs and always visible, 32-pixel wide
 466 2011-07-24 00:55:12 MattJD has joined
 467 2011-07-24 00:55:12 <Diablo-D3> and on ppc, 10.5 is MUCH slower than 10.4
 468 2011-07-24 00:55:17 <jrmithdobbs> Diablo-D3: have you used 10.6?
 469 2011-07-24 00:55:18 <Diablo-D3> and a lot of apps wont run on 10.4
 470 2011-07-24 00:55:21 <Diablo-D3> jrmithdobbs: hello, ppc
 471 2011-07-24 00:55:23 phantomcircuit has quit (Disconnected by services)
 472 2011-07-24 00:55:23 <BlueMatt> Diablo-D3: they gave up on PPC, thats why
 473 2011-07-24 00:55:27 phantomcircuit_ is now known as phantomcircuit
 474 2011-07-24 00:55:30 <jrmithdobbs> Diablo-D3: right, they stopped optimizing for ppc
 475 2011-07-24 00:55:36 <senseles> are you guys talking about OSX/
 476 2011-07-24 00:55:42 <senseles> osx sucks my left nut
 477 2011-07-24 00:55:45 <Diablo-D3> jrmithdobbs: yeah, that excuse never made sense to me
 478 2011-07-24 00:55:45 <jrmithdobbs> Diablo-D3: all that stuff that was annoying in 10.5? ya it's fixed in 10.6 on x86 ;p
 479 2011-07-24 00:55:49 <senseles> apple is incapable of following standards
 480 2011-07-24 00:55:59 jimon has joined
 481 2011-07-24 00:56:00 <jrmithdobbs> Diablo-D3: no 10.5 had those issues on x86 too
 482 2011-07-24 00:56:02 <Diablo-D3> jrmithdobbs: gcc accepts the same flags it has for a decade
 483 2011-07-24 00:56:09 <Diablo-D3> gcc -march=ppc, -march=686
 484 2011-07-24 00:56:11 <Diablo-D3> same shit
 485 2011-07-24 00:56:21 <Diablo-D3> its not like they had reams of assembly that they removed
 486 2011-07-24 00:56:23 <jrmithdobbs> apple's a hardware company
 487 2011-07-24 00:56:26 <senseles> just do yourself a favor and format/reinstall windows :/
 488 2011-07-24 00:56:29 <jrmithdobbs> they write software to sell their hardware
 489 2011-07-24 00:56:29 <Diablo-D3> apple is a SOLUTION compan
 490 2011-07-24 00:56:31 <jrmithdobbs> mystery solved
 491 2011-07-24 00:56:32 <BlueMatt> yea, osx kernel/drivers/etc can tend to be shit
 492 2011-07-24 00:56:35 <jrmithdobbs> Diablo-D3: nope.
 493 2011-07-24 00:56:40 <Diablo-D3> never EVER think they're a hardware company
 494 2011-07-24 00:56:43 <Diablo-D3> same as cisco
 495 2011-07-24 00:56:46 <Diablo-D3> cisco doesnt sell routers
 496 2011-07-24 00:56:48 <Diablo-D3> they sell solutions
 497 2011-07-24 00:56:53 <senseles> Diablo-d3: solutions that don't work? :p
 498 2011-07-24 00:56:53 <jrmithdobbs> we sell support contracts
 499 2011-07-24 00:56:55 <jrmithdobbs> ;p
 500 2011-07-24 00:57:01 peck has joined
 501 2011-07-24 00:57:02 <Diablo-D3> like I said, solutions
 502 2011-07-24 00:57:17 <Diablo-D3> theres a bunch of companies like apple and cisco
 503 2011-07-24 00:57:21 <senseles> try to setup an L2TP vpn client on an OSX client over wireless from a Mac Airport device
 504 2011-07-24 00:57:26 <Diablo-D3> totally vertical, control all aspects of the solution
 505 2011-07-24 00:57:27 <senseles> a fun time will ensue
 506 2011-07-24 00:57:29 <jrmithdobbs> Diablo-D3: no, apple actually loses money on all of their "services"/"solutions" divisions
 507 2011-07-24 00:57:34 <jrmithdobbs> Diablo-D3: they really are a hardware company
 508 2011-07-24 00:57:35 <Diablo-D3> jrmithdobbs: "loses"
 509 2011-07-24 00:57:35 <BlueMatt> osx gui is nice, their dev tools are nice, they have a good devel community that churns out pretty apps, but the os core is really just not up to spec...
 510 2011-07-24 00:57:43 <Diablo-D3> jrmithdobbs: just like microsoft "loses" money on every 360 sold
 511 2011-07-24 00:57:50 <BlueMatt> jrmithdobbs: they make a shitton on each app sold
 512 2011-07-24 00:57:54 Bwild has joined
 513 2011-07-24 00:58:00 <cjdelisle> good observation re solutions. kind of like motorcycle companies that don't sell a bike, they sell a way of life.
 514 2011-07-24 00:58:05 <Diablo-D3> jrmithdobbs: its just that, I cant buy a mac without osx
 515 2011-07-24 00:58:10 <Diablo-D3> and I generally cant buy osx without a mac
 516 2011-07-24 00:58:22 <jrmithdobbs> Diablo-D3: actually they basically give away osx, even upgrades
 517 2011-07-24 00:58:23 <senseles> ?? you can buy osx software online
 518 2011-07-24 00:58:27 <BlueMatt> Diablo-D3: you know you can run osx in virtualbox with no bootloaders/drivers/etc
 519 2011-07-24 00:58:33 <Diablo-D3> bluematt: wrong
 520 2011-07-24 00:58:37 <Diablo-D3> ONLY the server edition
 521 2011-07-24 00:58:41 <BlueMatt> Diablo-D3: right, Im doing it now
 522 2011-07-24 00:58:43 <senseles> bluematt: how? i tried to setup OSX in vmware could not get it working
 523 2011-07-24 00:58:45 <BlueMatt> well ok, legal blah blah
 524 2011-07-24 00:58:48 <Diablo-D3> the server edition is fucking expensive
 525 2011-07-24 00:58:53 <phantomcircuit> ok forums are officially worthless
 526 2011-07-24 00:58:54 <jrmithdobbs> Diablo-D3: not any more
 527 2011-07-24 00:59:07 <Diablo-D3> bluematt: I had one of the most used osx-in-virtualbox guides on the net before vbox officially supported osx
 528 2011-07-24 00:59:08 <jrmithdobbs> Diablo-D3: server edition is now an add-on pack to base os and is $49
 529 2011-07-24 00:59:11 <phantomcircuit> there are no replies to any of the topics i've posted in
 530 2011-07-24 00:59:13 <jrmithdobbs> Diablo-D3: base os is now $29
 531 2011-07-24 00:59:20 <BlueMatt> though I spent all last night hacking virtualbox and osx wont finish loading...at least I got the kernel to load
 532 2011-07-24 00:59:22 Clipse has quit (Ping timeout: 255 seconds)
 533 2011-07-24 00:59:26 <BlueMatt> (for 10.6)
 534 2011-07-24 00:59:36 <Diablo-D3> jrmithdobbs: yeah, but, for example, can I get full 2D accel on a linux host in vbox?
 535 2011-07-24 00:59:37 <jrmithdobbs> Diablo-D3: base os license includes wording that allows installation on any machine associated with your appleid
 536 2011-07-24 00:59:38 <Diablo-D3> nope, I cant.
 537 2011-07-24 00:59:42 <denisx> osx server is 40$
 538 2011-07-24 00:59:45 <jrmithdobbs> Diablo-D3: including 2 vms per machine
 539 2011-07-24 00:59:54 <jrmithdobbs> Diablo-D3: like I said, they're *giving it away* now
 540 2011-07-24 00:59:54 <denisx> oh, 40€
 541 2011-07-24 01:00:02 <Diablo-D3> because I wouldnt mind OSX in vbox if it ran like a mac
 542 2011-07-24 01:00:06 <BlueMatt> jrmithdobbs: 40$ >> 0$
 543 2011-07-24 01:00:17 <BlueMatt> Diablo-D3: ?
 544 2011-07-24 01:00:19 <Diablo-D3> bluematt: I shit out larger denomination bills, and Im poor as a church mouse.
 545 2011-07-24 01:00:29 <senseles> Bluematt: good luck, i tried for days to get OSX fully working in vmware could not. It was easier to just buy a crappy mac mini off of ebay for 100$.
 546 2011-07-24 01:00:34 <BlueMatt> Diablo-D3: meh, us students are cheap-asses
 547 2011-07-24 01:00:36 <Diablo-D3> bluematt: ever try to make a hackintosh on decidedly un-mac hardware?
 548 2011-07-24 01:00:37 <jrmithdobbs> BlueMatt: show me another commercial unix i can get for <$30usd
 549 2011-07-24 01:00:44 <Diablo-D3> it sucks
 550 2011-07-24 01:00:45 <Diablo-D3> it really
 551 2011-07-24 01:00:46 <Diablo-D3> really
 552 2011-07-24 01:00:46 <Diablo-D3> really
 553 2011-07-24 01:00:47 <Diablo-D3> sucks
 554 2011-07-24 01:00:56 <senseles> jrmithdobbs: centos is free
 555 2011-07-24 01:01:00 <BlueMatt> senseles: no, I have 10.5 and 6 working really well, its lion's kernel thats being a bitch
 556 2011-07-24 01:01:03 <jrmithdobbs> senseles: i said commercial
 557 2011-07-24 01:01:08 <BlueMatt> jrmithdobbs: who the hell uses commercial linux?
 558 2011-07-24 01:01:15 <jrmithdobbs> I didn't say linux
 559 2011-07-24 01:01:17 <jrmithdobbs> i said unix
 560 2011-07-24 01:01:18 <Diablo-D3> if I can throw osx in vbox and have that be a fucking real honest to fucking god mac
 561 2011-07-24 01:01:19 <senseles> jrmithdobbs: works on my production servers just fine :p
 562 2011-07-24 01:01:24 <Diablo-D3> I'd buy it
 563 2011-07-24 01:01:26 <senseles> why would you want unix?
 564 2011-07-24 01:01:27 <BlueMatt> Diablo-D3: I run osx on vbox now and its running perfectly?
 565 2011-07-24 01:01:30 <senseles> go install solaris i guess
 566 2011-07-24 01:01:35 <Diablo-D3> bluematt: what about 2D accel?
 567 2011-07-24 01:01:38 <Diablo-D3> can I tell its not a mac?
 568 2011-07-24 01:01:41 <Diablo-D3> does sound work yet?
 569 2011-07-24 01:01:47 <BlueMatt> sound works
 570 2011-07-24 01:01:53 <BlueMatt> and no lags/graphic lags
 571 2011-07-24 01:01:57 <BlueMatt> ok you cant game or anything
 572 2011-07-24 01:02:03 <Diablo-D3> I dont need to game
 573 2011-07-24 01:02:07 <Diablo-D3> but does video work right too?
 574 2011-07-24 01:02:07 <sacarlson> jrmithdobbs: I think sun solaris or now oricle has a unix for <$30
 575 2011-07-24 01:02:14 <BlueMatt> yea, video works perfectly
 576 2011-07-24 01:02:24 <Diablo-D3> like, I can I buy itunes mvoies
 577 2011-07-24 01:02:26 <Diablo-D3> and play them?
 578 2011-07-24 01:02:32 <senseles> do be people still use solaris, irix, etc?
 579 2011-07-24 01:02:38 <Diablo-D3> senseles: yes.
 580 2011-07-24 01:02:40 <BlueMatt> well havent done too much of that...but I can check
 581 2011-07-24 01:02:41 <senseles> for what?
 582 2011-07-24 01:02:42 <jrmithdobbs> not irix no
 583 2011-07-24 01:02:45 <Diablo-D3> although oracle killed solaris
 584 2011-07-24 01:02:46 <jrmithdobbs> irix is basically dead
 585 2011-07-24 01:02:51 <Diablo-D3> jrmithdobbs: irix yes.... but as existing installations
 586 2011-07-24 01:02:54 <jrmithdobbs> hp-ux, solaris, and aix are alive
 587 2011-07-24 01:02:59 <Diablo-D3> even "new" irix installs are for existing shit
 588 2011-07-24 01:03:01 <jrmithdobbs> and going strong
 589 2011-07-24 01:03:03 <Diablo-D3> same goes with hpux and aix
 590 2011-07-24 01:03:04 mrb_ has quit (Ping timeout: 276 seconds)
 591 2011-07-24 01:03:08 <jrmithdobbs> hp-ux is actually not bad at all
 592 2011-07-24 01:03:10 mrb_ has joined
 593 2011-07-24 01:03:13 <Diablo-D3> oracle is trying to stab solaris in the face until it stops moving though
 594 2011-07-24 01:03:20 <jrmithdobbs> good
 595 2011-07-24 01:03:21 <Diablo-D3> they shit on the entire solaris community
 596 2011-07-24 01:03:24 <senseles> so its just existing corporate/college clients that are forced with supporting an existing system
 597 2011-07-24 01:03:26 <Diablo-D3> bad, imo
 598 2011-07-24 01:03:27 <jrmithdobbs> solaris is the worst of the commercial unices anyways
 599 2011-07-24 01:03:38 <senseles> like all those corporations still trying to migrate from cobol to 2011
 600 2011-07-24 01:03:38 <Diablo-D3> they killed basically every open solaris effort
 601 2011-07-24 01:03:39 <jrmithdobbs> only thing good to come out of solaris in the last 15 years were dtrace and zfs
 602 2011-07-24 01:03:40 <[Tycho]> irix had nice interface style. I don't want to use it, but it looks stylish :)
 603 2011-07-24 01:04:02 <Diablo-D3> jrmithdobbs: yeah but zfs begat btrfs, and now oracle basically abandoned btrfs
 604 2011-07-24 01:04:16 <Diablo-D3> and btrfs was supposed to be the post-ext4 canidate
 605 2011-07-24 01:04:26 <jrmithdobbs> aware
 606 2011-07-24 01:04:42 <Diablo-D3> I cant help stare at Oracle and go mymindisfulloffuck.jpg
 607 2011-07-24 01:04:49 <jrmithdobbs> i conceded that dtrace and zfs were awesome ;p
 608 2011-07-24 01:04:56 fnord0 has joined
 609 2011-07-24 01:04:59 <Diablo-D3> yes, but what Im saying is
 610 2011-07-24 01:05:05 <Diablo-D3> who cares if its awesome if I cant use it
 611 2011-07-24 01:05:05 <jrmithdobbs> oracle is now trying to kill openfirmware
 612 2011-07-24 01:05:10 <jrmithdobbs> :(
 613 2011-07-24 01:05:12 <Diablo-D3> hah
 614 2011-07-24 01:05:16 <Diablo-D3> openfirmware will be alive forever
 615 2011-07-24 01:05:26 <denisx> I'm using zfs all the time...
 616 2011-07-24 01:05:31 AStove has quit (Ping timeout: 257 seconds)
 617 2011-07-24 01:05:31 <jrmithdobbs> they're basically killing sparc on their mid-to-low range
 618 2011-07-24 01:05:33 <senseles> oracle is great at killing stuff, they killed one of my ex girlfriend's fathers jobs in sales
 619 2011-07-24 01:05:36 <Diablo-D3> ibm uses it, apple uses it, intel's.... whatever the fuck it is of theirs uses it
 620 2011-07-24 01:05:41 <jrmithdobbs> someone needs to port OF to EFI
 621 2011-07-24 01:05:42 <jrmithdobbs> lol
 622 2011-07-24 01:05:43 <senseles> which killed his 400,000$ house he just built
 623 2011-07-24 01:05:44 <jrmithdobbs> <3 OF
 624 2011-07-24 01:05:47 <Diablo-D3> EFI _is_ OF
 625 2011-07-24 01:05:52 <Diablo-D3> its just a fucking insane impl of it
 626 2011-07-24 01:05:59 <senseles> and he had to sell and move out of a house he just built after losing his job :/
 627 2011-07-24 01:06:01 <senseles> sucks
 628 2011-07-24 01:06:01 <Diablo-D3> and has a manidtory x86 bios compat kit thrown on
 629 2011-07-24 01:06:11 <jrmithdobbs> Diablo-D3: I want a go prompt goddamn it
 630 2011-07-24 01:06:20 <[Tycho]> senseles, why sell ?
 631 2011-07-24 01:06:24 <Diablo-D3> jrmithdobbs: oh, the OF command line shit?
 632 2011-07-24 01:06:33 <Diablo-D3> jrmithdobbs: EFI doesnt provide it
 633 2011-07-24 01:06:42 <jrmithdobbs> "command line shit" == "interactive forth interpreter" and yes
 634 2011-07-24 01:06:44 <jrmithdobbs> i wants it
 635 2011-07-24 01:06:45 <Diablo-D3> actually, Im not even sure if it even has a forth interpre... yeah
 636 2011-07-24 01:06:47 <senseles> [tycho]: no more job oracle laid him off. he was a sales guy, and one of their top guys. surprised everyone involved.
 637 2011-07-24 01:06:55 <Diablo-D3> senseles: dude
 638 2011-07-24 01:06:55 <jrmithdobbs> Diablo-D3: then it's not OF
 639 2011-07-24 01:06:59 <Diablo-D3> Im not surprised
 640 2011-07-24 01:07:03 <Diablo-D3> jrmithdobbs: I dunno, intel says it is
 641 2011-07-24 01:07:06 <Diablo-D3> but fuck intel
 642 2011-07-24 01:07:06 <[Tycho]> senseles, but why sell the house ?
 643 2011-07-24 01:07:08 <jrmithdobbs> intel lies a lot
 644 2011-07-24 01:07:10 <Diablo-D3> fuck them hard and long
 645 2011-07-24 01:07:10 <senseles> no money
 646 2011-07-24 01:07:14 <Diablo-D3> [Tycho]: mortgaged.
 647 2011-07-24 01:07:18 Astrohacker has joined
 648 2011-07-24 01:07:26 <Diablo-D3> senseles: I know a lot of Java people that are and used to be at Sun
 649 2011-07-24 01:07:33 <Diablo-D3> senseles: oracle simply doesnt give a fuck
 650 2011-07-24 01:07:43 <Diablo-D3> they chased out their most important java devs
 651 2011-07-24 01:07:44 <jrmithdobbs> senseles: he deserved to lose his job
 652 2011-07-24 01:07:49 <jrmithdobbs> senseles: for worknig for that fucking company
 653 2011-07-24 01:07:51 <jrmithdobbs> fuck oracle ;p
 654 2011-07-24 01:07:53 <senseles> i wish i was a mega corporation so i could not give a fuck, that must be a nice life
 655 2011-07-24 01:07:56 <denisx> even most of the zfs programmers left oracle
 656 2011-07-24 01:07:58 <Diablo-D3> jrmithdobbs: he was probably a sun employee
 657 2011-07-24 01:08:08 <jrmithdobbs> Diablo-D3: does not change my statement
 658 2011-07-24 01:08:17 <Diablo-D3> I didnt mind the Sun guys
 659 2011-07-24 01:08:18 <jrmithdobbs> i hated sun as much as i hate oracle tbqh
 660 2011-07-24 01:08:26 <Diablo-D3> like, I would hump a Thumper given the chance
 661 2011-07-24 01:08:38 <Diablo-D3> I dont care if I can build one using supermicro's one new rig
 662 2011-07-24 01:08:44 <Diablo-D3> its just a sexy box
 663 2011-07-24 01:08:53 <[Tycho]> It's hard to understand those loan/credit people...
 664 2011-07-24 01:08:54 <jrmithdobbs> there hardware was always pretty nice
 665 2011-07-24 01:08:59 <jrmithdobbs> but their software
 666 2011-07-24 01:09:09 <jrmithdobbs> jesus christ drop you broken userland and make people fix their retarded broken scripts
 667 2011-07-24 01:09:12 <jrmithdobbs> for the love of fucking god
 668 2011-07-24 01:09:44 <jrmithdobbs> rsync -aP /usr/xpg4/bin/. /usr/bin/.; rsync -aP /usr/xpg4/sbin /usr/sbin/.;
 669 2011-07-24 01:09:48 <jrmithdobbs> g!
 670 2011-07-24 01:09:53 <jrmithdobbs> s/g//
 671 2011-07-24 01:10:03 <Diablo-D3> heh
 672 2011-07-24 01:10:06 <jrmithdobbs> THERE YOU GO SUN/ORACLE I JUST FIXED YOUR OS, YOU'RE WELCOME
 673 2011-07-24 01:10:07 <Diablo-D3> wait, why rsync?
 674 2011-07-24 01:10:09 <senseles> wasn't irix an alpha/dec product?
 675 2011-07-24 01:10:11 <Diablo-D3> ln -s
 676 2011-07-24 01:10:25 <Diablo-D3> jrmithdobbs: dude, you know what the worst problem in solaris is?
 677 2011-07-24 01:10:32 <Diablo-D3> unix style net if naming.
 678 2011-07-24 01:10:33 <jrmithdobbs> besides it being solaris?
 679 2011-07-24 01:10:41 <denisx> irix was sgi
 680 2011-07-24 01:10:58 <Diablo-D3> just name it fucking eth0 and get it over with goddamnit
 681 2011-07-24 01:11:01 <jrmithdobbs> god, i want to plubm whoever's ass who designed the solaris net if subsystem
 682 2011-07-24 01:11:07 <jrmithdobbs> s/plubm/plumb/
 683 2011-07-24 01:11:14 <jrmithdobbs> and the ip assignment crap
 684 2011-07-24 01:11:18 <jrmithdobbs> and their nss
 685 2011-07-24 01:11:22 <jrmithdobbs> and pam implementation
 686 2011-07-24 01:11:23 Zagitta has quit (Ping timeout: 252 seconds)
 687 2011-07-24 01:11:26 <denisx> jrmithdobbs: haha, yeah, I remember those...
 688 2011-07-24 01:11:51 <Diablo-D3> and the *. all of it.
 689 2011-07-24 01:11:51 <jrmithdobbs> and their pkg* crap
 690 2011-07-24 01:11:54 <Diablo-D3> fuck
 691 2011-07-24 01:11:56 <Diablo-D3> lets just use debian
 692 2011-07-24 01:12:00 <Diablo-D3> Ive used debian for over a decade
 693 2011-07-24 01:12:01 <Diablo-D3> and
 694 2011-07-24 01:12:02 <Diablo-D3> well
 695 2011-07-24 01:12:05 <Diablo-D3> Ive used debian for over a decade
 696 2011-07-24 01:12:09 <jrmithdobbs> and all the fuckin java crap they heaped on top of it that no decent admin ever uses anyways
 697 2011-07-24 01:12:32 <Diablo-D3> jrmithdobbs: btw, you do know I only started using java once sun let go of it, right?
 698 2011-07-24 01:12:46 <Diablo-D3> pre-foss java, I hated it just like everyone else
 699 2011-07-24 01:13:06 <jrmithdobbs> that wasn't a stab at you/java but their horribly implemented "administration tools" written in java
 700 2011-07-24 01:13:09 <jrmithdobbs> ;p
 701 2011-07-24 01:13:15 <jrmithdobbs> but I don't like java either, fwiw
 702 2011-07-24 01:13:17 <jrmithdobbs> lol
 703 2011-07-24 01:13:41 <Diablo-D3> wait, solaris had admin tools?
 704 2011-07-24 01:16:23 <jrmithdobbs> ya and if you ever tried to report a problem to them about something they would get all beligerent if you weren't using them
 705 2011-07-24 01:16:26 <jrmithdobbs> haha
 706 2011-07-24 01:17:11 <jrmithdobbs> much like rh does
 707 2011-07-24 01:17:22 <senseles> just imagine being RH though
 708 2011-07-24 01:17:28 <senseles> the kind of calls you get
 709 2011-07-24 01:17:29 <jrmithdobbs> actually, rh isn't nearly as bad about it
 710 2011-07-24 01:17:49 <jrmithdobbs> senseles: ya i don't call them, i figured out a long time ago the only real support mechanism is through the dev bug tracker
 711 2011-07-24 01:17:55 <senseles> Probably not as bad but on the level with "my winders dunt werk"
 712 2011-07-24 01:17:58 <jrmithdobbs> because so long as you're not retarded you get real help
 713 2011-07-24 01:18:01 <jrmithdobbs> AND it's free, lol
 714 2011-07-24 01:18:33 <senseles> Pisses me off so bad when people say "it don't work". As if that's helpful.
 715 2011-07-24 01:18:34 <jrmithdobbs> Diablo-D3: also, java the language/stdlib isn't horrible, sun's jvm and derivatives can go DIAF though
 716 2011-07-24 01:18:41 <jrmithdobbs> Diablo-D3: and all thei j2ee crap too
 717 2011-07-24 01:18:59 <jrmithdobbs> Diablo-D3: especially the ESB shit, fuck that retardedness
 718 2011-07-24 01:19:13 <jrmithdobbs> basically
 719 2011-07-24 01:19:20 <jrmithdobbs> i hate most computer software
 720 2011-07-24 01:19:34 <jrmithdobbs> lets all start doing aes-xts on abacuses
 721 2011-07-24 01:22:22 mrb_ has quit (Ping timeout: 252 seconds)
 722 2011-07-24 01:23:35 mrb_ has joined
 723 2011-07-24 01:24:00 m86 has joined
 724 2011-07-24 01:25:53 <Diablo-D3> jrmithdobbs: j2ee is dead
 725 2011-07-24 01:26:01 Sedra has joined
 726 2011-07-24 01:26:04 <Diablo-D3> jee6 basically sank all the old shit
 727 2011-07-24 01:27:06 zeropointo has quit (Quit: leaving)
 728 2011-07-24 01:29:20 Sedra- has quit (Ping timeout: 264 seconds)
 729 2011-07-24 01:30:39 <jrmithdobbs> Diablo-D3: ya but retarded java devs are still using it
 730 2011-07-24 01:30:53 <jrmithdobbs> i know it's *supposed* to be dead
 731 2011-07-24 01:31:02 <jrmithdobbs> where the fuck did cryptography@metzdowd.com move to
 732 2011-07-24 01:31:06 <jrmithdobbs> anyone know?
 733 2011-07-24 01:35:22 RobinPKR_ has joined
 734 2011-07-24 01:35:45 <Diablo-D3> jrmithdobbs: well, they're technically NOT using it
 735 2011-07-24 01:35:52 <Diablo-D3> seeing as its not exactly supported
 736 2011-07-24 01:36:01 <Diablo-D3> I mean, ever see how sun support contracts work?
 737 2011-07-24 01:36:07 <Diablo-D3> if its not in the contract, it doesnt exist
 738 2011-07-24 01:36:51 <Diablo-D3> jrmithdobbs: btw, as a side note
 739 2011-07-24 01:36:58 <Diablo-D3> no one gives a fuck about jee on sun
 740 2011-07-24 01:37:12 <Diablo-D3> glassfish is slow, out of date, and dumb
 741 2011-07-24 01:37:18 Zarutian has joined
 742 2011-07-24 01:37:19 <Diablo-D3> and it also might be gay
 743 2011-07-24 01:37:25 <Diablo-D3> not that theres anything wrong with that
 744 2011-07-24 01:37:45 <jrmithdobbs> weblogic and jboss are no better though
 745 2011-07-24 01:37:49 RobinPKR has quit (Ping timeout: 250 seconds)
 746 2011-07-24 01:37:49 RobinPKR_ is now known as RobinPKR
 747 2011-07-24 01:39:07 <Diablo-D3> jrmithdobbs: actually, as7 is fucking awesome
 748 2011-07-24 01:39:18 <Diablo-D3> and I should smack you for even mentioning weblogic
 749 2011-07-24 01:39:36 <Diablo-D3> I'd rather be stuck on glassfish then even have to admit weblogic exists
 750 2011-07-24 01:41:43 gribble has joined
 751 2011-07-24 01:43:37 Burgundy has joined
 752 2011-07-24 01:45:33 freakazoid has quit (Quit: Computer has gone to sleep.)
 753 2011-07-24 01:48:22 BitcoinForNewegg has joined
 754 2011-07-24 01:52:29 osmosis has quit (Ping timeout: 258 seconds)
 755 2011-07-24 01:54:54 Clipse has joined
 756 2011-07-24 01:55:31 underscor has joined
 757 2011-07-24 02:08:59 erus` has quit (Remote host closed the connection)
 758 2011-07-24 02:13:29 [7] is now known as {
 759 2011-07-24 02:13:36 denisx has quit (Quit: denisx)
 760 2011-07-24 02:14:24 phungus has joined
 761 2011-07-24 02:14:50 harrigan1 has joined
 762 2011-07-24 02:16:31 harrigan has quit (Ping timeout: 276 seconds)
 763 2011-07-24 02:19:34 arthurb has quit (Ping timeout: 276 seconds)
 764 2011-07-24 02:19:43 TheSeven has joined
 765 2011-07-24 02:21:34 { has quit (Ping timeout: 260 seconds)
 766 2011-07-24 02:21:56 citiz3n has quit (Ping timeout: 246 seconds)
 767 2011-07-24 02:23:00 fingster_ has joined
 768 2011-07-24 02:23:19 <fingster_> hi
 769 2011-07-24 02:23:35 cuqa has joined
 770 2011-07-24 02:23:38 cuqa has quit (Changing host)
 771 2011-07-24 02:23:38 cuqa has joined
 772 2011-07-24 02:23:57 <fingster_> anybody interested in a thin client?
 773 2011-07-24 02:24:26 <fingster_> without downloading any blocks or headers
 774 2011-07-24 02:26:22 asuk has quit (Quit: leaving)
 775 2011-07-24 02:27:14 huk has quit (Ping timeout: 246 seconds)
 776 2011-07-24 02:27:23 cuqaa has quit (Ping timeout: 260 seconds)
 777 2011-07-24 02:27:30 <jrmithdobbs> fingster_: no because that means you're probably lieing
 778 2011-07-24 02:27:40 <jrmithdobbs> at least, it can't be validated byt the client that you're not
 779 2011-07-24 02:27:58 <fingster_> try it
 780 2011-07-24 02:28:01 <fingster_> :)
 781 2011-07-24 02:28:07 <jrmithdobbs> try what
 782 2011-07-24 02:28:16 <fingster_> my client
 783 2011-07-24 02:28:22 <BlueMatt> link?
 784 2011-07-24 02:28:24 MobiusL is now known as Enoch
 785 2011-07-24 02:28:27 <jrmithdobbs> you privatekey stealer?
 786 2011-07-24 02:28:29 <jrmithdobbs> ;p
 787 2011-07-24 02:28:32 <jrmithdobbs> s/you/your/
 788 2011-07-24 02:28:38 <BlueMatt> (to source, not binary)
 789 2011-07-24 02:28:45 <fingster_> I will make open source
 790 2011-07-24 02:28:55 arthurb has joined
 791 2011-07-24 02:29:31 <senseles> i guess i should add the source to my centos release
 792 2011-07-24 02:29:35 Enoch is now known as MobiusL
 793 2011-07-24 02:29:58 MobiusL is now known as Enoch
 794 2011-07-24 02:30:02 <senseles> when is 0.3.25 due out?
 795 2011-07-24 02:30:10 <senseles> how is the wallet encryption coming along
 796 2011-07-24 02:30:24 nullrouten has quit (Ping timeout: 246 seconds)
 797 2011-07-24 02:30:26 Enoch is now known as MobiusL
 798 2011-07-24 02:30:40 <BlueMatt> .3.25 is set to never come out, instead when wallet import/export gets merged it becomes 0.4
 799 2011-07-24 02:31:56 clarkbox has quit (Ping timeout: 246 seconds)
 800 2011-07-24 02:32:15 <jrmithdobbs> BlueMatt: how did you end up doing the salt storage btw?
 801 2011-07-24 02:32:21 <senseles> any etas? would be really nice to have that
 802 2011-07-24 02:32:30 <BlueMatt> jrmithdobbs: random salt stored in wallet
 803 2011-07-24 02:32:49 <BlueMatt> 8-bytes as that is all openssl-evp will use
 804 2011-07-24 02:32:49 Stellar has quit (Ping timeout: 246 seconds)
 805 2011-07-24 02:32:52 <jrmithdobbs> BlueMatt: i mean specifics of storage, it's not fixed length or anything right?
 806 2011-07-24 02:32:52 clarkbox has joined
 807 2011-07-24 02:33:08 <BlueMatt> dont remember tbh...
 808 2011-07-24 02:33:20 <jrmithdobbs> where is that in the code i'll look ;p
 809 2011-07-24 02:33:28 <BlueMatt> no its a vector
 810 2011-07-24 02:33:31 <BlueMatt> so it can be any length
 811 2011-07-24 02:33:34 <jrmithdobbs> k
 812 2011-07-24 02:33:38 <BlueMatt> its the very top of crypter.h
 813 2011-07-24 02:34:07 <jrmithdobbs> and bdb is just using the serializer stuff so it doesn't set any limits right?
 814 2011-07-24 02:35:28 <BlueMatt> yea
 815 2011-07-24 02:35:39 <fingster_> If a client can't verify a received transaction, is it a promblem?
 816 2011-07-24 02:35:45 <jrmithdobbs> yes
 817 2011-07-24 02:35:46 huk has joined
 818 2011-07-24 02:35:55 <jrmithdobbs> a big one
 819 2011-07-24 02:36:19 <jrmithdobbs> especially for a thin client. see: http://en.wikipedia.org/wiki/Sybil_attack
 820 2011-07-24 02:36:26 <fingster_> but the client won't lose anything
 821 2011-07-24 02:36:33 <jrmithdobbs> especially for a thin client. see: http://en.wikipedia.org/wiki/Sybil_attack
 822 2011-07-24 02:37:00 <fingster_> reading...
 823 2011-07-24 02:37:42 lolwat` has joined
 824 2011-07-24 02:37:56 <jrmithdobbs> BlueMatt: cool gmaxwell convinced you on going ahead and adding the nDerivationMethod value
 825 2011-07-24 02:38:01 nullrouten has joined
 826 2011-07-24 02:38:13 <BlueMatt> dont remember who convinced me of that, but yea its there
 827 2011-07-24 02:38:23 <jrmithdobbs> ya not important
 828 2011-07-24 02:39:46 <fingster_> we set up a server, which is a normal bitcoin peer
 829 2011-07-24 02:40:16 <fingster_> thin clients connect to the server
 830 2011-07-24 02:40:19 BlueMatt has quit (Quit: Ex-Chat)
 831 2011-07-24 02:40:25 <jrmithdobbs> oh, you already know that such a client exists that just uses json-rpc right?
 832 2011-07-24 02:40:30 <jrmithdobbs> spelismo or whatever
 833 2011-07-24 02:40:43 <[Tycho]> spesmilo, imho
 834 2011-07-24 02:41:04 <fingster_> ?
 835 2011-07-24 02:41:06 <jrmithdobbs> fingster_: and that's not a thin client
 836 2011-07-24 02:41:15 <fingster_> no json-rpc
 837 2011-07-24 02:41:49 <fingster_> client keeps its wallet locally
 838 2011-07-24 02:41:50 <jrmithdobbs> well if it requires running a bitcoin/bitcoind instance it's not really a thin client
 839 2011-07-24 02:42:14 <jrmithdobbs> and you still need to download the blocks and headers
 840 2011-07-24 02:42:15 <AAA_awright> Couldn't a recieveing client potentially send a signal to the network to voluntarially undo a transaction before the next block? I can't think of a reason it's technically impossible
 841 2011-07-24 02:42:19 <jrmithdobbs> just on the bitcoin instance
 842 2011-07-24 02:42:22 <fingster_> move from p2p to cs
 843 2011-07-24 02:43:33 <fingster_> oh, your thin client definition is strong
 844 2011-07-24 02:43:59 huk has quit (Ping timeout: 246 seconds)
 845 2011-07-24 02:44:11 <jrmithdobbs> no your definition of "don't need to download blocks/headers" is weak
 846 2011-07-24 02:44:15 <fingster_> mine is running on mobi device, such as iPhone
 847 2011-07-24 02:44:41 nullrouten has quit (Ping timeout: 246 seconds)
 848 2011-07-24 02:44:55 jivvz has joined
 849 2011-07-24 02:45:06 <jrmithdobbs> so long as the client only uses configured trusted peers that sounds fine to me
 850 2011-07-24 02:45:34 <jrmithdobbs> i see what you're getting at, only requesting data from the bitcoin/bitcoind instance as it's necessary instead of storing it
 851 2011-07-24 02:45:42 <jrmithdobbs> but that requires implicit trust of the node you're connecting to
 852 2011-07-24 02:46:04 <fingster_> yeah
 853 2011-07-24 02:46:09 <jrmithdobbs> to the point where you might as well use json-rpc
 854 2011-07-24 02:46:24 <fingster_> but i don't see any harm to the client
 855 2011-07-24 02:46:25 <jrmithdobbs> since json-rpc has it's own listener pool
 856 2011-07-24 02:46:37 nullrouten has joined
 857 2011-07-24 02:46:44 <jrmithdobbs> well, if your trusted node has all it's connection slots full the client becomes unusable
 858 2011-07-24 02:47:05 <jrmithdobbs> but no, no harm so long as it is very clearly indicated that you need to implicitly trust the node you're connecting to
 859 2011-07-24 02:47:10 <fingster_> use p2p
 860 2011-07-24 02:47:22 <jrmithdobbs> yes, i understand how you've implemented
 861 2011-07-24 02:47:30 <fingster_> use extended bitcoin protocol
 862 2011-07-24 02:47:31 mmoya has joined
 863 2011-07-24 02:47:45 <jrmithdobbs> i'm saying that you might as well use json-rpc and let the bitcoin/bitcoind instance handle the keys at that point
 864 2011-07-24 02:47:53 <jrmithdobbs> since you already require an implicitly trusted node.
 865 2011-07-24 02:48:11 <fingster_> right
 866 2011-07-24 02:48:27 <jrmithdobbs> because then you have no key storage issues on the mobile device
 867 2011-07-24 02:48:36 <jrmithdobbs> which is a huge concern
 868 2011-07-24 02:48:45 <fingster_> totally
 869 2011-07-24 02:49:09 <jrmithdobbs> so I *personally* have no interest in it, but it's not really harmful ;p
 870 2011-07-24 02:49:32 <fingster_> what harm?
 871 2011-07-24 02:49:41 <jrmithdobbs> i said it's not?
 872 2011-07-24 02:49:44 <fingster_> I wanna improve it
 873 2011-07-24 02:49:46 <jrmithdobbs> not harmful
 874 2011-07-24 02:49:56 <fingster_> oh...
 875 2011-07-24 02:50:11 <jrmithdobbs> it's risky since you're storing privkeys on a mobile device though, and that complicates backups of them and such on most platforms
 876 2011-07-24 02:50:32 <jrmithdobbs> (besides the obvious security implications of lost/stolen phone/etc)
 877 2011-07-24 02:50:53 <fingster_> agree
 878 2011-07-24 02:51:21 nullrouten has quit (Ping timeout: 246 seconds)
 879 2011-07-24 02:51:30 <fingster_> what if such a client on PC?
 880 2011-07-24 02:51:45 huk has joined
 881 2011-07-24 02:51:53 <jrmithdobbs> not very useful
 882 2011-07-24 02:52:01 <fingster_> though power and storage is not problem
 883 2011-07-24 02:52:01 <jrmithdobbs> since you still have to run bitcoin
 884 2011-07-24 02:52:07 <jrmithdobbs> yes it is
 885 2011-07-24 02:52:19 <jrmithdobbs> because you still have to run a trusted bitcoin/bitcoind somewhere for it to connect to
 886 2011-07-24 02:52:48 <fingster_> if the remote server is not trusted?
 887 2011-07-24 02:52:57 <jrmithdobbs> then sybil attack applies
 888 2011-07-24 02:53:00 <fingster_> any harm?
 889 2011-07-24 02:53:19 <jrmithdobbs> since you store no state that you can validate you can't check the responses you receive from peers
 890 2011-07-24 02:53:23 <fingster_> lost btc?
 891 2011-07-24 02:53:49 <jrmithdobbs> possibly
 892 2011-07-24 02:54:03 <jrmithdobbs> though it would be a complicated attack
 893 2011-07-24 02:54:07 <fingster_> not seen...
 894 2011-07-24 02:54:52 eian has joined
 895 2011-07-24 02:55:13 <fingster_> when fake coins selected, the client is able to verify them
 896 2011-07-24 02:55:25 <fingster_> find them fake
 897 2011-07-24 02:55:46 cypher5001 has joined
 898 2011-07-24 02:55:52 <eian> Can someone explain the 'Format' column for Bitcoin's variable length integers? https://en.bitcoin.it/wiki/Protocol_specification#Variable_length_integer
 899 2011-07-24 02:55:58 <eian> I have no idea what that means
 900 2011-07-24 02:56:09 toffoo has joined
 901 2011-07-24 02:57:11 <jrmithdobbs> fingster_: i've answered your question, i can't say much more without code to look at.
 902 2011-07-24 02:57:27 <eian> what data types have storage lengths of 3, 5 and 9 bytes respectively?
 903 2011-07-24 02:57:35 <fingster_> thank you, ;)
 904 2011-07-24 02:57:42 gjs278 has quit (Ping timeout: 252 seconds)
 905 2011-07-24 02:58:12 <jrmithdobbs> eian: not understanding your question
 906 2011-07-24 02:58:37 <fingster_> uint8_t, unsigned int of 8 bit long
 907 2011-07-24 02:58:40 <eian> Under the format column, what does this mean: "0xfd + uint16_t"
 908 2011-07-24 02:58:53 <eian> uint16 to mean represents an 'unsigned short'
 909 2011-07-24 02:59:01 <fingster_> begin with 0xfd
 910 2011-07-24 02:59:13 <eian> OH
 911 2011-07-24 02:59:20 <jrmithdobbs> heh
 912 2011-07-24 02:59:26 <eian> holy hell
 913 2011-07-24 02:59:29 <eian> ok, got it
 914 2011-07-24 02:59:38 <eian> I was staring at that for at least 15 minutes
 915 2011-07-24 03:00:28 <eian> fingster, thanks
 916 2011-07-24 03:01:09 <fingster_> never mind
 917 2011-07-24 03:01:28 <eian> what? is that not the correct interpretation?
 918 2011-07-24 03:02:15 <fingster_> I'm not speaking English...
 919 2011-07-24 03:02:30 <eian> strange. I guess I'm not either
 920 2011-07-24 03:03:22 <fingster_> really?
 921 2011-07-24 03:04:19 toffoo has quit (Remote host closed the connection)
 922 2011-07-24 03:06:05 fnord0 has quit (Remote host closed the connection)
 923 2011-07-24 03:06:42 fnord0 has joined
 924 2011-07-24 03:07:57 nullrouten has joined
 925 2011-07-24 03:08:00 fnord0 has quit (Remote host closed the connection)
 926 2011-07-24 03:09:04 <lfm> + is concatenate there
 927 2011-07-24 03:10:55 <RenaKunisaki> bitcoind keeps giving me this error
 928 2011-07-24 03:10:58 <RenaKunisaki> error: {"code":-2,"message":"Safe mode: WARNING: Displayed transactions may not be correct!  You may need to upgrade, or other nodes may need to upgrade."}
 929 2011-07-24 03:11:13 <RenaKunisaki> I've upgraded, and everyone else has had plenty of time to upgrade, so what's with that
 930 2011-07-24 03:11:50 <lfm> are you sure that is a bitcoin message?
 931 2011-07-24 03:12:09 <RenaKunisaki> I don't know where else it could come from
 932 2011-07-24 03:13:16 sgstair has quit (Quit: .•«UPP»•.)
 933 2011-07-24 03:13:53 <lfm> RenaKunisaki: I think it means you are trying to run a bitcoind that doesnt match the server version. some old version in your path or something
 934 2011-07-24 03:14:10 <RenaKunisaki> yeah, it shows up with bitcoin getinfo. version 32300
 935 2011-07-24 03:14:18 d4de has quit (Ping timeout: 258 seconds)
 936 2011-07-24 03:14:32 <lfm> try 0.3.24 maybe?
 937 2011-07-24 03:15:27 <lfm> 3.23 is bad for other reasons
 938 2011-07-24 03:15:40 * RenaKunisaki downloads slowly
 939 2011-07-24 03:15:42 <fingster_> db damaged
 940 2011-07-24 03:15:56 <RenaKunisaki> hm, how to repair it?
 941 2011-07-24 03:16:17 <lfm> if it db damage, it depends which file it is
 942 2011-07-24 03:16:43 <fingster_> this happens when 'disconnect block' wrong
 943 2011-07-24 03:17:42 <fingster_> shutdown machine not properly
 944 2011-07-24 03:17:46 <lfm> RenaKunisaki: you say you upgraded but you still have 0.3.23?
 945 2011-07-24 03:18:11 <RenaKunisaki> well 0.3.22 was giving that error a while ago so I upgraded then
 946 2011-07-24 03:18:40 <RenaKunisaki> I've just been ignoring it for a while
 947 2011-07-24 03:18:58 ahihi2 has quit (Ping timeout: 252 seconds)
 948 2011-07-24 03:20:26 <lfm> you can wipe blk*.dat and addr.dat and keep wallet.dat. let it redownload the block chain and see if it fixes it. you should prolly get 0.3.24 tho for that anyway.
 949 2011-07-24 03:20:49 <RenaKunisaki> yeah 0.3.24 is downloading
 950 2011-07-24 03:21:05 <RenaKunisaki> it's a rather slow connection and already tied up with other downloads so it'll take a while :p
 951 2011-07-24 03:21:46 <phantomcircuit>  it's only ~ 600 MB of data
 952 2011-07-24 03:22:26 jimon has quit (Ping timeout: 240 seconds)
 953 2011-07-24 03:22:35 <RenaKunisaki> nah I'm still downloading the executable :p
 954 2011-07-24 03:22:51 <RenaKunisaki> another 10MB to go at 20k/s
 955 2011-07-24 03:23:38 freakazoid has joined
 956 2011-07-24 03:24:06 <phantomcircuit> oh
 957 2011-07-24 03:24:08 <phantomcircuit> lol
 958 2011-07-24 03:24:10 <lfm> RenaKunisaki: if you link is always that slow you might wanna get a preloaded blcok chain zip file to speed it up.
 959 2011-07-24 03:24:44 <RenaKunisaki> I can wait :p
 960 2011-07-24 03:25:20 ahihi2 has joined
 961 2011-07-24 03:25:44 <fingster_> someone should write a tool to fix db error
 962 2011-07-24 03:26:29 <lfm> fingster_: there is many ways to get db error, you cant always just fix em.
 963 2011-07-24 03:26:58 <fingster_> In fact, just need to erase last block
 964 2011-07-24 03:27:18 <fingster_> then re-download from the last block
 965 2011-07-24 03:27:29 <lfm> fingster_: then the index wont match!?
 966 2011-07-24 03:27:36 jimon has joined
 967 2011-07-24 03:28:06 <fingster_> read last block info from blkindex.dat
 968 2011-07-24 03:28:34 <fingster_> delete any transactions in the last block from blkindex.dat
 969 2011-07-24 03:29:36 <fingster_> Then client would download from the last block to the newest one, and setBestChain correctly
 970 2011-07-24 03:30:57 <lfm> fingster_: assuming first that the error is in the last block.
 971 2011-07-24 03:31:38 <lfm> and not in addr.dqat or wallet.dat
 972 2011-07-24 03:31:49 noagendamarket has joined
 973 2011-07-24 03:32:29 wolfspraul has joined
 974 2011-07-24 03:32:35 <fingster_> addr.dat and wallet.dat errors different, heh
 975 2011-07-24 03:33:39 <fingster_> the Error comes out at Reorganize > CBlock::Disconnect
 976 2011-07-24 03:37:18 <upb> wow you're from china, nice :)
 977 2011-07-24 03:38:17 arthurb has quit (Ping timeout: 258 seconds)
 978 2011-07-24 03:38:51 <fingster_> er, yeah, i'm from China
 979 2011-07-24 03:39:44 eastender has joined
 980 2011-07-24 03:39:48 Clipse has quit (Read error: Connection reset by peer)
 981 2011-07-24 03:40:33 <fingster_> I publish my client open source at : https://github.com/fingster/BTC-Express
 982 2011-07-24 03:40:45 Clipse has joined
 983 2011-07-24 03:43:06 thefinn93 has joined
 984 2011-07-24 03:43:19 <fingster_> howto add a new page to btc wiki?
 985 2011-07-24 03:45:19 thefinn93 has quit (Client Quit)
 986 2011-07-24 03:45:52 arthurb has joined
 987 2011-07-24 03:46:15 arthurb has quit (Remote host closed the connection)
 988 2011-07-24 03:49:54 hwolf has left ()
 989 2011-07-24 03:50:56 mmoya has quit (Ping timeout: 264 seconds)
 990 2011-07-24 03:52:49 TheZimm has quit (Read error: Connection reset by peer)
 991 2011-07-24 03:56:21 <RenaKunisaki> alright it looks like it's working after wiping and upgrading, although it hasn't downloaded any blocks yet
 992 2011-07-24 03:59:13 <fingster_> hug, :)
 993 2011-07-24 03:59:23 TheZimm has joined
 994 2011-07-24 04:00:25 spirals_away has quit (Ping timeout: 252 seconds)
 995 2011-07-24 04:02:24 Guest87553 is now known as abishai
 996 2011-07-24 04:10:27 Silverpike has joined
 997 2011-07-24 04:12:35 ar4s has joined
 998 2011-07-24 04:13:35 spirals_away has joined
 999 2011-07-24 04:14:14 Clipse has quit (Ping timeout: 255 seconds)
1000 2011-07-24 04:14:25 Folklore has joined
1001 2011-07-24 04:16:27 Clipse has joined
1002 2011-07-24 04:20:01 sgstair has joined
1003 2011-07-24 04:27:52 asuk has joined
1004 2011-07-24 04:37:25 aviadbd has joined
1005 2011-07-24 04:37:31 <aviadbd> anyone awake?
1006 2011-07-24 04:38:10 <aviadbd> here's another thing I don't get: the wiki says that miners in pooled mining get "simpler" tasks. But how can you simplify finding a double-hash that is smaller than a certain target number?
1007 2011-07-24 04:38:39 <aviadbd> what are the hashes an individual miner "finds"?
1008 2011-07-24 04:40:52 <Diablo-D3> aviadbd: pools just give a bigger target
1009 2011-07-24 04:43:14 <aviadbd> i figured as much; but how is that helpful?
1010 2011-07-24 04:43:37 <aviadbd> what do they do then? verify the smaller target? submit the hash to other workers?
1011 2011-07-24 04:44:06 <aviadbd> how does it help them reach the target, is what I don't get :)
1012 2011-07-24 04:44:44 gjs278 has joined
1013 2011-07-24 04:44:48 <Diablo-D3> aviadbd: it doesnt
1014 2011-07-24 04:45:03 <Diablo-D3> you do over a million individual, complete, attempts at producing a new block
1015 2011-07-24 04:45:24 <upb> it doesnt help them at all
1016 2011-07-24 04:45:33 <Diablo-D3> the only thing pools do is make miners return hashes that could produce new blocks if the diff was much lower
1017 2011-07-24 04:45:34 <upb> it only helps the pool to verify that the miner is doing work
1018 2011-07-24 04:45:51 <Diablo-D3> aviadbd: the non-block valid hashes, ie shares, are used as proof of work
1019 2011-07-24 04:46:02 <Diablo-D3> which the entire system is ALREADY a proof of work system, the problem was already solved
1020 2011-07-24 04:46:15 <aviadbd> ok... so i'm confused. I thought the idea of pooled mining is the workers helping the pool at reaching the target.
1021 2011-07-24 04:46:46 <aviadbd> that is to say: what incentive does the pool owner have to give shares to the miners, if they don't aid him in finding a block?
1022 2011-07-24 04:46:54 <upb> afaik the only point of pool mining is to smooth out the rate of income for miners
1023 2011-07-24 04:46:57 <upb> nothing more
1024 2011-07-24 04:47:39 <Diablo-D3> aviadbd: you have no clue how mining works.
1025 2011-07-24 04:47:42 <Diablo-D3> read what I wrote carefully.
1026 2011-07-24 04:47:49 <Diablo-D3> [12:41:04] <Diablo-D3> you do over a million individual, complete, attempts at producing a new block
1027 2011-07-24 04:47:55 <aviadbd> Diablo-D3: your assessment is correct. I have no clue. :)
1028 2011-07-24 04:48:08 <Diablo-D3> 1 mhash == 1 million attempts.
1029 2011-07-24 04:48:33 <aviadbd> ok. so far so good. with a lower target, or a real target?
1030 2011-07-24 04:48:41 <Diablo-D3> target is immaterial
1031 2011-07-24 04:48:48 <aviadbd> Okay.
1032 2011-07-24 04:48:48 <Diablo-D3> target has zero effect on speed
1033 2011-07-24 04:48:56 <aviadbd> that's true of course.
1034 2011-07-24 04:49:08 <Diablo-D3> the sha256 output is considered as a 256 bit integer, ie, a very large number
1035 2011-07-24 04:49:14 <aviadbd> so the pool owner gets all these lower target blocks, and searches for a correct target block from them?
1036 2011-07-24 04:49:21 <Diablo-D3> aviadbd: yeah
1037 2011-07-24 04:49:24 <aviadbd> and then pays off according to the shares?
1038 2011-07-24 04:49:25 TheZimm has quit (Quit: Computer has gone to sleep.)
1039 2011-07-24 04:49:25 <Diablo-D3> but thats after the fact
1040 2011-07-24 04:49:37 <aviadbd> yeah of course.
1041 2011-07-24 04:49:43 <Diablo-D3> the miners dont know what the network is actually at, nor even know what bitcoin really is
1042 2011-07-24 04:49:52 <aviadbd> and that gives him the incentive, and the workers the incentive.
1043 2011-07-24 04:49:53 <Diablo-D3> mining software is just given a 80 byte header, and sha256s it twice.
1044 2011-07-24 04:49:57 <Diablo-D3> thats _all_ it does.
1045 2011-07-24 04:50:17 <Diablo-D3> the target is a 256 bit number, the sha256 produced has to be smaller than that.
1046 2011-07-24 04:50:39 <Diablo-D3> so, the higher the difficulty, the smaller the number is
1047 2011-07-24 04:51:18 <Diablo-D3> pools just set the bar higher just to keep track of how much work you're actually doing
1048 2011-07-24 04:51:28 <Diablo-D3> er, bar lower
1049 2011-07-24 04:51:35 <aviadbd> Diablo-D3: that makes sense.
1050 2011-07-24 04:51:41 <Diablo-D3> some of those shares will be actual blocks
1051 2011-07-24 04:51:57 <aviadbd> Diablo-D3: i assumed that's what's happening; i just couldn't find it Written anywhere.
1052 2011-07-24 04:52:35 <aviadbd> maybe it should be made more clear on the wiki.
1053 2011-07-24 04:52:35 <TuxBlackEdo> hm
1054 2011-07-24 04:52:49 <aviadbd> (I might edit it if I have the permissions to ... I'll check later. :) )
1055 2011-07-24 04:52:56 <Diablo-D3> aviadbd: well, you cant explain how pools work without explaining how actual mining works
1056 2011-07-24 04:53:01 <TuxBlackEdo> sha256 produces a 256bit hash or a 256 byte hash?
1057 2011-07-24 04:53:07 gjs278 has quit (Read error: Connection reset by peer)
1058 2011-07-24 04:53:07 <Diablo-D3> TuxBlackEdo: its bits.
1059 2011-07-24 04:53:19 <Diablo-D3> its a very large number, but not THAT large.
1060 2011-07-24 04:53:27 <TuxBlackEdo> so a 32 character hash
1061 2011-07-24 04:53:37 <Diablo-D3> TuxBlackEdo: uh, no.
1062 2011-07-24 04:53:38 <aviadbd> Diablo-D3: actually, mining is explained well, if you look deep enough.
1063 2011-07-24 04:53:41 <Diablo-D3> wait yes
1064 2011-07-24 04:53:45 <Diablo-D3> if you mean by bytes.
1065 2011-07-24 04:53:46 <TuxBlackEdo> :)
1066 2011-07-24 04:53:55 <TuxBlackEdo> yeah bytes
1067 2011-07-24 04:54:03 <aviadbd> hehe. Java vs C?
1068 2011-07-24 04:54:16 <TuxBlackEdo> yep
1069 2011-07-24 04:55:28 MetaV has quit (Quit: Leaving)
1070 2011-07-24 04:55:41 MetaV has joined
1071 2011-07-24 04:55:53 gjs278 has joined
1072 2011-07-24 04:56:05 <aviadbd> anyway, thanks guys. you've been a great help. :)
1073 2011-07-24 04:56:55 <aviadbd> ooh, one more thing :
1074 2011-07-24 04:57:11 <aviadbd> it says in the wiki that the difficulty is set according to the amount of miners out there.
1075 2011-07-24 04:57:18 <Diablo-D3> not quite
1076 2011-07-24 04:57:24 <Diablo-D3> its set to the total hashing power out there
1077 2011-07-24 04:57:33 <aviadbd> and its supposed to be about 10 minutes / block
1078 2011-07-24 04:57:39 <aviadbd> yeah..
1079 2011-07-24 04:57:40 <Diablo-D3> it targets 10 minutes/block
1080 2011-07-24 04:57:49 <Diablo-D3> but theres a bunch of math on the actual difficulty set
1081 2011-07-24 04:57:51 <aviadbd> so how is that governed?
1082 2011-07-24 04:58:00 <aviadbd> and by who/what?
1083 2011-07-24 04:58:03 <Diablo-D3> involves stuff like time during the past few blocks, current block number, and others tuff
1084 2011-07-24 04:58:06 <Diablo-D3> *stuff
1085 2011-07-24 04:58:34 <aviadbd> i assume there's a bitcoin server somewhere that looks at how fast blocks are being made and and adjusts the next difficulty accordingly
1086 2011-07-24 04:58:38 <Diablo-D3> nope
1087 2011-07-24 04:58:41 <aviadbd> but that would be against the idea..
1088 2011-07-24 04:58:42 <Diablo-D3> each client knows the math
1089 2011-07-24 04:59:04 <Diablo-D3> so if a client gets an incoming new block that fails the math, it rejects it
1090 2011-07-24 04:59:22 <aviadbd> when you say client.. you mean wallet, miner, pool? all?
1091 2011-07-24 04:59:33 <aviadbd> probably not the miner - as we said, its dumb.
1092 2011-07-24 04:59:34 <aviadbd> right?
1093 2011-07-24 04:59:49 <Diablo-D3> the client, as in, bitcoin
1094 2011-07-24 05:00:07 <Diablo-D3> and technically pools too, but for obscure reasons, they all have an instance of bitcoin running behind them
1095 2011-07-24 05:00:22 <aviadbd> ok, that makes sense.
1096 2011-07-24 05:00:33 <aviadbd> and, i suppose they just want to be sure they're running the correct math?
1097 2011-07-24 05:00:45 <Diablo-D3> well, they have to make sure its not an attack on the network
1098 2011-07-24 05:01:19 <Diablo-D3> everything is mathematically provable
1099 2011-07-24 05:01:46 <aviadbd> the pools? but they generate blocks. what do they care about an attack? unless you mean an attack by creating a lot of blocks quickly, thus reducing their calculated target?
1100 2011-07-24 05:01:58 MobiusL has quit (Quit: Leaving)
1101 2011-07-24 05:02:14 <Diablo-D3> everyone has to worry about it
1102 2011-07-24 05:02:16 <Diablo-D3> pools arent magical
1103 2011-07-24 05:02:20 <Diablo-D3> they're normal clients too
1104 2011-07-24 05:02:31 <Diablo-D3> they just have a shitload of miners connected to them
1105 2011-07-24 05:02:42 <Diablo-D3> the pool software just proxies the bitcoin the pool owner is running
1106 2011-07-24 05:02:46 <aviadbd> hmm. so they make sure that transactions are legit using a bitcoin client before placing them in a block?
1107 2011-07-24 05:02:56 <Diablo-D3> not quite
1108 2011-07-24 05:03:04 <Diablo-D3> pool software does not manage transactions
1109 2011-07-24 05:03:05 <Diablo-D3> or anything
1110 2011-07-24 05:03:12 <aviadbd> bitcoin does
1111 2011-07-24 05:03:18 <Diablo-D3> yes, bitcoin does
1112 2011-07-24 05:03:24 <aviadbd> but pool software Could - it just uses bitcoin to do it for it.
1113 2011-07-24 05:03:38 <Diablo-D3> the only way to prove the entire currency pool is valid is to mathematically prove EVERYTHING
1114 2011-07-24 05:03:44 <Diablo-D3> every block, every transaction, everything
1115 2011-07-24 05:03:48 <aviadbd> yeah.
1116 2011-07-24 05:03:56 <aviadbd> crazy concept. :P
1117 2011-07-24 05:04:17 <Diablo-D3> and btw, if pool software does start doing that... it just means the pool software is now a client too
1118 2011-07-24 05:04:32 <aviadbd> yeah, that's what i meant. that pool software could be a client.
1119 2011-07-24 05:04:41 <aviadbd> or rather, it is - it just encapsulates bitcoin
1120 2011-07-24 05:04:42 <Diablo-D3> it could
1121 2011-07-24 05:04:48 <Diablo-D3> its just kinda pointless
1122 2011-07-24 05:04:59 <Diablo-D3> although Ive considered it with my pool software
1123 2011-07-24 05:05:05 <fingster_> a few days ago, I tried to explain btc to a professor. He ended up with full of questions.
1124 2011-07-24 05:05:13 <Diablo-D3> heh
1125 2011-07-24 05:05:17 <Diablo-D3> btc is easy to explain to noobs imo
1126 2011-07-24 05:05:20 MobiusL has joined
1127 2011-07-24 05:05:22 <aviadbd> depends on your target platform, i guess. not everyone is comfortable running bitcoin.
1128 2011-07-24 05:05:23 Folklore has quit (Ping timeout: 250 seconds)
1129 2011-07-24 05:05:32 <Diablo-D3> its a cryptographically self-proving transaction system
1130 2011-07-24 05:05:38 <TuxBlackEdo> so all pools are running the official client found on bitcoin.org
1131 2011-07-24 05:05:40 <Diablo-D3> the btc currency itself is a side effect.
1132 2011-07-24 05:05:43 Folklore has joined
1133 2011-07-24 05:05:49 <aviadbd> Diablo-D3: i think noobs fail at "crypto-self-what?"
1134 2011-07-24 05:05:50 <Diablo-D3> TuxBlackEdo: yes and/or with extra patches
1135 2011-07-24 05:05:58 <TuxBlackEdo> what patches?
1136 2011-07-24 05:06:04 <Diablo-D3> aviadbd: in the modern day of wikipedia, there is no reason for that.
1137 2011-07-24 05:06:07 <TuxBlackEdo> there are patches for bitcoind?
1138 2011-07-24 05:06:14 <Diablo-D3> TuxBlackEdo: its open source. duh.
1139 2011-07-24 05:06:32 <TuxBlackEdo> any cool mods? i do solo mining
1140 2011-07-24 05:06:38 <Diablo-D3> its useful for pools only
1141 2011-07-24 05:06:54 <TuxBlackEdo> i wanna see my total hashrate of all my connected miners (external)
1142 2011-07-24 05:06:58 <fingster_> his opinion is btc is useless and untrusted
1143 2011-07-24 05:06:58 <Diablo-D3> like, eligius uses a patch that allows luke to use gen txes to send btc out
1144 2011-07-24 05:07:12 <Diablo-D3> fingster_: btc itself is useless, but its NOT untrusted
1145 2011-07-24 05:07:19 <Diablo-D3> bitcoin itself is of the highest level of trust
1146 2011-07-24 05:07:19 <TuxBlackEdo> oh yeah, i like getting the blocks as "generated" blocks
1147 2011-07-24 05:07:24 <Diablo-D3> as in, its mathematically provable
1148 2011-07-24 05:07:37 <Diablo-D3> every single last bit of the currency pool cannot be faked
1149 2011-07-24 05:07:44 <fingster_> though he understood btc is safe proved by math
1150 2011-07-24 05:07:51 <Diablo-D3> the btc itself is useless
1151 2011-07-24 05:07:55 <Diablo-D3> they're just crypto tokens
1152 2011-07-24 05:08:11 <Diablo-D3> the same system could be used lots of places
1153 2011-07-24 05:08:14 <aviadbd> Diablo-D3: how is it useless? the numbers are, but its useful by the virtue of the meaning people put to it (i.e. currency)
1154 2011-07-24 05:08:26 <Diablo-D3> aviadbd: exactly, thats why its useless
1155 2011-07-24 05:08:37 <aviadbd> the same way a dollar bill is useless
1156 2011-07-24 05:08:39 <aviadbd> i guess
1157 2011-07-24 05:08:45 <Diablo-D3> even calling btc a currency us kinda disingenuous
1158 2011-07-24 05:08:48 <Diablo-D3> *is
1159 2011-07-24 05:08:52 <fingster_> btc' value is related to the power cost when produced them, right?
1160 2011-07-24 05:09:00 <Diablo-D3> fingster_: nope.
1161 2011-07-24 05:09:14 <Diablo-D3> btc value is what people are willing to pay for them
1162 2011-07-24 05:09:21 <Diablo-D3> THAT must exceed the actual cost of production
1163 2011-07-24 05:09:22 <nanotube> <Diablo-D3> even calling btc a currency us kinda disingenuous <- every currency is nothing more than a "token that other people agree has value"
1164 2011-07-24 05:09:29 <Diablo-D3> nanotube: kind of.
1165 2011-07-24 05:09:38 <Diablo-D3> btc has no direct equivilent
1166 2011-07-24 05:09:45 <fingster_> Then it is dangerous
1167 2011-07-24 05:09:48 <TuxBlackEdo> just like any other currency
1168 2011-07-24 05:10:08 <Diablo-D3> the software behind it, however, its incredibly interesting
1169 2011-07-24 05:10:12 <TuxBlackEdo> you could say the USD is useless with the same argument
1170 2011-07-24 05:10:21 <aviadbd> TuxBlackEdo: i just did but noone paid attention :P
1171 2011-07-24 05:10:29 <fingster_> heh
1172 2011-07-24 05:10:34 <Diablo-D3> it could be used to crypto-prove a lot of things that are logged or transactioned across untrusted or semitrusted nodes in parallel
1173 2011-07-24 05:10:52 <Diablo-D3> the fact that its used to make btc a currency is immaterial
1174 2011-07-24 05:11:46 <aviadbd> Diablo-D3: i agree. the software, and even the academic concept, is quite something here.
1175 2011-07-24 05:13:28 <fingster_> can btc resolve a problem task like seti@home, then produce btc?
1176 2011-07-24 05:13:57 <Diablo-D3> fingster_: no, that would be nonsensical
1177 2011-07-24 05:14:07 eastender has quit (Ping timeout: 258 seconds)
1178 2011-07-24 05:14:08 <Diablo-D3> such a system would require a problem that CAN be solved
1179 2011-07-24 05:14:10 <nanotube> fingster_: you're like the thousandth person to ask that, and the answer is no.
1180 2011-07-24 05:14:21 <Diablo-D3> it needs something that makes actual logical sense
1181 2011-07-24 05:14:23 <TuxBlackEdo> no because the bitcoin generating algorithm is generating sha256 hashes, and seti@home doesn't do any hashing
1182 2011-07-24 05:14:33 <Diablo-D3> seti@home has no defined outcome
1183 2011-07-24 05:14:37 <Diablo-D3> its not suitable
1184 2011-07-24 05:14:40 <TuxBlackEdo> no proof of work
1185 2011-07-24 05:14:51 <Diablo-D3> basically, if seti@home would be used... the system would generate btc when aliens are found.
1186 2011-07-24 05:15:22 <TuxBlackEdo> Diablo-D3: Finnish this sentence "Bitcoin is useless because..."
1187 2011-07-24 05:15:39 <fingster_> proof of work ought to be checking the computation amount
1188 2011-07-24 05:15:47 <Diablo-D3> "... I fucking said so, and I was here long before any of you."
1189 2011-07-24 05:15:57 <aviadbd> haha :)
1190 2011-07-24 05:16:03 <Diablo-D3> fingster_: it is. the more proof of work you have, the more you computated.
1191 2011-07-24 05:16:06 <TuxBlackEdo> yep pretty much
1192 2011-07-24 05:16:13 Optimo has quit (Ping timeout: 250 seconds)
1193 2011-07-24 05:16:23 <TuxBlackEdo> get off my lawn you damn kids
1194 2011-07-24 05:16:32 <TuxBlackEdo> :P
1195 2011-07-24 05:16:35 <aviadbd> :D
1196 2011-07-24 05:16:49 <Diablo-D3> seriously, I remember when this channel had like 12 people
1197 2011-07-24 05:17:04 <fingster_> zombies
1198 2011-07-24 05:17:04 <aviadbd> Diablo-D3: you sound as if you're sitting on a rocking chair with a laptop and a shotgun.
1199 2011-07-24 05:17:05 <aviadbd> :P
1200 2011-07-24 05:17:13 <TuxBlackEdo> all non-programmers must leave
1201 2011-07-24 05:17:22 <Diablo-D3> aviadbd: well, dude
1202 2011-07-24 05:17:23 <Diablo-D3> think about it
1203 2011-07-24 05:17:29 <Diablo-D3> I produced 650 btc with a cpu.
1204 2011-07-24 05:17:34 <aviadbd> hmm
1205 2011-07-24 05:17:41 <Diablo-D3> a single cpu.
1206 2011-07-24 05:17:46 <Diablo-D3> and it wasnt particularly lucky.
1207 2011-07-24 05:17:53 <Diablo-D3> and I didnt write a gpu miner yet.
1208 2011-07-24 05:17:55 <fingster_> no gpu involved?
1209 2011-07-24 05:18:08 <aviadbd> now that's more impressive than being here with 12 people :)
1210 2011-07-24 05:18:09 <fingster_> wow
1211 2011-07-24 05:18:18 <TuxBlackEdo> wow, you generated 650 btc with a single unlucky cpu?
1212 2011-07-24 05:18:23 <Diablo-D3> TuxBlackEdo: yup
1213 2011-07-24 05:18:25 <nanotube> fingster_: difficulty wasn't always in the millions, you know.
1214 2011-07-24 05:18:27 <Diablo-D3> and it wasnt even on 24/7.
1215 2011-07-24 05:18:29 <aviadbd> but alas, I must go - thanks for all the help, Diablo-D3 and the rest! :)
1216 2011-07-24 05:19:33 aviadbd has quit (Quit: back later)
1217 2011-07-24 05:19:51 <TuxBlackEdo> he waited for someone to say bye to him
1218 2011-07-24 05:19:55 <TuxBlackEdo> but no one did
1219 2011-07-24 05:20:30 <nanotube> heh
1220 2011-07-24 05:21:07 <fingster_> https://github.com/fingster/BTC-Express
1221 2011-07-24 05:21:11 <fingster_> my client for iOS
1222 2011-07-24 05:21:54 <TuxBlackEdo> nice, how does it work without downloading the blockchain?
1223 2011-07-24 05:22:08 <fingster_> in fact, can be compiled without gui for every platform
1224 2011-07-24 05:22:22 <fingster_> we set up a server
1225 2011-07-24 05:22:51 <nanotube> ah, so it basically relies on a central server, which acts as a full node on the bitcoin network?
1226 2011-07-24 05:23:05 <fingster_> server relay the trans client sent
1227 2011-07-24 05:23:19 <fingster_> exactly
1228 2011-07-24 05:24:11 <TuxBlackEdo> interesting... but is it secure?
1229 2011-07-24 05:24:28 Optimo has joined
1230 2011-07-24 05:24:40 <TuxBlackEdo> could a man in the middle attack happen?
1231 2011-07-24 05:25:01 <fingster_> impossible
1232 2011-07-24 05:25:20 <jrmithdobbs> why should anyone trust *your* server?
1233 2011-07-24 05:25:24 <fingster_> server can verify every trans like btc peers do
1234 2011-07-24 05:25:29 <TuxBlackEdo> i mean isnt that the point of having a bitcoin client with the full blockcain downloaded?
1235 2011-07-24 05:25:43 <fingster_> server cannot modify any trans
1236 2011-07-24 05:25:56 <jrmithdobbs> server could feed bad data to the client
1237 2011-07-24 05:25:58 <nanotube> the private keys are stored on the client only
1238 2011-07-24 05:26:32 <fingster_> right!
1239 2011-07-24 05:26:46 <fingster_> client keeps it wallet locally
1240 2011-07-24 05:27:01 <fingster_> client signs its sending trans
1241 2011-07-24 05:27:24 <nanotube> jrmithdobbs: well, one attack i can think of, server sends bad data as in, underreport amount of btc stored on $address, so that the client ends up sending a transaction with a bunch of extra btc lost to transaction fees.
1242 2011-07-24 05:27:32 <nanotube> anything else?
1243 2011-07-24 05:27:44 <nanotube> fingster_: ^ also see :)
1244 2011-07-24 05:28:40 <jrmithdobbs> nanotube: since the client doesn't have any blocks/hashes locally and didn't, on my cursory look, do any height checksum checking the possibilities are endless
1245 2011-07-24 05:29:06 <jrmithdobbs> s/height checksum/height checkpointing/
1246 2011-07-24 05:29:32 <nanotube> well, i guess client could easily be made to believe that it received BTC when it actually didn't
1247 2011-07-24 05:30:05 <nanotube> which would be a problem if you're selling something for btc
1248 2011-07-24 05:30:05 <fingster_> @nanotube: yep
1249 2011-07-24 05:30:09 <TuxBlackEdo> good point nanotube
1250 2011-07-24 05:30:25 <nanotube> and maybe jrmithdobbs can think of a few other sneaky things hehe
1251 2011-07-24 05:30:32 <jrmithdobbs> if it's only connecting to his server, and not a configurable node that defaults to "doesn't connect until you configure it" i wouldn't trust it
1252 2011-07-24 05:30:34 <nanotube> fingster_: hope the connection goes through ssl then. :)
1253 2011-07-24 05:30:39 <jrmithdobbs> it's basically willingly sybil'ing yourself
1254 2011-07-24 05:30:49 <fingster_> but bad data would be rejected by btc network
1255 2011-07-24 05:31:35 <nanotube> jrmithdobbs: well, presumably if you're running the sw, you at least trust the operator. mitm can then be avoided by using ssl and knowing the correct ssl key
1256 2011-07-24 05:31:55 LightRider has quit (afk!~LightRide@unaffiliated/lightrider|Ping timeout: 260 seconds)
1257 2011-07-24 05:32:04 <nanotube> if you don't trust the server operator, you shouldn't be running his software either.
1258 2011-07-24 05:32:20 LightRider has joined
1259 2011-07-24 05:32:20 <jrmithdobbs> nanotube: and if his server gets compromised?
1260 2011-07-24 05:32:40 <nanotube> well, it's no worse than if something like mtgox gets compromised. :)
1261 2011-07-24 05:32:57 <nanotube> well, it's actually better
1262 2011-07-24 05:33:30 <nanotube> because with mtgox-style wallet, if it gets rooted, your coins are gone
1263 2011-07-24 05:33:53 <TuxBlackEdo> yeah
1264 2011-07-24 05:33:54 <nanotube> fingster-style wallet, there are no coins to steal, the guy has to be patient and feed bad tx to people when they expect to receive a tx, etc.
1265 2011-07-24 05:34:05 <fingster_> agree
1266 2011-07-24 05:34:07 <TuxBlackEdo> use mtgox server to do transactions
1267 2011-07-24 05:34:18 <TuxBlackEdo> just write a iphone app for mtgox haha
1268 2011-07-24 05:34:26 <jrmithdobbs> i'm pretty sure there's some more evil things that could be done
1269 2011-07-24 05:34:28 <TuxBlackEdo> its pretty much like sending and receiving btc
1270 2011-07-24 05:34:35 <nanotube> TuxBlackEdo: that's called a web browser :)
1271 2011-07-24 05:34:40 <TuxBlackEdo> oh wait
1272 2011-07-24 05:34:54 <TuxBlackEdo> i got a really old phone :(
1273 2011-07-24 05:34:57 <jrmithdobbs> nanotube: since the client looks up every piece of data it needs from this "trusted" server on demand every time
1274 2011-07-24 05:34:58 spirals_away is now known as spirals
1275 2011-07-24 05:35:05 spirals has quit (Changing host)
1276 2011-07-24 05:35:06 spirals has joined
1277 2011-07-24 05:35:21 <nanotube> jrmithdobbs: well ideally, maybe the client would store a cache of the previous transactions it has received.
1278 2011-07-24 05:35:32 <nanotube> that would make it not completely memoryless.
1279 2011-07-24 05:35:35 <TuxBlackEdo> maybe a text message transfer would be cool
1280 2011-07-24 05:35:38 <jrmithdobbs> which could have been spoofed by the server in the first place
1281 2011-07-24 05:35:39 <TuxBlackEdo> ok guys check this out
1282 2011-07-24 05:36:05 <jrmithdobbs> it decentralizes unnecessarily
1283 2011-07-24 05:36:10 <jrmithdobbs> in an untrustworthy way
1284 2011-07-24 05:36:17 <fingster_> @nanotube: wallet stores previous trans
1285 2011-07-24 05:36:18 <nanotube> you mean, centralizes
1286 2011-07-24 05:36:22 E-sense has quit (Ping timeout: 258 seconds)
1287 2011-07-24 05:36:22 <jrmithdobbs> err ya
1288 2011-07-24 05:36:34 <jrmithdobbs> and it does it to a client that can't directly validate it's data
1289 2011-07-24 05:36:34 <nanotube> fingster_: cool
1290 2011-07-24 05:36:41 <TuxBlackEdo> you make a bitcoin wallet site, and people can text peoples email addresses instead of bitcoin addresses (like paypal)
1291 2011-07-24 05:37:01 <TuxBlackEdo> so i could text "send 5 bill@gates.com"
1292 2011-07-24 05:37:16 <nanotube> jrmithdobbs: well, it's an interesting compromise between a fully-remote wallet, and a local full node.
1293 2011-07-24 05:37:18 <TuxBlackEdo> only problem is when someone spoofs your phone number and sends bitcoins to himself
1294 2011-07-24 05:37:42 <jrmithdobbs> nanotube: i'd be fine with it if it required you run your own bitcoin instance
1295 2011-07-24 05:37:54 <nanotube> jrmithdobbs: then you're looking for bitcoin-js-remote :)
1296 2011-07-24 05:38:03 <nanotube> which is great, imo :)
1297 2011-07-24 05:38:04 <jrmithdobbs> i'm not looking for anything
1298 2011-07-24 05:38:12 <nanotube> well, i mean "you are thinking of"
1299 2011-07-24 05:38:13 <nanotube> :)
1300 2011-07-24 05:38:34 <nanotube> but not everyone is savvy enough to set up a vps with a bitcoind on it.
1301 2011-07-24 05:39:01 <jrmithdobbs> this puts every user of this platform at the server ops' mercy though
1302 2011-07-24 05:39:11 <nanotube> i for one would go with bitcoin-js-remote... but joe shmoe probably can't.
1303 2011-07-24 05:39:13 <TuxBlackEdo> but thats a cool idea, no?? make it like paypal, when you text the service "send 5btc to user@gmail.com" it will look up user@gmail.com's phone and text them that they received the bitcoins, and if user@gmail.com doesnt have an account he will just get an email
1304 2011-07-24 05:39:23 <nanotube> yes it does. every "bitcoin bank"-like system does.
1305 2011-07-24 05:39:24 <TuxBlackEdo> I think we could really stick it to paypal if we did something like this
1306 2011-07-24 05:39:50 <jrmithdobbs> who cares about sticking it to paypal
1307 2011-07-24 05:39:54 <nanotube> TuxBlackEdo: bitcoinmail.com
1308 2011-07-24 05:40:08 <TuxBlackEdo> yes ok
1309 2011-07-24 05:40:14 <TuxBlackEdo> that site is 50% of the way there
1310 2011-07-24 05:40:18 <TuxBlackEdo> it needs to implement SMS
1311 2011-07-24 05:40:38 <nanotube> http://yro.slashdot.org/story/11/07/22/2345217/PayPal-Joins-London-Police-Effort <- paypal is sticking it to itself. practically begging people to use some other funds transfer service. :)
1312 2011-07-24 05:41:07 <jrmithdobbs> TuxBlackEdo: why, only poor people don't have email on their phone
1313 2011-07-24 05:41:12 <jrmithdobbs> and poor people don't have bitcoins
1314 2011-07-24 05:41:13 <jrmithdobbs> the end
1315 2011-07-24 05:41:43 <TuxBlackEdo> so i can text bitcoinmail.com's server to send bitcoins to user@domain.com and if that person has an account it will text their phone that they received the bitcoins
1316 2011-07-24 05:41:54 <nanotube> jrmithdobbs: haha, nice try, but unfortunately i am a counterexample :P
1317 2011-07-24 05:42:09 <nanotube> jrmithdobbs: that said, i don't have sms on my phone either...
1318 2011-07-24 05:42:52 <TuxBlackEdo> i just want to be able to call someone up while i am driving, and say "hey ill send you some bitcoins right now" and I'd be able to text bitcoinmail.com and then they would text the person who i am sending the bitcoins to, so both parties dont have to be at a computer
1319 2011-07-24 05:43:34 <TuxBlackEdo> and the other party would know they received bitcoins because the service (let's say bitcoinmail.com) would tell them in a SMS
1320 2011-07-24 05:44:11 <fingster_> good idea
1321 2011-07-24 05:44:22 <TuxBlackEdo> so nobody has to be by a computer
1322 2011-07-24 05:44:31 <TuxBlackEdo> and it would be simple as hell
1323 2011-07-24 05:44:43 jargon has joined
1324 2011-07-24 05:44:43 jargon has quit (Changing host)
1325 2011-07-24 05:44:43 jargon has joined
1326 2011-07-24 05:45:19 <fingster_> but you should trust on bitcoinmail.com
1327 2011-07-24 05:46:03 <jrmithdobbs> nanotube: baristas aren't people so fall outside of the set ;p
1328 2011-07-24 05:46:03 <fingster_> maybe need to allow bitcoinmail.com keep your wallet
1329 2011-07-24 05:46:04 * jrmithdobbs runs
1330 2011-07-24 05:46:59 <TuxBlackEdo> fingster_ yes because it would be impossible for bitcoinmail.com to take money out of your bitcoin wallet when you send them a text to send someone else bitcoins
1331 2011-07-24 05:47:51 <TuxBlackEdo> but the receiving party could be able to make automatic/instant withdrawls to a bitcoin address
1332 2011-07-24 05:48:31 <TuxBlackEdo> anyhow i dont care, all I want is bitcoin transfers to be done over SMS with the receiving party getting an SMS notification. Everything else doesn't matter
1333 2011-07-24 05:48:32 <nanotube> jrmithdobbs: haha
1334 2011-07-24 05:49:34 <TuxBlackEdo> http://en.wikipedia.org/wiki/File:SMS_messages_sent_monthly_in_USA_%28in_billions%29.svg
1335 2011-07-24 05:52:09 LightRider has quit (Ping timeout: 260 seconds)
1336 2011-07-24 05:52:16 <TuxBlackEdo> twitter is an SMS based social network
1337 2011-07-24 05:52:22 <TuxBlackEdo> people use SMS
1338 2011-07-24 05:52:36 LightRider has joined
1339 2011-07-24 05:55:54 sgornick has quit (Quit: Leaving.)
1340 2011-07-24 05:58:17 <eian> Is this current: https://en.bitcoin.it/wiki/Protocol_specification#version
1341 2011-07-24 05:58:56 <eian> I'm looking at a wireshark capture and it looks like addr_you comes before addr_me
1342 2011-07-24 06:00:49 sgornick has joined
1343 2011-07-24 06:03:24 cuddlefish has joined
1344 2011-07-24 06:03:45 <cuddlefish> oh sweet jesus
1345 2011-07-24 06:03:46 <cuddlefish> http://textcoin.co.uk/
1346 2011-07-24 06:04:43 <TuxBlackEdo> yeah i got your message in #bitcoin
1347 2011-07-24 06:04:54 <nanotube> cuddlefish: pretty cool. though i wonder how chargebackable txt message charges are.
1348 2011-07-24 06:05:07 <cuddlefish> nanotube: you'd probably have to fight pretty hard
1349 2011-07-24 06:05:38 <nanotube> mmm
1350 2011-07-24 06:05:57 <jrmithdobbs> nanotube: very
1351 2011-07-24 06:05:59 <upb> they arent
1352 2011-07-24 06:06:12 <jrmithdobbs> upb: you can dispute them with carrier in some cases
1353 2011-07-24 06:06:27 <upb> sure but they'll just track it was really sent from your phone
1354 2011-07-24 06:06:35 <nanotube> hmm well, whatever the case is... i can just say "if chargebackable: service will die soon"
1355 2011-07-24 06:06:44 <nanotube> heh
1356 2011-07-24 06:06:50 <cuddlefish> until then...
1357 2011-07-24 06:06:56 <cuddlefish> grab friend's phone
1358 2011-07-24 06:06:59 <cuddlefish> buy 10 BTC
1359 2011-07-24 06:07:01 LightRider has quit (Read error: Connection reset by peer)
1360 2011-07-24 06:07:02 <jrmithdobbs> haha
1361 2011-07-24 06:07:11 noagendamarket has quit (Ping timeout: 258 seconds)
1362 2011-07-24 06:07:30 <cuddlefish> oh, I only get up to 5 euros
1363 2011-07-24 06:07:31 <cuddlefish> damn
1364 2011-07-24 06:07:53 <jrmithdobbs> per text
1365 2011-07-24 06:08:01 <cuddlefish> yep :3
1366 2011-07-24 06:08:08 <cuddlefish> eh, pounds
1367 2011-07-24 06:08:16 <cuddlefish> silly font
1368 2011-07-24 06:08:21 <jrmithdobbs> even better
1369 2011-07-24 06:09:09 <cuddlefish> but seriously, this could change everything
1370 2011-07-24 06:09:14 <cuddlefish> bitcoin android app could integreate
1371 2011-07-24 06:09:27 <cuddlefish> 'tap here to purchase Bitcoins!'
1372 2011-07-24 06:09:36 <cuddlefish> nanotube: the fact is, it will survive
1373 2011-07-24 06:09:42 <cuddlefish> because they first, get a god markup
1374 2011-07-24 06:09:44 <TuxBlackEdo> hey
1375 2011-07-24 06:09:48 <cuddlefish> *good
1376 2011-07-24 06:09:49 <TuxBlackEdo> gmaxwell hangs out here?
1377 2011-07-24 06:09:52 <jrmithdobbs> pretty sure they'll get payment services cut off eventually
1378 2011-07-24 06:09:57 <cuddlefish> and second, it's NON-TRIVIAL to chargeback
1379 2011-07-24 06:10:03 <upb> bitcoin over sms will never work because the telcos get a too big cut
1380 2011-07-24 06:10:20 <cuddlefish> upb: sure. but for a quick-and-easy solution
1381 2011-07-24 06:10:20 <upb> and if they use an intermediary they too take a big cut
1382 2011-07-24 06:10:26 <cuddlefish> perhaps to seed your initial wallet
1383 2011-07-24 06:11:00 <nanotube> TuxBlackEdo: yes he does.
1384 2011-07-24 06:11:06 <TuxBlackEdo> oh
1385 2011-07-24 06:11:07 <cuddlefish> 'Welcome to Bitcoin Wallet for Android! You currently have 0 BTC. To purchase your first coins (charged to your phone bill), tap here."
1386 2011-07-24 06:11:13 <TuxBlackEdo> he did something really cool nanotube
1387 2011-07-24 06:11:18 <nanotube> TuxBlackEdo: i know :)
1388 2011-07-24 06:11:18 <cuddlefish> it doesn't matter if it's super-expensive at first
1389 2011-07-24 06:11:27 <jrmithdobbs> TuxBlackEdo: we know
1390 2011-07-24 06:11:32 <nanotube> cuddlefish: you have a point. with a large enough margin, they can affort to eat some chargebacks.
1391 2011-07-24 06:11:32 <TuxBlackEdo> oh ok never mind then
1392 2011-07-24 06:11:51 <cuddlefish> nanotube: and unlike paypal, the intermediaries actually care about the sellers
1393 2011-07-24 06:12:10 <cuddlefish>  because it's at LEAST $300/month just for the number
1394 2011-07-24 06:12:23 <cuddlefish> and then, a percentage of each charge...
1395 2011-07-24 06:12:24 <nanotube> well, prepaid phones...
1396 2011-07-24 06:12:39 <cuddlefish> nanotube: don't know how prepaid SMS works there
1397 2011-07-24 06:12:45 <cuddlefish> believe it fails if balance is low
1398 2011-07-24 06:13:05 <nanotube> probably does. but i mean a charge/back/ is about getting your money /back/ :)
1399 2011-07-24 06:13:21 <cuddlefish> right...
1400 2011-07-24 06:13:23 <nanotube> but anyway, if it's not entirely trivial, and if they make it small amounts per person.
1401 2011-07-24 06:13:33 <nanotube> it can work if they have a hefty profit margin
1402 2011-07-24 06:13:35 <cuddlefish> nanotube: I'm fairly sure you basically have to sue :D
1403 2011-07-24 06:13:49 <nanotube> heh
1404 2011-07-24 06:15:01 <nanotube> cuddlefish: also: new way to cash in on a stolen cellphone? :)
1405 2011-07-24 06:15:16 <jrmithdobbs> ya it's too abusable
1406 2011-07-24 06:15:22 <jrmithdobbs> they'll get shutdown
1407 2011-07-24 06:15:30 <cuddlefish> not really
1408 2011-07-24 06:15:40 <cuddlefish> lots of services take SMS payments
1409 2011-07-24 06:15:43 <nanotube> well let's hope not. stealing cellphones requires physical theft, which is a whole different animal from digital theft.
1410 2011-07-24 06:15:59 <cuddlefish> boku.com
1411 2011-07-24 06:21:06 <cuddlefish> If someone brings textcoin to the US
1412 2011-07-24 06:21:25 <cuddlefish> I will send them a pizza
1413 2011-07-24 06:21:53 <cuddlefish> and, if they get a newspaper article about it in a reputable source, I'll ship them a kitten
1414 2011-07-24 06:22:04 <TuxBlackEdo> dead or alive?
1415 2011-07-24 06:22:13 <cuddlefish> either
1416 2011-07-24 06:25:00 rethaw has joined
1417 2011-07-24 06:28:00 <jrmithdobbs> what is digital theft?
1418 2011-07-24 06:28:03 <jrmithdobbs> theft is theft
1419 2011-07-24 06:29:29 <Zeiris> Schrodinger's Kitten
1420 2011-07-24 06:31:11 zyb has quit (Ping timeout: 258 seconds)
1421 2011-07-24 06:31:57 zyb has joined
1422 2011-07-24 06:32:35 MobiusL has quit (Read error: Connection reset by peer)
1423 2011-07-24 06:35:50 eian has quit (Quit: Leaving)
1424 2011-07-24 06:36:28 MobiusL has joined
1425 2011-07-24 06:40:14 bitcoinbulletin has quit (Remote host closed the connection)
1426 2011-07-24 06:44:19 bitcoinbulletin has joined
1427 2011-07-24 06:45:24 <kreal-> Help me translate my website: http://forum.bitcoin.org/index.php?topic=31307.0
1428 2011-07-24 06:46:00 Zarutian has quit (Quit: Zarutian)
1429 2011-07-24 06:47:53 briareus has joined
1430 2011-07-24 06:47:53 briareus has quit (Changing host)
1431 2011-07-24 06:47:53 briareus has joined
1432 2011-07-24 06:54:54 Stellar has joined
1433 2011-07-24 06:55:36 zeropointo has joined
1434 2011-07-24 06:58:25 <zeropointo> anyone know where i can find an alternative to json-rpc.org to d/l python-jsonrpc?
1435 2011-07-24 07:01:41 evelyn66 has joined
1436 2011-07-24 07:03:27 <rethaw> yes
1437 2011-07-24 07:03:49 <rethaw> https://github.com/bmjames/python-jsonrpc
1438 2011-07-24 07:04:53 <zeropointo> rethaw: thanks i'll check it out
1439 2011-07-24 07:05:16 <rethaw> literally?
1440 2011-07-24 07:05:58 <zeropointo> lol no sorry didn't realize that could be misinterpreted
1441 2011-07-24 07:07:41 karnac has joined
1442 2011-07-24 07:08:32 <zeropointo> rethaw: all the updates are 2007, was the origional that old?
1443 2011-07-24 07:08:45 <rethaw> it works
1444 2011-07-24 07:09:17 <zeropointo> in that case i guess it's better than nothing
1445 2011-07-24 07:12:02 senseles has quit (Read error: Connection reset by peer)
1446 2011-07-24 07:12:14 senseles has joined
1447 2011-07-24 07:13:45 E-sense has joined
1448 2011-07-24 07:15:02 lolwat` has quit (Remote host closed the connection)
1449 2011-07-24 07:15:26 freakazoid has quit (Quit: Computer has gone to sleep.)
1450 2011-07-24 07:20:22 wardearia has quit (Read error: Operation timed out)
1451 2011-07-24 07:23:21 FractalUniverse has joined
1452 2011-07-24 07:23:38 Folklore has quit (Ping timeout: 250 seconds)
1453 2011-07-24 07:24:21 Folklore has joined
1454 2011-07-24 07:24:32 nick has joined
1455 2011-07-24 07:24:58 nick is now known as Guest52604
1456 2011-07-24 07:25:02 imsaguy has quit (Ping timeout: 255 seconds)
1457 2011-07-24 07:25:51 karnac has quit (Quit: karnac)
1458 2011-07-24 07:29:43 Qatz has joined
1459 2011-07-24 07:29:43 DaQatz has quit (Read error: Connection reset by peer)
1460 2011-07-24 07:29:45 fingster_ has quit (Ping timeout: 252 seconds)
1461 2011-07-24 07:32:03 gjs278 has quit (Remote host closed the connection)
1462 2011-07-24 07:33:21 asuk has quit (Ping timeout: 240 seconds)
1463 2011-07-24 07:33:27 Guest52604 has quit (Ping timeout: 252 seconds)
1464 2011-07-24 07:33:49 imsaguy has joined
1465 2011-07-24 07:33:49 imsaguy has quit (Changing host)
1466 2011-07-24 07:33:49 imsaguy has joined
1467 2011-07-24 07:36:29 jargon has quit ()
1468 2011-07-24 07:38:02 wardearia has joined
1469 2011-07-24 07:39:37 Rabbit67890 has joined
1470 2011-07-24 07:45:32 <zeropointo> exit
1471 2011-07-24 07:45:34 zeropointo has quit (Quit: leaving)
1472 2011-07-24 07:49:26 Rabbit67890 has quit (Quit: Colloquy for iPad - http://colloquy.mobi)
1473 2011-07-24 07:50:06 molecular has joined
1474 2011-07-24 07:53:48 <abishai> hmm any suggestions on a good free online collaboration tool to use for a new project?
1475 2011-07-24 07:54:22 <abishai> if you used that you liked, spread the word
1476 2011-07-24 07:55:12 <abishai> if you used one that you liked, spread the word, even
1477 2011-07-24 07:56:23 <abishai> looking for something different than Basecamp and Collabtive we've been using till now
1478 2011-07-24 07:59:34 jivvz has quit (Quit: Lämnar)
1479 2011-07-24 08:05:22 devon_hillard has joined
1480 2011-07-24 08:11:02 koleg has joined
1481 2011-07-24 08:11:47 ujjain has quit (Ping timeout: 240 seconds)
1482 2011-07-24 08:12:45 Quetzalcoatl_ has quit (Ping timeout: 260 seconds)
1483 2011-07-24 08:16:25 Quetzalcoatl_ has joined
1484 2011-07-24 08:23:28 gjs278 has joined
1485 2011-07-24 08:24:27 Clipse has quit (Ping timeout: 255 seconds)
1486 2011-07-24 08:26:38 MrTiggr has quit (Ping timeout: 252 seconds)
1487 2011-07-24 08:28:26 MrTiggr has joined
1488 2011-07-24 08:30:04 kish`_ has joined
1489 2011-07-24 08:32:51 kish` has quit (Ping timeout: 252 seconds)
1490 2011-07-24 08:34:02 AStove has joined
1491 2011-07-24 08:43:40 blueadept has quit (Max SendQ exceeded)
1492 2011-07-24 08:46:03 molecular has quit (Ping timeout: 255 seconds)
1493 2011-07-24 08:50:43 moa7 has joined
1494 2011-07-24 08:55:16 TheAncientGoat has joined
1495 2011-07-24 08:57:52 imsaguy has quit (Read error: Connection reset by peer)
1496 2011-07-24 08:58:39 molecular has joined
1497 2011-07-24 08:59:19 denisx has joined
1498 2011-07-24 09:01:48 Beccara__ has quit (Ping timeout: 255 seconds)
1499 2011-07-24 09:01:51 Guest64059 has joined
1500 2011-07-24 09:07:14 abishai has quit (Quit: Off)
1501 2011-07-24 09:07:21 spirals has quit (Ping timeout: 258 seconds)
1502 2011-07-24 09:09:16 Guest64059 has quit (Ping timeout: 258 seconds)
1503 2011-07-24 09:17:11 ejbvanc has quit (Ping timeout: 240 seconds)
1504 2011-07-24 09:19:09 nhodges_k has joined
1505 2011-07-24 09:19:43 dvide has joined
1506 2011-07-24 09:19:45 ejbvanc has joined
1507 2011-07-24 09:21:56 nhodges has quit (Ping timeout: 250 seconds)
1508 2011-07-24 09:22:47 spirals has joined
1509 2011-07-24 09:24:17 ujjain has joined
1510 2011-07-24 09:28:00 nhodges has joined
1511 2011-07-24 09:28:40 nick has joined
1512 2011-07-24 09:29:11 nick is now known as Guest95733
1513 2011-07-24 09:29:15 somuchwin has quit (Ping timeout: 255 seconds)
1514 2011-07-24 09:30:03 somuchwin has joined
1515 2011-07-24 09:30:11 nhodges_k has quit (Ping timeout: 250 seconds)
1516 2011-07-24 09:30:22 Incitatus has joined
1517 2011-07-24 09:30:54 kytibe has quit (Remote host closed the connection)
1518 2011-07-24 09:34:10 nhodges_g has joined
1519 2011-07-24 09:36:40 nhodges has quit (Ping timeout: 250 seconds)
1520 2011-07-24 09:38:01 Incitatus has quit (Ping timeout: 258 seconds)
1521 2011-07-24 09:40:20 maikmerten has joined
1522 2011-07-24 09:45:40 mosimo has joined
1523 2011-07-24 09:46:47 nhodges has joined
1524 2011-07-24 09:49:40 nhodges_g has quit (Ping timeout: 250 seconds)
1525 2011-07-24 09:51:18 molecular has quit (Ping timeout: 255 seconds)
1526 2011-07-24 09:52:10 molecular has joined
1527 2011-07-24 09:52:20 Stryker has quit (Read error: Connection reset by peer)
1528 2011-07-24 09:54:19 Stryker has joined
1529 2011-07-24 09:55:09 suriv has joined
1530 2011-07-24 09:55:46 <da2ce7> how do I check if a transaction looks valid to other clients?
1531 2011-07-24 09:56:02 <da2ce7> other than waiting for a confimation?
1532 2011-07-24 09:59:10 <cuddlefish> da2ce7: check bitcoincharts.com/bitcoin
1533 2011-07-24 09:59:29 <da2ce7> cuddlefish, I have... I have 6 connections... and it isn't showing up
1534 2011-07-24 09:59:45 <cuddlefish> da2ce7: hmm
1535 2011-07-24 09:59:50 <cuddlefish> TXid?
1536 2011-07-24 09:59:57 <cuddlefish> i'll check debug.log
1537 2011-07-24 10:00:32 <da2ce7> to: 18QWm2sXiYX8p2kKSKTtr8wMFtPRtcgNWb
1538 2011-07-24 10:00:50 <cuddlefish> right
1539 2011-07-24 10:00:53 <cuddlefish> but I need the TX id
1540 2011-07-24 10:01:09 <da2ce7> *check logs.
1541 2011-07-24 10:01:21 <da2ce7> is there a easy way to get it?
1542 2011-07-24 10:01:29 <denisx> this address is not known yet
1543 2011-07-24 10:01:48 <denisx> bitcoind listtransactions '' 50
1544 2011-07-24 10:01:56 <da2ce7> k
1545 2011-07-24 10:02:36 <cuddlefish> hmm
1546 2011-07-24 10:03:16 Joric has joined
1547 2011-07-24 10:06:36 molecular has quit (Ping timeout: 255 seconds)
1548 2011-07-24 10:06:37 ktbe has joined
1549 2011-07-24 10:06:56 <da2ce7> has heart attack
1550 2011-07-24 10:07:42 <da2ce7> I will kill myself if I have lost my coins.
1551 2011-07-24 10:07:42 <cuddlefish> da2ce7: ?
1552 2011-07-24 10:08:11 <da2ce7> checks address in explorer
1553 2011-07-24 10:10:03 <da2ce7> I will kill myself.
1554 2011-07-24 10:10:03 Folklore has quit (Remote host closed the connection)
1555 2011-07-24 10:10:06 <da2ce7> or maybe not
1556 2011-07-24 10:10:11 Qatz has quit (Ping timeout: 260 seconds)
1557 2011-07-24 10:10:13 <da2ce7> maybe I'll start from scratch
1558 2011-07-24 10:11:02 <da2ce7> unexpected spends don't show up?
1559 2011-07-24 10:11:06 ujjain has quit (Ping timeout: 255 seconds)
1560 2011-07-24 10:11:12 ktbe has quit (Remote host closed the connection)
1561 2011-07-24 10:13:15 <da2ce7> I will really kill myself... or maybe start from scratch
1562 2011-07-24 10:13:28 <da2ce7> doing a rescan now
1563 2011-07-24 10:13:43 <da2ce7> oh thank god!
1564 2011-07-24 10:13:44 <da2ce7> :)
1565 2011-07-24 10:13:56 <da2ce7> I have at least some coins...
1566 2011-07-24 10:14:32 DaQatz has joined
1567 2011-07-24 10:14:49 * da2ce7 has his bitcoin locked up like ft knox
1568 2011-07-24 10:14:57 <da2ce7> it would be supprising if somebody got in
1569 2011-07-24 10:15:03 <da2ce7> but you never know
1570 2011-07-24 10:16:03 <da2ce7> ok the fist transaction is still not showing up
1571 2011-07-24 10:16:23 <da2ce7> loading up bitcoind is going to be annoying.
1572 2011-07-24 10:16:53 <denisx> which version do you use?
1573 2011-07-24 10:17:04 <da2ce7> .24 beta
1574 2011-07-24 10:17:17 <denisx> on linux?
1575 2011-07-24 10:17:20 <da2ce7> winodws
1576 2011-07-24 10:17:26 <da2ce7> windows.
1577 2011-07-24 10:17:32 <denisx> lolz, ft knox ;)
1578 2011-07-24 10:17:42 <da2ce7> lol... I know how to secure windows.
1579 2011-07-24 10:17:55 <da2ce7> windows can be pritty secure if you do it rite.
1580 2011-07-24 10:18:33 <da2ce7> well... no unexpect transactions have shown up yet...
1581 2011-07-24 10:19:02 <senseles> have you downloaded all the block?
1582 2011-07-24 10:19:02 <senseles> payment wont show until your client is up to date
1583 2011-07-24 10:19:14 <da2ce7> ummm....
1584 2011-07-24 10:19:18 <senseles> it should read 137783 blocks
1585 2011-07-24 10:19:23 <da2ce7> yep got em'
1586 2011-07-24 10:19:28 <denisx> da2ce7: yeah, what does getinfo say
1587 2011-07-24 10:19:29 Zagitta has joined
1588 2011-07-24 10:19:31 <denisx> ok
1589 2011-07-24 10:25:07 rethaw has quit (Quit: rethaw)
1590 2011-07-24 10:26:42 larsivi has joined
1591 2011-07-24 10:27:19 ujjain has joined
1592 2011-07-24 10:30:07 <da2ce7> ok
1593 2011-07-24 10:30:23 <da2ce7> "txid" : "0993c4c0aed11c7b9d31c616bf24f175c0a1374da64c35f7061f25bcba0f2be6",
1594 2011-07-24 10:30:30 <da2ce7> *prays*
1595 2011-07-24 10:30:51 <da2ce7> cuddlefish ^
1596 2011-07-24 10:31:00 <da2ce7> I cannot think clearly anymore.
1597 2011-07-24 10:37:08 <da2ce7> anyone?
1598 2011-07-24 10:39:08 sipa has joined
1599 2011-07-24 10:39:11 <denisx> its not there
1600 2011-07-24 10:39:53 <da2ce7> werid... it may not be relayed
1601 2011-07-24 10:40:59 <da2ce7> what is your clients IP address... I can I please try connecting to it? do you accept transactions with bad/low fees?
1602 2011-07-24 10:41:23 Daviey has quit (Excess Flood)
1603 2011-07-24 10:41:24 fsntation has joined
1604 2011-07-24 10:41:50 <da2ce7> if the transaction is having spent coins... it wont' be transmitted :(
1605 2011-07-24 10:42:05 <da2ce7> it seems like not all my coins have been spent.
1606 2011-07-24 10:43:10 Guest95733 has quit (Ping timeout: 264 seconds)
1607 2011-07-24 10:43:54 Daviey has joined
1608 2011-07-24 10:46:39 BlueMatt has joined
1609 2011-07-24 10:46:46 RazielZ has joined
1610 2011-07-24 10:46:54 imsaguy has joined
1611 2011-07-24 10:47:00 <da2ce7> moved some coins to a outside wallet...
1612 2011-07-24 10:47:03 <da2ce7> *offline
1613 2011-07-24 10:47:11 <da2ce7> and will remake wallet.
1614 2011-07-24 10:50:45 <da2ce7> hmmm.... my first bitcoin shock.
1615 2011-07-24 10:50:52 <da2ce7> I think that I have most of the coins.
1616 2011-07-24 10:50:56 <da2ce7> if not all...
1617 2011-07-24 10:51:07 <sipa> ...
1618 2011-07-24 10:51:17 <da2ce7> still tring to work out why the 0993c4c0aed11c7b9d31c616bf24f175c0a1374da64c35f7061f25bcba0f2be6 insn't getting transmitted.
1619 2011-07-24 10:51:57 * da2ce7 hopes it i just a bug in bitcoin
1620 2011-07-24 10:52:36 <Eliel> da2ce7: if it's not getting transmitted, recreating it might work.
1621 2011-07-24 10:52:56 suriv has quit (Remote host closed the connection)
1622 2011-07-24 10:53:22 fingster has joined
1623 2011-07-24 10:53:28 <da2ce7> I'm gonna backup my wallet now.
1624 2011-07-24 10:53:36 <da2ce7> (made 100 new addresses)
1625 2011-07-24 10:53:38 suriv has joined
1626 2011-07-24 10:54:26 denisx_ has joined
1627 2011-07-24 10:56:13 erus` has joined
1628 2011-07-24 10:56:14 denisx_ has quit (Read error: Connection reset by peer)
1629 2011-07-24 10:57:27 denisx has quit (Ping timeout: 255 seconds)
1630 2011-07-24 10:57:53 dr_win has quit (Remote host closed the connection)
1631 2011-07-24 10:57:59 denisx has joined
1632 2011-07-24 10:59:15 dr_win has joined
1633 2011-07-24 11:01:39 <da2ce7> ok
1634 2011-07-24 11:01:48 <da2ce7> I have at least a domomant ammount of my coins.
1635 2011-07-24 11:02:19 <da2ce7> if some are stolen.... who would only steal a few?
1636 2011-07-24 11:02:33 <da2ce7> maybe a nicer hacker.
1637 2011-07-24 11:02:57 <denisx> how much is missing?
1638 2011-07-24 11:03:07 <da2ce7> don't know yet... going throogh my address now.
1639 2011-07-24 11:03:41 nus has quit (Ping timeout: 240 seconds)
1640 2011-07-24 11:05:28 <sacarlson> I just found  my problem with merge-mine-proxy  the password and username are reversed in the documentation help='connect to the aux RPC at this address (default: http://pw:un@127.0.0.1:8342/)
1641 2011-07-24 11:06:12 <sacarlson> this results in return from miner  "message": "Unknown error: No JSON object could be decoded",
1642 2011-07-24 11:06:27 <da2ce7> ok
1643 2011-07-24 11:06:32 denisx_ has joined
1644 2011-07-24 11:06:46 <da2ce7> how do I work where coins were send from?
1645 2011-07-24 11:07:19 dr_win has quit (Remote host closed the connection)
1646 2011-07-24 11:07:29 <sipa> use the -debug flag and click the transaction, it will tell you when double-clicking the tx
1647 2011-07-24 11:07:38 <da2ce7> thx
1648 2011-07-24 11:08:18 <sacarlson> so tomaro we should have a new test proto chain with merge mining using MultiCoin-exp
1649 2011-07-24 11:09:58 dr_win has joined
1650 2011-07-24 11:10:03 denisx has quit (Ping timeout: 255 seconds)
1651 2011-07-24 11:10:04 denisx_ is now known as denisx
1652 2011-07-24 11:11:25 Prof_BiG_BanG has quit (Changing host)
1653 2011-07-24 11:11:25 Prof_BiG_BanG has joined
1654 2011-07-24 11:13:06 aga has joined
1655 2011-07-24 11:13:48 fsntation has quit (Read error: Connection reset by peer)
1656 2011-07-24 11:16:34 <gmaxwell> da2ce7: a stuck transaction isn't usually a sign of theft...
1657 2011-07-24 11:16:51 <gmaxwell> if you were online you'd actually see the theif's transactions as if you'd made them.
1658 2011-07-24 11:16:59 <da2ce7> I'm not online.
1659 2011-07-24 11:17:11 <da2ce7> I leave my wallet offline most of the time.
1660 2011-07-24 11:17:12 <BlueMatt> well then of course the tx is stuck
1661 2011-07-24 11:17:19 <gmaxwell> If you're not online how do you know you have a stuck transaction??
1662 2011-07-24 11:17:24 <BlueMatt> the tx wont confirm if no one else knows about it
1663 2011-07-24 11:17:32 <da2ce7> I'm online now.
1664 2011-07-24 11:17:39 <da2ce7> and updated the block chain.
1665 2011-07-24 11:17:43 <BlueMatt> then you have to wait for a reabroadcast
1666 2011-07-24 11:18:20 <da2ce7> ah ok
1667 2011-07-24 11:18:26 <da2ce7> how long will that be?
1668 2011-07-24 11:18:59 <sipa> wait an hour
1669 2011-07-24 11:18:59 <gmaxwell> it's random with a maximum of .. er. hm I don't have the source handy atm.
1670 2011-07-24 11:19:20 <sipa> it's random, with a max of half an hour, iirc
1671 2011-07-24 11:19:27 <da2ce7> ok
1672 2011-07-24 11:19:32 nus has joined
1673 2011-07-24 11:19:38 <da2ce7> (fingers crossed)
1674 2011-07-24 11:19:46 <da2ce7> it looks like the transction is valid.
1675 2011-07-24 11:19:47 <da2ce7> :)
1676 2011-07-24 11:20:11 <sipa> unless you've done some wallet surgery, it probably is
1677 2011-07-24 11:20:27 <gmaxwell> or are running a modified client.
1678 2011-07-24 11:21:04 <da2ce7> na stock client.
1679 2011-07-24 11:21:15 <da2ce7> and stock (adbet old) wallet.
1680 2011-07-24 11:21:18 <gmaxwell> then it shouldn't be possible to create an invalid transaction.
1681 2011-07-24 11:22:28 * da2ce7 's heart starts to beat normal again.
1682 2011-07-24 11:23:43 clarkbox has quit (Ping timeout: 250 seconds)
1683 2011-07-24 11:25:23 karnac has joined
1684 2011-07-24 11:25:27 clarkbox has joined
1685 2011-07-24 11:26:48 denisx has quit (Remote host closed the connection)
1686 2011-07-24 11:27:34 cuddlefish has quit (Read error: Connection reset by peer)
1687 2011-07-24 11:28:09 Tiraspol has quit (Ping timeout: 252 seconds)
1688 2011-07-24 11:28:11 imsaguy has quit (Ping timeout: 260 seconds)
1689 2011-07-24 11:28:39 denisx_ has joined
1690 2011-07-24 11:31:45 koleg has quit (Read error: Connection reset by peer)
1691 2011-07-24 11:32:50 Tiraspol has joined
1692 2011-07-24 11:34:02 danbri has quit (Read error: Connection timed out)
1693 2011-07-24 11:34:56 flok has quit (Quit: ZNC - http://znc.sourceforge.net)
1694 2011-07-24 11:35:07 danbri has joined
1695 2011-07-24 11:35:21 <da2ce7> ooooh it confirmed!
1696 2011-07-24 11:35:22 <da2ce7> :)
1697 2011-07-24 11:37:02 * da2ce7 slowly starts to relax.
1698 2011-07-24 11:37:03 <da2ce7> :P
1699 2011-07-24 11:37:13 <da2ce7> my private key is safe.
1700 2011-07-24 11:37:16 <da2ce7> *still safe
1701 2011-07-24 11:38:37 flok has joined
1702 2011-07-24 11:39:06 Clipse has joined
1703 2011-07-24 11:39:22 <moa7> for now
1704 2011-07-24 11:39:35 dr_win has quit (Remote host closed the connection)
1705 2011-07-24 11:41:08 mmoya has joined
1706 2011-07-24 11:42:19 <gmaxwell> da2ce7: why are you freaking out?
1707 2011-07-24 11:42:26 <gmaxwell> s/are/were/
1708 2011-07-24 11:42:29 <Joric> speaking of safety... i've added downlodable wallet http://goo.gl/1iuEy :]
1709 2011-07-24 11:42:45 <gmaxwell> da2ce7: just because the stuck txn?
1710 2011-07-24 11:44:04 <da2ce7> yeah... I thought that if I made a transaction with old data, then a new block invalidated the coin's it used, that would make the (theif) transaction not appear.
1711 2011-07-24 11:44:23 <gmaxwell> Joric: you really shouldn't encourage people to use such weakly derrived private keys…
1712 2011-07-24 11:45:30 <gmaxwell> da2ce7: thats true except the theif would probably have shown up before you made the txn. But what you're describing could happen if you made a txn while offline.
1713 2011-07-24 11:45:41 <da2ce7> yep
1714 2011-07-24 11:45:52 <da2ce7> I did make the transaction offline... then went online to send it
1715 2011-07-24 11:46:07 <da2ce7> by mistake... didn't check the number of connections.
1716 2011-07-24 11:46:10 <gmaxwell> Yea, don't do that, you're going to always frustrate yourself with slow sends.
1717 2011-07-24 11:46:14 <gmaxwell> ah. okay.
1718 2011-07-24 11:51:40 Clipse has quit (Read error: Connection reset by peer)
1719 2011-07-24 11:53:17 <sipa> ipv6 support pull request is up :)
1720 2011-07-24 11:53:27 moa7 has quit (Quit: Page closed)
1721 2011-07-24 11:53:31 <da2ce7> sipa :) bigups!
1722 2011-07-24 11:53:32 <BlueMatt> do any dnsseeds support it yet?
1723 2011-07-24 11:53:36 <da2ce7> sipa++
1724 2011-07-24 11:54:30 <Zagitta> non-programming programmers = sad face
1725 2011-07-24 11:55:24 <sipa> BlueMatt: not that i know of
1726 2011-07-24 11:55:36 <BlueMatt> do you have a static ip I can add?
1727 2011-07-24 11:55:40 Rabbit67890 has joined
1728 2011-07-24 11:55:45 koleg has joined
1729 2011-07-24 11:55:45 koleg has joined
1730 2011-07-24 11:56:07 <sipa> BlueMatt: 2a02:348:5e:5a29::1
1731 2011-07-24 11:58:39 peper has quit (Quit: leaving)
1732 2011-07-24 11:59:59 Titeuf_87 has joined
1733 2011-07-24 12:00:05 dr_win has joined
1734 2011-07-24 12:00:10 koleg has joined
1735 2011-07-24 12:00:52 <Eliel> gmaxwell: are you saying Joric's method is incapable of producing trustworthy keys?
1736 2011-07-24 12:01:08 koleg has quit (Ping timeout: 240 seconds)
1737 2011-07-24 12:01:16 <Joric> It's not the same for everyone...
1738 2011-07-24 12:01:23 <gmaxwell> Ha, well anything from a webpage wouldn't be! but thats not what I was talking about.
1739 2011-07-24 12:01:31 Diablo-D3 has quit (Ping timeout: 252 seconds)
1740 2011-07-24 12:01:36 koleg has quit (2!kvirc@79.133.154.165|Ping timeout: 260 seconds)
1741 2011-07-24 12:01:42 <Eliel> at least to me, it feels like keys made that way could be quite secure if you use a good passphrase.
1742 2011-07-24 12:02:03 <Eliel> 6+ words makes for quite a bit of security.
1743 2011-07-24 12:02:07 <gmaxwell> If you ask users for passwords/phrases they will choose passphrases which are _far_ less secure than they think they are.
1744 2011-07-24 12:02:25 <gmaxwell> And their coins will be stolen by someone who does the computationally trivial search.
1745 2011-07-24 12:02:33 <Eliel> gmaxwell: yes, most will do that.
1746 2011-07-24 12:03:14 <gmaxwell> Right. And if you are one actually going to remember a string which is strong, you are not far from just having a 256 bit random string, and being completely sure of it.
1747 2011-07-24 12:03:49 <Eliel> Joric: 1. do make the javascript version that doesn't compromise the keys by transfering them through the net. 2. Add some warnings about using short passwords.
1748 2011-07-24 12:03:50 <sipa> also, the bitcoin community has access to some massive sha256 hashing power :)
1749 2011-07-24 12:04:09 <Eliel> gmaxwell: a passphrase with 6 words chosen properly has quite a bit of security and is not difficult to remember
1750 2011-07-24 12:04:33 <Joric> i'm wondering maybe that's why more than 10% mtgox passwords were recovered
1751 2011-07-24 12:04:47 <BlueMatt> sipa: and the only way an ipv6-only node will bootstrap is dnsseed, correct?
1752 2011-07-24 12:05:20 <gmaxwell> Eliel: six words? like six dictionary words? ... so maybe you'll get 48 bits of entropy. That is searchable.
1753 2011-07-24 12:05:21 <sipa> BlueMatt: it will join IRC, and connect to ipv4 nodes
1754 2011-07-24 12:05:33 <BlueMatt> sipa: ipv6-only, but ok
1755 2011-07-24 12:05:34 peper has joined
1756 2011-07-24 12:05:48 <gmaxwell> Joric: if what is why? people choosing poor passwords? sure.
1757 2011-07-24 12:06:04 <sipa> BlueMatt: if ipv6-only means "no connectivity to the ipv4 network", sure, but those people have to be lonely on the internet :D
1758 2011-07-24 12:06:10 <Joric> gmaxwell, gpu bruteforcing
1759 2011-07-24 12:06:18 <gmaxwell> Joric: of course
1760 2011-07-24 12:06:33 <sipa> i guess 5% was doable using trivial cpu search
1761 2011-07-24 12:06:41 <sipa> and another 5% using gpu bruteforcing :)
1762 2011-07-24 12:06:45 <diki> so..my bitcoin gui froze
1763 2011-07-24 12:06:50 <diki> please fix these freezes
1764 2011-07-24 12:07:10 Stellar has quit (Ping timeout: 246 seconds)
1765 2011-07-24 12:07:28 <Eliel> gmaxwell: hmm... yes, you're right. Just words is not enough.
1766 2011-07-24 12:08:30 <gmaxwell> Joric: 32% btw.
1767 2011-07-24 12:08:54 <Eliel> diki: I haven't seen a single freeze so far.
1768 2011-07-24 12:09:01 <gmaxwell> Eliel: but you thought it was. And I totally believe that you're not a foolish person. Thats why I think this sort of thing is a trap. :)
1769 2011-07-24 12:10:02 <Joric> gmaxwell, 20 thousands?
1770 2011-07-24 12:10:09 <gmaxwell> Joric: oh sorry. 24.4% I fail at math
1771 2011-07-24 12:10:22 <gmaxwell> (I computed cracked/uncracked, rather than cracked/total)
1772 2011-07-24 12:10:56 <diki> it happens when sending coins
1773 2011-07-24 12:11:07 <Joric> most ppl fail in math, especially in cryptography
1774 2011-07-24 12:11:09 <diki> and if in the future this also happens to some big shot...
1775 2011-07-24 12:11:51 <gmaxwell> And that was using 1BTC mining-equivalent computing time.
1776 2011-07-24 12:15:01 MrTiggr has quit (Ping timeout: 276 seconds)
1777 2011-07-24 12:15:21 <Joric> could anyone check how much time would take calculating a set of matching http://bitcoin-kamikaze.com 'honesty proofs'?
1778 2011-07-24 12:15:40 <Joric> i.e. how easy it would be to make 2 strings "(1,2,3,4,5)randomgarbage" and "(2,1,3,4,5)anothergarbage" with identical md5 hashes?
1779 2011-07-24 12:16:20 <Joric> i've read 8800/cuda calculates atleast 200m md5's a second but not quite sure about difficulty
1780 2011-07-24 12:16:51 <cjdelisle> 200m md5 or 200m md5crypt?
1781 2011-07-24 12:16:55 <cjdelisle> they are not the same
1782 2011-07-24 12:17:42 <cjdelisle> md5crypt is a relitively solid function which is interesting because it takes a long time to execute, thus cracking is difficult.
1783 2011-07-24 12:18:23 <cjdelisle> This is why linux still uses md5crypt hashes in it's /etc/shadow file, it's not because they are stupid.
1784 2011-07-24 12:19:06 <Zagitta> It's actually kinda creepy how little miners and pool software needs to be changed to have a near perfect distributed password cracker
1785 2011-07-24 12:20:55 <Joric> i was talking about the effective speed of md5 collision attack
1786 2011-07-24 12:21:49 <sipa> cjdelisle: as far as i can see, md5crypt does not use iterated hashing
1787 2011-07-24 12:21:54 <sipa> it does include a salt
1788 2011-07-24 12:22:03 diki has quit (Read error: Connection reset by peer)
1789 2011-07-24 12:22:06 <Joric> btw looks like it's about 5 billion hashes a second currently
1790 2011-07-24 12:22:11 <Joric> on 5970
1791 2011-07-24 12:22:11 Guest17379 has joined
1792 2011-07-24 12:22:23 <gmaxwell> Joric: I'm trying an attack now.
1793 2011-07-24 12:22:26 Guest17379 is now known as diki
1794 2011-07-24 12:22:43 <da2ce7> md5 should never be used... use at least whirlpool.
1795 2011-07-24 12:22:55 <sipa> it should be around 1.9x the number of bitcoin hashes that device does
1796 2011-07-24 12:23:54 <gmaxwell> Joric: https://people.xiph.org/~greg/msg1.bin
1797 2011-07-24 12:23:57 <gmaxwell> Joric: https://people.xiph.org/~greg/msg2.bin
1798 2011-07-24 12:24:14 <gmaxwell> not exactly the attack you wanted though.
1799 2011-07-24 12:24:32 <da2ce7> looks like Grøstl is the best next-gen hashing algorithm
1800 2011-07-24 12:24:48 <da2ce7> quite, novel
1801 2011-07-24 12:24:50 <sipa> why?
1802 2011-07-24 12:24:52 <gmaxwell> 'best' and 'next-gen' don't really go well togeather.
1803 2011-07-24 12:24:56 <cjdelisle> http://stackoverflow.com/questions/23511/algorithm-behind-md5crypt/155696#155696
1804 2011-07-24 12:24:59 <gmaxwell> Certantly 'novel' doesn't go well with 'best.
1805 2011-07-24 12:25:26 <da2ce7> gmaxwell, yeah... unless it is novel in a creative good way.
1806 2011-07-24 12:25:28 <gmaxwell> Joric: like my example?
1807 2011-07-24 12:25:56 <gmaxwell> da2ce7: novel means that is security depends on the non-existance of a novel attack. An unknown value.
1808 2011-07-24 12:25:59 <sipa> i prefer an old tried-and-true with some known potential attacks on 30 out of 64 rounds over a brand-new one with is super fast and has no known attacks
1809 2011-07-24 12:26:19 <Joric> gmaxwell, yep, but random garbage should be constrained :) fixed length and fixed set of shars
1810 2011-07-24 12:27:03 <gmaxwell> I told you the other day that constraining the garbage probably makes the attacks as known impossible.
1811 2011-07-24 12:27:18 <da2ce7> gmaxwell, well it is novel for a hash function... not novel cypto entirely... using standard maths techniques that are used elsewhere, such as in AES.
1812 2011-07-24 12:27:34 <Joric> gmaxwell, http://bitcoin-kamikaze.com there's a free play button, check those hashes yourself
1813 2011-07-24 12:27:36 <sipa> and those techniques have to be tested in the context of hashing
1814 2011-07-24 12:27:37 <gmaxwell> Certantly it does if the garbage is not very larged compared to the min(prefix,hash output size)
1815 2011-07-24 12:28:04 <gmaxwell> da2ce7: go look at the success of provably secure hashes.
1816 2011-07-24 12:28:12 <gmaxwell> They keep getting broken.
1817 2011-07-24 12:28:21 <da2ce7> ya
1818 2011-07-24 12:29:03 <Joric> gmaxwell, what speed you got?
1819 2011-07-24 12:29:15 <gmaxwell> But thats "using standard maths techniques that are used elsewhere" in fact, along with proofs about their behavior, and yet they fail because their assuptions fail, or they fail because the miss a property of hashes we consider important.
1820 2011-07-24 12:29:26 <gmaxwell> Joric: that took 26.92 seconds on my laptop.
1821 2011-07-24 12:29:46 <gmaxwell> I spent more time figuring out all the library flags for boost in order to compile the code.
1822 2011-07-24 12:30:10 <da2ce7> well have a lil read: https://secure.wikimedia.org/wikipedia/en/wiki/Gr%C3%B8stl
1823 2011-07-24 12:30:20 <cjdelisle> "good" hashes by their nature tend to resist attempts to prove anything about them.
1824 2011-07-24 12:30:29 <gmaxwell> I'm familar with it.
1825 2011-07-24 12:30:35 <da2ce7> k
1826 2011-07-24 12:30:51 <gmaxwell> (I'm on the NIST SHA-3 mailing list)
1827 2011-07-24 12:30:58 <da2ce7> kk
1828 2011-07-24 12:31:27 <senseles> its odd because this site when you hit a bomb
1829 2011-07-24 12:31:31 <senseles> it takes longer to load than if you succeed
1830 2011-07-24 12:31:45 <senseles> are you trying to figure out if it's legit or not?
1831 2011-07-24 12:32:19 <gmaxwell> (I was going to submit a joke entry to the SHA-3 contest— a hash based on the hex-expansion of Pi which was probably a random oracle if Pi is normal. But I got caught up on the proof and lost interest.
1832 2011-07-24 12:32:29 <gmaxwell> )
1833 2011-07-24 12:33:13 AStove has quit (Ping timeout: 276 seconds)
1834 2011-07-24 12:34:10 ar4s has left ()
1835 2011-07-24 12:35:43 Rabbit67890 has quit (Remote host closed the connection)
1836 2011-07-24 12:35:49 <da2ce7> gmaxwell, that sounds like fun :)
1837 2011-07-24 12:38:30 <da2ce7> gmaxwell, have you thought about making an 'emergency' version of bitcoin that changes the hashing algorithm, in the case that an adervercy gains much more than 50% hashing power (say goveremnt gets asic's)
1838 2011-07-24 12:38:38 Clipse has joined
1839 2011-07-24 12:39:02 <da2ce7> make a hashing algorithm that only works well on graphics gpu's...
1840 2011-07-24 12:39:28 <gmaxwell> da2ce7: well, what do you think gpus are made of? :)
1841 2011-07-24 12:39:44 <sipa> i'd prefer a hashing algorithm that only works on general purpose cpu's
1842 2011-07-24 12:39:53 MrTiggr has joined
1843 2011-07-24 12:40:04 <gmaxwell> sipa: you prefer bitcoin be mined almost entirely by botnet then? ;)
1844 2011-07-24 12:40:10 <da2ce7> we need something that makes use of the 1gb of high-speed ram that the gpu's have to themn
1845 2011-07-24 12:40:14 <sipa> gmaxwell: meh
1846 2011-07-24 12:40:40 <gmaxwell> da2ce7: A government attacker has much better options than just >50% mining.
1847 2011-07-24 12:40:46 <Eliel> yes, the specialized hardware requirement does make botnet mining quite a bit more ineffective :)
1848 2011-07-24 12:41:07 <gmaxwell> Like outlawing it and putting miners who are caught to death. Pretty effective and _much_ cheaper.
1849 2011-07-24 12:41:11 <da2ce7> and an algorithm that cannot easly be unwound and done in a single clock.
1850 2011-07-24 12:41:28 <Eliel> gmaxwell: that is unlikely to fly, at least in the short term, in any western democracy.
1851 2011-07-24 12:41:54 <Eliel> but throwing miners to jail is pretty close in effect.
1852 2011-07-24 12:42:17 <gmaxwell> Eliel: Attacking it explicitly would be unlawful, so if you're going to assue that the state is going to ignore the law or change the law then you can assume a lot of less expensive options.
1853 2011-07-24 12:42:25 <gmaxwell> Right. Fair enough.
1854 2011-07-24 12:42:25 <da2ce7> gmaxwell, it only cost say, 100M to make 99% mining power with asic's.
1855 2011-07-24 12:42:47 <gmaxwell> you don't need 99%..  60% would be fine.
1856 2011-07-24 12:42:54 <senseles> it'd be easier to just force exchanges to report transactions
1857 2011-07-24 12:43:00 <senseles> then tax the hell out of everyone with capital gains tax
1858 2011-07-24 12:43:01 <fingster> a license for mining?
1859 2011-07-24 12:43:06 <gmaxwell> and it costs less than 100M to get that just buying GPUs from AMD.
1860 2011-07-24 12:43:20 aga is now known as agath
1861 2011-07-24 12:43:26 <gmaxwell> senseles: you're not paying taxes on bitcoin sales?
1862 2011-07-24 12:43:29 <da2ce7> but gpu dies are biggg... would take a long lead-up time.
1863 2011-07-24 12:43:36 <senseles> no because i havent sold any
1864 2011-07-24 12:43:44 <gmaxwell> Fair enough.
1865 2011-07-24 12:44:29 <gmaxwell> da2ce7: They already exist. AMD sells >100M in gpus per quarter.
1866 2011-07-24 12:44:48 <gmaxwell> It doesn't change that it's stilly.
1867 2011-07-24 12:45:06 <gmaxwell> Outlawing it takes paper and a pen.
1868 2011-07-24 12:45:15 <da2ce7> hmm...
1869 2011-07-24 12:45:27 <senseles> most free nations couldn't do that though
1870 2011-07-24 12:45:28 <gmaxwell> Running TV ads "Bitcoin: The TERROR DOLLAR" is also fairly cheap.
1871 2011-07-24 12:45:33 <senseles> china could, and a few other tolitarian nations
1872 2011-07-24 12:45:39 <Eliel> if bitcoin exchange rates goes up much more, we might see mining capable card prices going through the roof.
1873 2011-07-24 12:45:42 <gmaxwell> senseles: then they couldn't attack it otherways.
1874 2011-07-24 12:45:46 <senseles> but even "terror dollar" or "kiddy porn buck"
1875 2011-07-24 12:45:47 <da2ce7> well it would be good to have... just to the goveremnt knows attacking throogh asic is going to be expencive.
1876 2011-07-24 12:45:51 <senseles> i dont think they'd ban it
1877 2011-07-24 12:45:58 <gmaxwell> Of course they wouldn't.
1878 2011-07-24 12:46:01 <gmaxwell> That would be stupid.
1879 2011-07-24 12:46:14 Tiraspol has quit (Ping timeout: 258 seconds)
1880 2011-07-24 12:46:27 <da2ce7> you will be supprised what an goverement dose when pressed aganst the econmomic wall.
1881 2011-07-24 12:46:29 <gmaxwell> Bitcoin is currently too unimportant to worry about, and not terribly problematic to the extent that it is important.
1882 2011-07-24 12:46:33 <senseles> kind of amazed that smaller governments havent got into the fold
1883 2011-07-24 12:46:40 Tiraspol has joined
1884 2011-07-24 12:47:03 <senseles> some place like nauru where they don't have any income since their mineral deposits are all mined out
1885 2011-07-24 12:47:07 <senseles> they've got nothing
1886 2011-07-24 12:47:17 <gmaxwell> da2ce7: long before any goverment felt "pressed against the economic wall" bitcoin would be 'too big to fail'.
1887 2011-07-24 12:47:22 karnac has quit (Quit: karnac)
1888 2011-07-24 12:47:23 <senseles> state sponsored bitcoin mining would bring a lot of income to a place where most people make less than 1$ a day
1889 2011-07-24 12:47:55 <da2ce7> gmaxwell, once we have open source asic chips... we have won.
1890 2011-07-24 12:48:10 Clipse has quit (Ping timeout: 255 seconds)
1891 2011-07-24 12:48:11 <senseles> even without asic, fpga are still pretty cost effective
1892 2011-07-24 12:48:21 <senseles> the thing about asic is that it's going to be expensive to get the latest tech
1893 2011-07-24 12:48:29 <senseles> 32nm would cost far more than >65nm
1894 2011-07-24 12:48:33 <gmaxwell> da2ce7: yea, this was an argument I made in the past: That we (the community) need asics to protect against someone else having them.
1895 2011-07-24 12:48:46 cjdelisle has quit (Read error: Connection reset by peer)
1896 2011-07-24 12:48:47 <da2ce7> yep
1897 2011-07-24 12:48:49 <senseles> so while everyone is using 65nm asic chips the FPGA'ers will be using <32nm chips
1898 2011-07-24 12:48:54 <sipa> strange, my ipv6 node only seems to make connections to itself :s
1899 2011-07-24 12:49:03 <gmaxwell> Thats okay, even at 45nm you can get such insane hash rates.. the impovement from top of the line process isn't _that_ much.
1900 2011-07-24 12:49:31 <gmaxwell> senseles: doesn't matter, you're looking at >>100x transistor inflation from FPGAs
1901 2011-07-24 12:49:35 <da2ce7> untill then we should have an emergency version of bitcoin spare that moves to something like SHA4096 or something.
1902 2011-07-24 12:49:53 <da2ce7> something that would be MUCH bigger to unroll
1903 2011-07-24 12:49:56 <gmaxwell> for 22nm FPGA would be smoked by a smaller (area wise) 45nm custom job.
1904 2011-07-24 12:50:10 <senseles> i just don't think anyone would pay for 45nm
1905 2011-07-24 12:50:16 <gmaxwell> da2ce7: then just don't unroll it.
1906 2011-07-24 12:50:18 <senseles> that's still too new cost wise
1907 2011-07-24 12:50:28 <gmaxwell> you can still pipeline a rolled engine
1908 2011-07-24 12:50:52 <gmaxwell> senseles: the MPW services have it, its not cost ineffective at all.
1909 2011-07-24 12:51:59 * Eliel wonders if it would be better to do one sha512 followed by sha256 than two sha256s
1910 2011-07-24 12:52:17 <gmaxwell> the only real barrier is that at 65nm and below you need fancy design tools because the wires end up slower than the gates, so the timing analysis gets tricky.
1911 2011-07-24 12:52:19 <sipa> only harder, not better
1912 2011-07-24 12:53:47 <Joric> speaking about technology - i'm wondering about 3d printers - there MUST be something that turns them into a gold mine right now, but i can't figure out what it may be :)
1913 2011-07-24 12:54:20 <sipa> just print gold!
1914 2011-07-24 12:54:31 <sipa> instant gold mine
1915 2011-07-24 12:54:31 * gmaxwell fires up the accelerator
1916 2011-07-24 12:54:39 <da2ce7> you could have an hashing algorithm that would create a 1gb file where each 'stream processer' mutates a small section of it, then it is hashed with a secure hash at the end.
1917 2011-07-24 12:54:50 <Eliel> Joric: I'm personally quite positive about 3d printers. If the prices go down a bit more, they'll definitely get quite popular.
1918 2011-07-24 12:54:52 <da2ce7> kinda like what a pixel shader works.
1919 2011-07-24 12:54:59 <da2ce7> but over a much larger file.
1920 2011-07-24 12:55:10 <Joric> gold expendables must be expensive
1921 2011-07-24 12:55:14 <davex__> someone should draw up a glbse contract or similar for ASIC to get investors and designers on board
1922 2011-07-24 12:55:15 <senseles> Joric: having a china in your house
1923 2011-07-24 12:55:18 * gmaxwell sends da2ce7 to the scrypt webpage.
1924 2011-07-24 12:55:31 <Eliel> Joric: although, I guess they need to be able to print metals in addition to plastic to be really useful.
1925 2011-07-24 12:55:36 <senseles> Joric: just need to sell an easy to use unit with tons of preprogrammed stuff that it can automatically build.
1926 2011-07-24 12:55:56 <senseles> stuff housewives would want
1927 2011-07-24 12:55:59 <gmaxwell> Wake me up when it can print itself.
1928 2011-07-24 12:56:02 <Eliel> senseles: or a good community website from where to get the stuff.
1929 2011-07-24 12:56:07 <senseles> gmaxwell: it can
1930 2011-07-24 12:56:11 <gmaxwell> not really.
1931 2011-07-24 12:56:40 <senseles> last time i checked was watching a youtube video where a guy had some modification that allowed it to solder PCB
1932 2011-07-24 12:56:47 <senseles> the rest of the plastic parts were easily built
1933 2011-07-24 12:57:02 <gmaxwell> you still have to add ICs, motors, gears, etc.
1934 2011-07-24 12:57:18 <senseles> well, it can partially build itself
1935 2011-07-24 12:57:51 <denisx_> python.org down?
1936 2011-07-24 12:57:56 <gmaxwell> Humans + water + carrots = more humans.  I want 3d printers that work like that.
1937 2011-07-24 12:57:57 <Eliel> I guess the tipping point is when they become able to print compact enough electronic circuits.
1938 2011-07-24 12:58:58 <senseles> Eliel: for use at home people don't want to have to search a community website. they just want to click a button and have it make what they want
1939 2011-07-24 12:59:10 <senseles> searching a website is too much work for your average user
1940 2011-07-24 12:59:14 <Joric> python.org is dead, really. at least!
1941 2011-07-24 12:59:17 <da2ce7> Eliel, the problem is when you get super small things get too 'sticky'
1942 2011-07-24 13:00:01 <da2ce7> things must work really differnty when really small.
1943 2011-07-24 13:00:26 <Eliel> senseles: searching a website or a large library of stuff to print... not much difference, frankly.
1944 2011-07-24 13:00:39 * Zagitta wonders about how many small PICS it would take to do bitcoin hashing
1945 2011-07-24 13:00:47 <senseles> other than downloading, importing, etc
1946 2011-07-24 13:01:03 <senseles> if it has a lib in a program on the desktop just click a few buttons and click print
1947 2011-07-24 13:01:12 <Eliel> senseles: that's not a difficult thing to integrate into the thing itself.
1948 2011-07-24 13:01:18 somuchwin has quit (Ping timeout: 252 seconds)
1949 2011-07-24 13:01:38 somuchwin has joined
1950 2011-07-24 13:02:01 <da2ce7> I suspect that the first self-replocating micro-mashine will be dna based.
1951 2011-07-24 13:02:26 <Eliel> senseles: a printer like that needs a computer in it anyway and a display to show the models.
1952 2011-07-24 13:03:02 <Eliel> so, it really would make no sense to not integrate things to a high degree, even with a community website.
1953 2011-07-24 13:03:10 <gmaxwell> da2ce7: well, it's believed that RNA based replicators came first. :)
1954 2011-07-24 13:03:17 <da2ce7> :)
1955 2011-07-24 13:03:28 <da2ce7> yada, yada, :P bio based.
1956 2011-07-24 13:03:34 <da2ce7> whatever form we incode the data.
1957 2011-07-24 13:04:07 <Eliel> there was a nice article introducing DNA from a programmers viewpoint I read once... I wonder if I can still find the link...
1958 2011-07-24 13:04:43 <Eliel> http://ds9a.nl/amazing-dna/
1959 2011-07-24 13:05:07 koleg has quit (3!kvirc@79.133.145.165|Read error: Connection reset by peer)
1960 2011-07-24 13:05:08 m00p has joined
1961 2011-07-24 13:05:38 Joric has quit (Ping timeout: 246 seconds)
1962 2011-07-24 13:06:12 Joric has joined
1963 2011-07-24 13:06:21 koleg has joined
1964 2011-07-24 13:06:35 <Joric> did anyone hear about the steganography method used in spore? they used png's to store 3d models
1965 2011-07-24 13:06:51 Stellar has joined
1966 2011-07-24 13:07:47 danbri has quit (Read error: Connection timed out)
1967 2011-07-24 13:08:41 danbri has joined
1968 2011-07-24 13:09:37 <Zagitta> Eliel: very interesting read
1969 2011-07-24 13:10:14 makomk has quit (Remote host closed the connection)
1970 2011-07-24 13:10:17 x6763 has joined
1971 2011-07-24 13:10:19 makomk has joined
1972 2011-07-24 13:12:18 cjdelisle has joined
1973 2011-07-24 13:12:36 <Eliel> Zagitta: that was my thought as well
1974 2011-07-24 13:12:57 <senseles> they already have dna coders dont they?
1975 2011-07-24 13:13:08 <senseles> you can just punch the buttons on the machine for what chemicals you want in sequence
1976 2011-07-24 13:13:12 TiggrBot has left ()
1977 2011-07-24 13:13:15 <senseles> and the machine just slaps the dna strand together
1978 2011-07-24 13:13:15 <senseles> ?
1979 2011-07-24 13:13:19 datagutt has joined
1980 2011-07-24 13:15:09 <Joric> at least some of git commits are science-based https://github.com/cariaso/dna/commit/dcc4b64f0b53f3dde2222b482e98ad2e869a3312 :)
1981 2011-07-24 13:15:49 huk has quit ()
1982 2011-07-24 13:16:52 mmoya has quit (Ping timeout: 260 seconds)
1983 2011-07-24 13:21:28 WildSoil has quit (Ping timeout: 240 seconds)
1984 2011-07-24 13:28:46 Zagitta has quit (Ping timeout: 252 seconds)
1985 2011-07-24 13:30:32 asuk has joined
1986 2011-07-24 13:31:19 fingster has quit (Ping timeout: 252 seconds)
1987 2011-07-24 13:34:46 danbri_ has joined
1988 2011-07-24 13:35:42 danbri has quit (Ping timeout: 240 seconds)
1989 2011-07-24 13:50:21 koleg has quit (Read error: Connection reset by peer)
1990 2011-07-24 13:50:48 blueadept has joined
1991 2011-07-24 13:52:53 koleg has joined
1992 2011-07-24 13:54:10 <da2ce7> still looking for a good linux admin who has exprence with iptables and openvpn.
1993 2011-07-24 13:54:29 <da2ce7> help me out on  getting a bitcoin site up and running.
1994 2011-07-24 13:54:42 jimon has quit (Excess Flood)
1995 2011-07-24 13:54:53 jimon has joined
1996 2011-07-24 13:55:00 <Eliel> da2ce7: what kind of a site?
1997 2011-07-24 13:55:06 <da2ce7> bitcoin.com.auy
1998 2011-07-24 13:55:08 dr_win has quit (Remote host closed the connection)
1999 2011-07-24 13:55:11 <da2ce7> *bitcoin.com.ay
2000 2011-07-24 13:55:16 <da2ce7> fuck... .au
2001 2011-07-24 13:55:43 <Eliel> is the idea that it'll be a localized bitcoin portal?
2002 2011-07-24 13:55:44 <da2ce7> umm... just a basic info site atm... we plan to add features soon.
2003 2011-07-24 13:56:07 <Eliel> I believe I could do what you need.
2004 2011-07-24 13:56:20 <Joric> kangaroo socks?
2005 2011-07-24 13:56:34 Happy0 has quit (Read error: Connection reset by peer)
2006 2011-07-24 13:56:35 <da2ce7> Joric, no not yet.
2007 2011-07-24 13:56:50 <da2ce7> we plan to add a email subscription like morning money.
2008 2011-07-24 13:57:04 <da2ce7> and provide community announcements.
2009 2011-07-24 13:57:05 <da2ce7> and news.
2010 2011-07-24 13:59:19 Happy0 has joined
2011 2011-07-24 13:59:52 <da2ce7> ;;gettrust Eliel
2012 2011-07-24 13:59:52 <gribble> Trust relationship from user da2ce7 to user Eliel: Level 1: 0, Level 2: 0 via 0 connections. Graph: http://serajewelks.bitcoin-otc.com/trustgraph.php?source=da2ce7&dest=Eliel
2013 2011-07-24 14:01:06 <Eliel> I haven't done any otc just yet
2014 2011-07-24 14:01:25 <da2ce7> k
2015 2011-07-24 14:01:53 <sacarlson> da2ce7: I didn't even know we had that on #bitcoin-dev  the gpg trust thing
2016 2011-07-24 14:02:14 <sipa> everywhere gribble lingers
2017 2011-07-24 14:04:18 wolfspraul has quit (Quit: leaving)
2018 2011-07-24 14:11:43 danbri_ is now known as danbri
2019 2011-07-24 14:12:43 Clipse has joined
2020 2011-07-24 14:15:31 Incitatus has joined
2021 2011-07-24 14:20:36 <Joric> does -rescan delete invalid transactions from the walllet?
2022 2011-07-24 14:20:43 <sipa> no
2023 2011-07-24 14:20:55 <sipa> it only adds and updates
2024 2011-07-24 14:21:17 <Joric> there probably should be an option to fix wallet
2025 2011-07-24 14:21:55 <sipa> i once wrote a patch that would detect transactions that conflict with the block chain or eachother
2026 2011-07-24 14:22:00 <sipa> but it was never merged
2027 2011-07-24 14:22:36 <sipa> mainly because it wasn't clear what the behaviour should be if a conflict was found
2028 2011-07-24 14:22:43 <diki> i oh man
2029 2011-07-24 14:22:48 <diki> an hour long fap
2030 2011-07-24 14:22:52 <diki> that hit the spot
2031 2011-07-24 14:22:54 <da2ce7> sipa, that would be a great patch... if it conflicts with the blockchain the transaction should go red... and after 100 blocks, it should be removed.
2032 2011-07-24 14:23:13 <sipa> da2ce7: yes, gavin eventually suggested something like that
2033 2011-07-24 14:24:39 anarchyx has joined
2034 2011-07-24 14:24:46 gasteve has quit (Quit: gasteve)
2035 2011-07-24 14:24:51 anarchyx has quit (Changing host)
2036 2011-07-24 14:24:51 anarchyx has joined
2037 2011-07-24 14:25:12 makomk has quit (Ping timeout: 240 seconds)
2038 2011-07-24 14:25:24 suriv_ has joined
2039 2011-07-24 14:27:35 Burgundy has quit (Ping timeout: 264 seconds)
2040 2011-07-24 14:27:42 <sipa> WTF?
2041 2011-07-24 14:27:47 denisx_ has quit (Quit: denisx_)
2042 2011-07-24 14:27:55 <sacarlson> sipa: I have seen a delete transaction would that fix it https://github.com/bitcoin/bitcoin/pull/256
2043 2011-07-24 14:28:22 <sipa> sacarlson: that's a last-resort solution, yes
2044 2011-07-24 14:28:34 <sipa> but it removes everything from the wallet, and recreates it
2045 2011-07-24 14:28:46 suriv has quit (Ping timeout: 250 seconds)
2046 2011-07-24 14:28:46 <sacarlson> sipa: It's one in my pull request
2047 2011-07-24 14:28:47 <sipa> so you lose all unconfirmed transactions, and all account information
2048 2011-07-24 14:29:38 <sipa> but wth, where did my pull req 264 go?
2049 2011-07-24 14:29:49 <sipa> it was merged, but apparently isn't in git head anymore
2050 2011-07-24 14:30:02 <sacarlson> sipa: have you seen this import export transaction method for offline transactions? https://github.com/piotrnar/bitcoin/tree/importexporttx
2051 2011-07-24 14:30:30 <b4epoche_> sipa:  would you have any idea how to track down this issue:  http://snapplr.com/kemb
2052 2011-07-24 14:30:42 <sacarlson> sipa: if so how do you feel about it?
2053 2011-07-24 14:31:18 <b4epoche_> basically it swamps both cores
2054 2011-07-24 14:31:23 <sipa> b4epoche_: :s
2055 2011-07-24 14:31:33 BitcoinForNewegg has quit ()
2056 2011-07-24 14:31:45 makomk has joined
2057 2011-07-24 14:31:45 <sipa> sacarlson: something like that will definitely be necessary somewhere in the future, but i think we've other things to nail now
2058 2011-07-24 14:32:16 <sacarlson> sipa: well I plan to merge it into MultiCoin
2059 2011-07-24 14:32:30 <b4epoche_> sipa:  well crap (or great) it's not doing it now
2060 2011-07-24 14:32:56 MrTiggr has quit (Ping timeout: 240 seconds)
2061 2011-07-24 14:33:05 <b4epoche_> it was doing it consistently yesterday.  very strange.
2062 2011-07-24 14:33:13 MrTiggr has joined
2063 2011-07-24 14:34:05 Kolky has joined
2064 2011-07-24 14:36:48 topi`_ has quit (Remote host closed the connection)
2065 2011-07-24 14:37:06 <sipa> crap #264 was lost in wallet class merge :S
2066 2011-07-24 14:37:22 <wumpus> uh oh
2067 2011-07-24 14:37:55 <sacarlson> sipa I've had my own merges loose things like that didn't find out till weeks later
2068 2011-07-24 14:39:55 E-sense has quit (Ping timeout: 252 seconds)
2069 2011-07-24 14:41:08 <sacarlson> I'm looking forward to some standard method of plugins that delete and add classes and somehow auto manage conflicts
2070 2011-07-24 14:41:13 p0s has joined
2071 2011-07-24 14:42:34 <wumpus> it's possible to use your own merging with git, not that I know of any good mergers that really take the c++ code into account...
2072 2011-07-24 14:44:39 <sacarlson> wumpus: well that's what I presenlty use git to merge but with many people working on the same project in a short time it becomes very hard to merge older code
2073 2011-07-24 14:45:40 <rlifchitz> ;;bc,stats
2074 2011-07-24 14:45:42 <gribble> Current Blocks: 137808 | Current Difficulty: 1690906.2047244 | Next Difficulty At Block: 139103 | Next Difficulty In: 1295 blocks | Next Difficulty In About: 1 week, 1 day, 14 hours, 50 minutes, and 25 seconds | Next Difficulty Estimate: 1766710.88241154
2075 2011-07-24 14:46:57 denisx has joined
2076 2011-07-24 14:47:38 BitcoinForNewegg has joined
2077 2011-07-24 14:47:48 BitcoinForNewegg has quit (Client Quit)
2078 2011-07-24 14:47:54 BitcoinForNewegg has joined
2079 2011-07-24 14:51:51 fsntation has joined
2080 2011-07-24 14:54:42 Marf has joined
2081 2011-07-24 14:56:20 <wumpus> sacarlson: it's crazy that we're living in 2011 and still doing automerging based on literal text content and lines instead of internal knowledge of c++ parse trees, it could be so much better
2082 2011-07-24 14:58:08 <sipa> a few years ago, we weren't doing automerging at all :)
2083 2011-07-24 14:58:51 eastender has joined
2084 2011-07-24 14:59:11 <sacarlson> wumpus: I guess there are some pointing into the lib direction that will hopfully have a standard plugin method of destruct and create classes,  I'm not really a programer just an integrator
2085 2011-07-24 14:59:36 AStove has joined
2086 2011-07-24 15:00:49 <sacarlson> but who knows maybe some day I can become one
2087 2011-07-24 15:00:49 m00p has quit (Ping timeout: 252 seconds)
2088 2011-07-24 15:00:51 <wumpus> sipa: auto merging based on text context is pretty old, I think, it's basically the same idea of diff/patch which have existed forever
2089 2011-07-24 15:01:17 <lfm> wumpus: we should all just move to 2021
2090 2011-07-24 15:01:44 <wumpus> lfm: lol
2091 2011-07-24 15:02:19 <sacarlson> lfm: I'm sure I'll be getting more senile than I am already by 2021
2092 2011-07-24 15:02:35 <AlonzoTG> om
2093 2011-07-24 15:02:46 * AlonzoTG has been reading through more of the docs and papers on the wiki.
2094 2011-07-24 15:02:54 <wumpus> lfm: would be interesting to wake up every 10 years and see when commonplace annoyances have finally been solved :P
2095 2011-07-24 15:02:59 <AlonzoTG> the docs are pretty good if not quite detailed enough.
2096 2011-07-24 15:03:40 <wumpus> yes the docs in the wiki are surprisingly good
2097 2011-07-24 15:03:50 <AlonzoTG> right now I'm trying to develop a mathematical understanding of the database operations on the transactions and blocks.
2098 2011-07-24 15:04:08 <AlonzoTG> so that I can develop a database engine for them.
2099 2011-07-24 15:04:20 <AlonzoTG> that can be audited, verified, and unit-tested.
2100 2011-07-24 15:04:42 <wumpus> sounds good
2101 2011-07-24 15:04:48 <sacarlson> AlonzoTG: I didn't know databases were very mathematicle
2102 2011-07-24 15:04:55 <AlonzoTG> Oh yes!
2103 2011-07-24 15:05:19 BlueMattBot has quit (Ping timeout: 276 seconds)
2104 2011-07-24 15:05:24 <AlonzoTG> Go wikipedia "relational algebra" It is the underlying theoretical framework for relational databases.
2105 2011-07-24 15:05:38 <AlonzoTG> This is somewhat different, it forms a multi-headed tree structure,
2106 2011-07-24 15:05:42 <wumpus> yes queries are logical mathematics
2107 2011-07-24 15:05:48 <sacarlson> AlonzoTG: I mostly play with mysql that has some math in it but I normaly do math in software other than increment and sort
2108 2011-07-24 15:05:52 WakiMiko_ has joined
2109 2011-07-24 15:06:04 Nicksasa has joined
2110 2011-07-24 15:06:47 <sipa> sacarlson: math is not the same thing as calculation :)
2111 2011-07-24 15:06:57 Technic has joined
2112 2011-07-24 15:07:02 Technic has quit (Client Quit)
2113 2011-07-24 15:07:17 <sacarlson> sipa: like sum of a colum?
2114 2011-07-24 15:07:38 <AlonzoTG> go read: http://en.wikipedia.org/wiki/Relational_algebra
2115 2011-07-24 15:07:51 <sipa> sacarlson: math is about general truths that hold for abstract notions
2116 2011-07-24 15:07:59 <sipa> sacarlson: calculation is playing with numbers
2117 2011-07-24 15:08:30 <sacarlson> sipa: well all I can do is accounting math I leave to you guys
2118 2011-07-24 15:08:31 WakiMiko has quit (Ping timeout: 240 seconds)
2119 2011-07-24 15:10:45 Zagitta has joined
2120 2011-07-24 15:11:35 danbri has quit (Read error: Connection timed out)
2121 2011-07-24 15:12:26 danbri has joined
2122 2011-07-24 15:12:37 <CIA-103> bitcoin: Pieter Wuille master * ra7dd11c / src/wallet.cpp : (log message trimmed)
2123 2011-07-24 15:12:37 <CIA-103> bitcoin: Fix for small change outputs
2124 2011-07-24 15:12:37 <CIA-103> bitcoin: With the separation of CENT and MIN_TX_FEE, it is now reasonable
2125 2011-07-24 15:12:37 <CIA-103> bitcoin: to create change outputs between 0.01 and 0.0005, as these are
2126 2011-07-24 15:12:37 <CIA-103> bitcoin: spendable according to the policy, even though they require a fee
2127 2011-07-24 15:12:38 <CIA-103> bitcoin: to be paid.
2128 2011-07-24 15:12:38 <CIA-103> bitcoin: Also, when enough fee was already present, everything can go into
2129 2011-07-24 15:12:39 <CIA-103> bitcoin: Pieter Wuille master * rb6d19ae / src/wallet.cpp :
2130 2011-07-24 15:12:40 <CIA-103> bitcoin: Merge pull request #428 from sipa/mintxfeefix
2131 2011-07-24 15:12:41 <CIA-103> bitcoin: Fix for small change outputs - https://github.com/bitcoin/bitcoin/commit/b6d19aefa0cf455dca57ca74eeb75e158a9390d0
2132 2011-07-24 15:12:55 <sipa> you call that "log message trimmed" ?
2133 2011-07-24 15:13:17 <wumpus> that's a novel, not a log message :p
2134 2011-07-24 15:13:44 <sipa> well, i like descriptive messages in public commits
2135 2011-07-24 15:13:55 <sipa> but they shouldn't be copied in full to IRC
2136 2011-07-24 15:14:13 <wumpus> yeah exactly....
2137 2011-07-24 15:14:34 <sacarlson> so min_tx_fee has chaged again?
2138 2011-07-24 15:14:45 <sipa> sacarlson: no, not at all
2139 2011-07-24 15:15:33 <sipa> it's a commit that was merged 2 months ago, and lost
2140 2011-07-24 15:16:13 Akiron has joined
2141 2011-07-24 15:16:25 <sipa> it was a bug that no subcent changes were created
2142 2011-07-24 15:17:33 karnac has joined
2143 2011-07-24 15:21:19 BlueMattBot has joined
2144 2011-07-24 15:21:43 <sacarlson> sipa: any think else you want me to add to MultiCoin that I'm missing?
2145 2011-07-24 15:22:05 asuk has quit (Ping timeout: 252 seconds)
2146 2011-07-24 15:22:38 <sacarlson> sipa: or take away for that mater
2147 2011-07-24 15:23:11 <sipa> it's your project, you're the judge over what follows its design principles
2148 2011-07-24 15:25:03 <sacarlson> sipa: I'm always open for critisism and sugestions
2149 2011-07-24 15:25:39 <BitcoinForNewegg> you are fat and ugly
2150 2011-07-24 15:25:48 <BitcoinForNewegg> go on a diet and get pastic surgery
2151 2011-07-24 15:26:02 <BitcoinForNewegg> sipa: I'm always open for critisism and sugestions
2152 2011-07-24 15:26:04 <BitcoinForNewegg> :P
2153 2011-07-24 15:26:36 <sacarlson> BitcoinForNewegg: I should cut down on all that beer I drink then ;-)
2154 2011-07-24 15:26:55 <BitcoinForNewegg> save it for others and it solves both problems
2155 2011-07-24 15:29:19 topi` has joined
2156 2011-07-24 15:31:12 <BlueMattBot> Project Bitcoin-Test build #13: STILL FAILING in 10 sec: http://www.bluematt.me/jenkins/job/Bitcoin-Test/13/
2157 2011-07-24 15:31:13 <BlueMattBot> pieter.wuille: Fix for small change outputs
2158 2011-07-24 15:31:59 eastender has quit (Ping timeout: 252 seconds)
2159 2011-07-24 15:33:48 topi` has quit (Remote host closed the connection)
2160 2011-07-24 15:33:52 topi` has joined
2161 2011-07-24 15:34:42 TiggrBot has joined
2162 2011-07-24 15:36:09 larsivi has quit (Ping timeout: 246 seconds)
2163 2011-07-24 15:40:48 eastender has joined
2164 2011-07-24 15:41:08 Tiraspol has quit (Ping timeout: 240 seconds)
2165 2011-07-24 15:42:06 Tiraspol has joined
2166 2011-07-24 15:43:46 Theo` has left ()
2167 2011-07-24 15:46:17 Joric has quit ()
2168 2011-07-24 15:47:47 phuzion has quit (Ping timeout: 260 seconds)
2169 2011-07-24 15:49:01 TheZagitta has joined
2170 2011-07-24 15:49:05 phuzion has joined
2171 2011-07-24 15:49:12 shawn-p has quit (Ping timeout: 240 seconds)
2172 2011-07-24 15:51:16 Zagitta has quit (Ping timeout: 258 seconds)
2173 2011-07-24 15:52:45 osmosis has joined
2174 2011-07-24 15:54:28 <senseles> i dont understand why transaction fees go to the miners
2175 2011-07-24 15:54:35 <senseles> the transaction fees should go to the peers a user is connected to
2176 2011-07-24 15:54:37 <senseles> no?
2177 2011-07-24 15:55:27 <freewil> senseles, the miners are peers
2178 2011-07-24 15:55:59 <mtrlt> and miners are the ones that confirm the transactions.
2179 2011-07-24 15:56:07 <mtrlt> i.e. add them to the block chain
2180 2011-07-24 15:56:08 <cjdelisle> because the miner is the one who decides if you're sorry transaction is going to make the chain, you really want to send him flowers and a bottle of wine tiwh that transaction.
2181 2011-07-24 15:56:28 <freewil> ha
2182 2011-07-24 15:56:41 <senseles> :(
2183 2011-07-24 15:57:35 * Eliel wonders if it would make sense to tie the sizes of the default transaction fee sizes to network hashrate. as in, hashrate drops, fees go up, hashrate goes up, fees go down.
2184 2011-07-24 15:58:17 shawn-p has joined
2185 2011-07-24 15:59:37 <sipa> senseles: that's not a bad question
2186 2011-07-24 16:00:10 <sipa> i think it would actually have been better if there was some way of awarding the non-mining part of the network
2187 2011-07-24 16:00:26 <gmaxwell> senseles: try to figure out how that would work.
2188 2011-07-24 16:00:29 <sipa> i believe satoshi envisioned +- everyone to be a miner
2189 2011-07-24 16:00:44 <gmaxwell> And yea I was about to say "the non-mining part of the network?"
2190 2011-07-24 16:00:50 <Eliel> I believe the initial idea was that every node would mine but... it unexpectedly didn't quite work out that way
2191 2011-07-24 16:00:52 <gmaxwell> Oh well.
2192 2011-07-24 16:00:55 <TheZagitta> award for what? In most modern countries bandwidth is free and storage doesn't exactly cost much either
2193 2011-07-24 16:01:07 <sipa> not much != nothing
2194 2011-07-24 16:02:24 <senseles> what do you guys think about hardcoding some nodes into the bitcoin client by ip? so if irc isnt available everything continues to function?
2195 2011-07-24 16:02:27 <phantomcircuit> sipa, the marginal cost of bandwidth/storage is 0
2196 2011-07-24 16:02:36 <phantomcircuit> senseles, already done
2197 2011-07-24 16:02:54 <senseles> can i get my dedicated node added to the list?
2198 2011-07-24 16:03:06 <senseles> been running a client on a server @ softlayer for quite awhile now
2199 2011-07-24 16:03:23 <sipa> senseles: in that case your node probably made it into the DNS seeder's lists
2200 2011-07-24 16:03:45 <senseles> dns seeder's list?
2201 2011-07-24 16:03:51 <sipa> senseles: there are: 1) hardcoded IP's 2) IRC bootstrapping 3) DNS seeding 4) -addnode 5) the bitcoin P2p network
2202 2011-07-24 16:03:59 <sipa> as sources of nodes to connect to
2203 2011-07-24 16:04:31 <sipa> since 0.3.24, dns seeding is on by default
2204 2011-07-24 16:05:18 <senseles> maybe ill build a small website where people can submit supernode hubs
2205 2011-07-24 16:05:30 <senseles> so users can addnode the hubs if they'd like
2206 2011-07-24 16:06:33 <sipa> https://en.bitcoin.it/wiki/Fallback_Nodes
2207 2011-07-24 16:07:20 <sipa> hmm, the list used to be a lot biffer
2208 2011-07-24 16:07:35 <senseles> can i add my ip to the list somehow?
2209 2011-07-24 16:07:43 <sipa> it's a wiki...
2210 2011-07-24 16:07:47 <phantomcircuit> lol
2211 2011-07-24 16:08:49 <senseles> well ya, but i thought maybe there was something special about it
2212 2011-07-24 16:08:57 <senseles> since it has what looks like some sort of automated checker
2213 2011-07-24 16:09:25 <sipa> it has an automated checker, yes
2214 2011-07-24 16:12:49 <gmaxwell> I think the automated checker went nuts and removed a bunch of working nodes because they filled up or something like that.
2215 2011-07-24 16:13:11 <lfm> thats ok
2216 2011-07-24 16:14:21 <sipa> they were re-added by WakiMiko, and subsequently removed by the checker
2217 2011-07-24 16:14:27 <TheZagitta> any of you linux fokes that happen to know anything about scsi hotswapping? my ubuntu server won't see the 2 drives i just added to it and i'd prefer to not restart :3
2218 2011-07-24 16:14:57 <JFK911> so you scanned the scsi bus and they didn't appear?
2219 2011-07-24 16:14:58 <MrTiggr> ;;bc,,wiki Fallback_Nodes
2220 2011-07-24 16:14:59 <gribble> Error: "bc,,wiki" is not a valid command.
2221 2011-07-24 16:15:02 <MrTiggr> :S
2222 2011-07-24 16:15:58 <gmaxwell> echo "- - - crapIcan'tremember" > /proc/morecrapIcan'trembmer'/scsi/0
2223 2011-07-24 16:16:34 TheZimm has joined
2224 2011-07-24 16:16:42 <TheZagitta> crap i can't remember? :P
2225 2011-07-24 16:16:59 justmoon has joined
2226 2011-07-24 16:17:14 <phantomcircuit> TheZagitta, http://misterd77.blogspot.com/2007/12/how-to-scan-scsi-bus-with-26-kernel.html
2227 2011-07-24 16:17:15 <sipa> ;;bc,wiki Fallback_nodes
2228 2011-07-24 16:17:15 <gribble> https://en.bitcoin.it/wiki/Fallback_Nodes | Fallback Nodes. From Bitcoin. Jump to: navigation, search. This is a list of nodes which are considered reliable. Nodes from this list which are down for ...
2229 2011-07-24 16:17:29 <MrTiggr> well done
2230 2011-07-24 16:17:31 <MrTiggr> :D
2231 2011-07-24 16:17:34 <gmaxwell> TheZagitta: surely you're as able to goole as I am. Perhaps more so. :)
2232 2011-07-24 16:17:58 <JFK911> part of googling is understanding the problem and knowing what you need
2233 2011-07-24 16:18:12 eastender has quit (Ping timeout: 252 seconds)
2234 2011-07-24 16:18:22 <gmaxwell> well, thats why I pointed out that it's needed to echo some crap to a file in proc. :)
2235 2011-07-24 16:18:28 <BlueMatt> sipa: you should probably add your ipv6 node to fallback nodes page...
2236 2011-07-24 16:18:56 <senseles> do you have to patch for ipv6?
2237 2011-07-24 16:18:59 <senseles> or how does that work
2238 2011-07-24 16:18:59 <TheZagitta> gmaxwell: already googled for a while however and i saw that echoing to that file but it wasn't sure about it
2239 2011-07-24 16:19:12 <gmaxwell> Thats what you need to do. It'll do magic.
2240 2011-07-24 16:19:16 <sipa> BlueMatt: good point
2241 2011-07-24 16:19:59 <sipa> BlueMatt: but i'd better wait until it's merged
2242 2011-07-24 16:20:06 <BlueMatt> yea, probably
2243 2011-07-24 16:20:10 <BlueMatt> best
2244 2011-07-24 16:20:21 <sipa> senseles: http://github.com/sipa/bitcoin/tree/ipv6
2245 2011-07-24 16:20:47 <BlueMatt> can I ask why you didnt just do everything ipv6 and map ipv4 to ipv6?
2246 2011-07-24 16:21:08 <gmaxwell> sipa: can I run your node out of diskspace by addr.dat rumoring the whole ipv6 address space?
2247 2011-07-24 16:21:31 <senseles> is it possible to do a dnat forwarding on port 8333 ipv4 to ip port 8333 ipv6?
2248 2011-07-24 16:21:39 <senseles> (with iptables)
2249 2011-07-24 16:21:47 <BlueMatt> senseles: you have to have an ipv6 address, but I think so
2250 2011-07-24 16:21:53 <Dagger2> senseles: no
2251 2011-07-24 16:21:54 <TheZagitta> ... /sys/class/scsi_host/host3/scan: Permission denied :3 (using sudo)
2252 2011-07-24 16:21:58 <sipa> gmaxwell: quite possible that you can
2253 2011-07-24 16:21:59 <senseles> ya, i didnt think it would be
2254 2011-07-24 16:22:18 <BlueMatt> TheZagitta: did you sudo tee or sudo echo > ...?
2255 2011-07-24 16:22:23 Stellar has quit (Read error: Connection reset by peer)
2256 2011-07-24 16:22:45 <sipa> BlueMatt: compatibility... so you can comment out the IPv6 part, and not depend on any IPv6 presence in the OS
2257 2011-07-24 16:23:09 <BlueMatt> if an os doesnt have ipv6, bitcoin isnt gonna compile on it anyway
2258 2011-07-24 16:23:09 <senseles> sipa: any .tar.gz? im running centos so there is no git in the rpm repo :/
2259 2011-07-24 16:23:11 <gmaxwell> TheZagitta: I expect your echo got run as root, the redirection as you. :)
2260 2011-07-24 16:23:17 <BlueMatt> xp sp what 2? had ipv6
2261 2011-07-24 16:23:19 <TheZagitta> BlueMatt: sudo echo "- - -" > /sys/class/scsi_host/host3/scan
2262 2011-07-24 16:23:38 <BlueMatt> TheZagitta: you are echoing as root, not writing try echo "..." | sudo tee /sys...
2263 2011-07-24 16:23:53 <sipa> gmaxwell: but i think you can already kill nodes now by creating addr messages for the whole ipv4 space, no?
2264 2011-07-24 16:23:55 <senseles> just get a root shell and quit screwing with sudo
2265 2011-07-24 16:24:10 <senseles> sudo's only good use is for limited access to root apps as web user
2266 2011-07-24 16:24:11 <BlueMatt> sipa: do current nodes not forward ipv6 addresses as well?
2267 2011-07-24 16:24:21 <gmaxwell> I thought it did a isroutable check
2268 2011-07-24 16:24:32 <gmaxwell> (before saving/forwarding any of them)
2269 2011-07-24 16:24:40 <BlueMatt> mm, maybe, Im just asking
2270 2011-07-24 16:24:51 <senseles> how hard would it be to get a dual stack client?
2271 2011-07-24 16:24:54 <TheZagitta> senseles: i'm using ssh and su doesn't work for some stupid reason i haven't cared to look into yet
2272 2011-07-24 16:25:01 <senseles> thezagitta: sudo bash
2273 2011-07-24 16:25:04 <sipa> BlueMatt: no, they're ignored now
2274 2011-07-24 16:25:09 <BlueMatt> ah, ok
2275 2011-07-24 16:25:15 <sipa> line 1912 in main.cpp in current git head
2276 2011-07-24 16:25:16 <BlueMatt> TheZagitta: sudo su doesnt work?
2277 2011-07-24 16:25:44 <BlueMatt> su wont work on default debian as root does have a password, sudo su will
2278 2011-07-24 16:25:46 <BlueMatt> or sudo bash
2279 2011-07-24 16:25:51 <TheZagitta> senseles: that worked
2280 2011-07-24 16:26:04 <sipa> AddAddress ingores unroutable addresses
2281 2011-07-24 16:26:14 <BlueMatt> ah, ok
2282 2011-07-24 16:26:26 <sipa> but the call to AddAddress in ProcessMessage doesn't check AddAddress's return status
2283 2011-07-24 16:26:27 <BlueMatt> so...we should do addr.dat pruning before pulling ipv6?
2284 2011-07-24 16:26:32 <sipa> probably
2285 2011-07-24 16:26:58 <TheZagitta> thanks guys that made the drives show up :D
2286 2011-07-24 16:27:16 <BlueMatt> well first off, Id say nodes that send more than 1k addresses should be dropped/addr messages ignored (or does that already happen?)
2287 2011-07-24 16:27:17 <senseles> woot
2288 2011-07-24 16:27:44 <senseles> im liking this patch thezagitta gave me with a -hub option
2289 2011-07-24 16:27:52 <senseles> but if im running a hub, wouldn't i send all 1k addresses to anyone who connects?
2290 2011-07-24 16:28:07 <senseles> would it be possible to just send 50 at a time, or limit the number that get sent in some way?
2291 2011-07-24 16:29:01 <sipa> by the way, if there are any other known large tunnel brokers where the client-dependent part is below /32... let me know
2292 2011-07-24 16:29:08 Namegduf has quit (Ping timeout: 240 seconds)
2293 2011-07-24 16:29:23 <sipa> sixxs uses different PoP's, with different IPv6 ranges for each
2294 2011-07-24 16:29:27 <senseles> what do you mean?
2295 2011-07-24 16:29:45 <senseles> by "client-dependent part below /32"
2296 2011-07-24 16:30:05 <sipa> well, he.net gives addresses of the form AAAA:BBBB:XXXX:YYYY:..., where AAAA:BBBB is constant for whole of he.net
2297 2011-07-24 16:30:21 <senseles> ah
2298 2011-07-24 16:30:23 <sipa> which means all he.net customers will end up in the same /32 group
2299 2011-07-24 16:30:31 <sipa> and bitcoin will never connect to more than one of those
2300 2011-07-24 16:30:33 <senseles> where do you live?
2301 2011-07-24 16:30:53 <senseles> .be?
2302 2011-07-24 16:30:56 <sipa> yes
2303 2011-07-24 16:30:57 osmosis has quit (Ping timeout: 260 seconds)
2304 2011-07-24 16:31:10 <senseles> i think i can get ipv6 in NL and UK if you wanted a static /64
2305 2011-07-24 16:31:19 <senseles> been meaning to get ipv6 working on my vpn anyway
2306 2011-07-24 16:31:25 <senseles> radvd is pissing me off something terrible
2307 2011-07-24 16:31:42 BlueMatt has quit (Ping timeout: 260 seconds)
2308 2011-07-24 16:32:20 <senseles> most server locations i host at are all runing native ipv6
2309 2011-07-24 16:32:28 <senseles> you could probably get a small VPS and do it yourself
2310 2011-07-24 16:32:54 <sipa> my VPS has native ipv6
2311 2011-07-24 16:33:19 <senseles> there you go, setup an xl2tpd & ipsec servers
2312 2011-07-24 16:35:30 Namegduf has joined
2313 2011-07-24 16:39:21 BlueMatt has joined
2314 2011-07-24 16:41:55 <TheZagitta> ones router and ISP needs to support ipv6 before being able to get an address right?
2315 2011-07-24 16:42:08 <sipa> yes
2316 2011-07-24 16:42:18 <senseles> you can get an ip for free from a dozen different brokers
2317 2011-07-24 16:42:20 <BlueMatt> senseles: my connection cut out and this message didnt go through:
2318 2011-07-24 16:42:23 <BlueMatt> <BlueMatt> DONT EVER USE THE -hub OPTION UNLESS YOU KNOW WHAT YOU ARE DOING
2319 2011-07-24 16:42:39 <sipa> BlueMatt: btw, a failed build message was shown here
2320 2011-07-24 16:42:41 <senseles> some private vpn companies offer ipv6, or you can get a VPS and setup your own vpn
2321 2011-07-24 16:42:45 <Dagger2> no, tunnels work through braindead ISPs (and, depending on the tunnel type, some routers)
2322 2011-07-24 16:42:46 <sipa> but i can't access jenkins
2323 2011-07-24 16:42:47 <senseles> bluematt: why? what does it do?
2324 2011-07-24 16:42:58 <BlueMatt> sipa: yep, happens after every commit, the current bitcoin testsuite doesnt build
2325 2011-07-24 16:43:01 * gmaxwell cries
2326 2011-07-24 16:43:10 <BlueMatt> senseles: then you shouldnt be using it
2327 2011-07-24 16:43:12 T_X has quit (Ping timeout: 260 seconds)
2328 2011-07-24 16:43:24 <BlueMatt> sipa: theres a pull, but its only partial and the guy hasnt updated it...
2329 2011-07-24 16:43:27 <senseles> i want to have a server that can accept 1000+ connections
2330 2011-07-24 16:43:37 <senseles> sounded like that was what it was for when i read the patch notes
2331 2011-07-24 16:43:37 <BlueMatt> senseles: so use -maxconnections=...
2332 2011-07-24 16:43:44 <BlueMatt> no, its not
2333 2011-07-24 16:43:51 <sipa> even with maxconnections, it is limited to 125
2334 2011-07-24 16:44:04 <gmaxwell> hm? I don't recall that.
2335 2011-07-24 16:44:08 <BlueMatt> sipa: no its not
2336 2011-07-24 16:44:12 <sipa> no? :o
2337 2011-07-24 16:44:13 <BlueMatt> mine gets 150 all the time
2338 2011-07-24 16:44:16 <gmaxwell> No.
2339 2011-07-24 16:44:21 <BlueMatt> I think its limited to what? 1024?
2340 2011-07-24 16:44:35 <sipa> oh, right
2341 2011-07-24 16:44:36 <gmaxwell> Well, you'll run out of FDs unless you deal with that.
2342 2011-07-24 16:44:37 <BlueMatt> whatever the max select will let you use iirc
2343 2011-07-24 16:44:55 <gmaxwell> though you should be careful, because if you run out of fd's it'll crap on the database. :)
2344 2011-07-24 16:45:14 <gmaxwell> but 1000 is fine.
2345 2011-07-24 16:45:31 <gmaxwell> Though you'll end up being pretty slow in some cases.
2346 2011-07-24 16:45:37 <senseles> bluematt: the server is an x3440 w/ 8 gig of ram that never gets used. so i dont really care how many resources it consumes.
2347 2011-07-24 16:45:51 <BlueMatt> well you can bump the fd limit
2348 2011-07-24 16:45:57 <gmaxwell> senseles: "-hub" should be renamed "-dosattackthenetwork"
2349 2011-07-24 16:45:59 <BlueMatt> but no patch will let you get around that
2350 2011-07-24 16:46:05 _Silverpike_ has joined
2351 2011-07-24 16:46:05 <senseles> oh :/
2352 2011-07-24 16:46:12 <gmaxwell> yea, thats just a ulimit.
2353 2011-07-24 16:46:21 <senseles> does it generate a ton of connections?
2354 2011-07-24 16:46:32 <BlueMatt> to be fair, it does have legitimate usage, but its really a hack to get around current network segmentation
2355 2011-07-24 16:46:42 <BlueMatt> but thats only a real problem for a pool server
2356 2011-07-24 16:47:10 <gmaxwell> pool servers would be far better served by peering with each other.
2357 2011-07-24 16:47:17 <BlueMatt> yep
2358 2011-07-24 16:47:28 <BlueMatt> its a hack, there are better hacks to solve the problem though
2359 2011-07-24 16:48:45 <iddo> did ClearCoin use the infrastructure described in https://en.bitcoin.it/wiki/Contracts ?
2360 2011-07-24 16:48:53 <sipa> no
2361 2011-07-24 16:49:01 <iddo> why not?
2362 2011-07-24 16:49:09 <sipa> because it isn't implemented
2363 2011-07-24 16:49:16 <sipa> or at least, wasn't
2364 2011-07-24 16:49:18 Zagitta has joined
2365 2011-07-24 16:49:22 <iddo> hmm
2366 2011-07-24 16:49:28 <iddo> is it implemented now?
2367 2011-07-24 16:49:43 <sipa> not in the main client
2368 2011-07-24 16:49:44 Silverpike has quit (Ping timeout: 252 seconds)
2369 2011-07-24 16:50:01 <iddo> i don't understand example 1 in the wiki, who has the private key for the output of Tx1 ?
2370 2011-07-24 16:50:40 <sipa> Tx1's output needs two signatures
2371 2011-07-24 16:50:45 <sipa> one from both parties
2372 2011-07-24 16:50:54 <sipa> so the answer to your question is: both
2373 2011-07-24 16:51:16 <iddo> so if the 2 signatures are provied, to whom the 10 BTC sent to?
2374 2011-07-24 16:51:22 TheZagitta has quit (Ping timeout: 260 seconds)
2375 2011-07-24 16:51:27 <iddo> s/provied/provided
2376 2011-07-24 16:52:00 <sipa> to wherever the transaction that claims that output sends it
2377 2011-07-24 16:52:15 bitcoinbulletin has quit (Remote host closed the connection)
2378 2011-07-24 16:52:39 <iddo> hmm
2379 2011-07-24 16:53:00 <gmaxwell> iddo: the txn must be signed by both.
2380 2011-07-24 16:53:25 <iddo> so in step 6, when signing tx1, what would be the output ?
2381 2011-07-24 16:53:57 <gmaxwell> The documentation here may be helpful to you: https://github.com/groffer/bitcoin/commit/dc2dfbab6a0f75070fc3b962da4eb2967e9659df
2382 2011-07-24 16:54:20 bitcoinbulletin has joined
2383 2011-07-24 16:54:32 Zarutian has joined
2384 2011-07-24 16:55:31 <iddo> the wiki seems more elaborate, unless i could actually understand the source code
2385 2011-07-24 16:56:09 T_X has joined
2386 2011-07-24 16:57:03 <iddo> i fail to understand what Tx1 specifies exactly when it broadcast in step 6 of example1 in wiki
2387 2011-07-24 16:58:01 <iddo> and seems like clearcoin would need some combination of example 1 (for future time) and example 2
2388 2011-07-24 16:58:10 <gmaxwell> oh you're asking about the deposit one.
2389 2011-07-24 16:58:28 <iddo> yes deposit one is example 1, it's hard for me to understand:(
2390 2011-07-24 16:59:10 <gmaxwell> tx1 is a regular escrow transaction.
2391 2011-07-24 16:59:36 <iddo> so who has the private key for the output ?
2392 2011-07-24 16:59:43 koleg has quit (Ping timeout: 263 seconds)
2393 2011-07-24 17:00:02 <gmaxwell> There isn't just one private key.
2394 2011-07-24 17:00:22 <gmaxwell> "into an output which requires both user and website to sign"
2395 2011-07-24 17:00:35 <gmaxwell> it's an escrow transaction. I needs signatures by two keys.
2396 2011-07-24 17:00:40 <gmaxwell> s/I/It/
2397 2011-07-24 17:00:58 <iddo> so if the signatures are provided, what happens?
2398 2011-07-24 17:01:21 <gmaxwell> The world ends.
2399 2011-07-24 17:01:34 <gmaxwell> I'm not sure what answer you want.
2400 2011-07-24 17:02:02 <iddo> the 10 BTC become controlled by someone else if 2 signatures are provided?
2401 2011-07-24 17:02:27 <gmaxwell> There is a txn. It needs two signatures in order to release its output. You create a new txn that does something with that input, sign it with both. Then that txn is valid and does whatever it does.
2402 2011-07-24 17:03:32 <gmaxwell> maybe it forwards the coins to 1bitcoineater .. maybe it just sends one to mtgox and the rest back into a txn like the original one.
2403 2011-07-24 17:03:48 <iddo> so tx2 specifies what to do with tx1 ? tx1 by itself just locks your coins ?
2404 2011-07-24 17:05:56 <sipa> the only thing a txn *ever* does is consume coins, create coins, and state what can be done with them
2405 2011-07-24 17:06:36 <sipa> tx1 here creates a coin which cannot be used unless both parties agree on what to do with it
2406 2011-07-24 17:07:18 <gmaxwell> iddo: it's like any other txn.
2407 2011-07-24 17:07:45 <iddo> "create a coin" here means taken from the balance of the user?
2408 2011-07-24 17:07:56 <gmaxwell> ...
2409 2011-07-24 17:08:00 <gmaxwell> Yea, I was just about to say:
2410 2011-07-24 17:08:02 <sipa> iddo: i call transaction outputs "coins"
2411 2011-07-24 17:08:14 <gmaxwell> You sound like you're working with a broken mental model where there are balances.
2412 2011-07-24 17:08:24 <sipa> balances don't exist, only coins
2413 2011-07-24 17:08:30 <gmaxwell> There aren't. Bitcoin tracks transactions (coins), not balances. Balances are a fiction of your wallet.
2414 2011-07-24 17:08:48 <sipa> balances are just the sum of the coins you have the (only necessary) key for
2415 2011-07-24 17:09:49 <iddo> "create a coin" in tx1 means locking a coin that was under your control, so now it's no longer under your control?
2416 2011-07-24 17:10:28 <sipa> tx1 consumes a coin you had the key for, and creates a coin that needs signatures from two keys
2417 2011-07-24 17:10:47 <sipa> so that new coin is indeed no longer (fully) under your control
2418 2011-07-24 17:10:54 <iddo> ok great... now i understand it better
2419 2011-07-24 17:12:21 <gmaxwell> I wonder how the balance output in the client should reflect escrowed coins.. surely there should be some indication of your partial control.
2420 2011-07-24 17:12:23 erus` has quit (Remote host closed the connection)
2421 2011-07-24 17:13:27 <sipa> i don't think you should try to map them to the ledger concept bitcoin tries to mimic in its user interface
2422 2011-07-24 17:13:35 <gmaxwell> otherwise we'll get fun stuff like "HELP: I checked out an older wallet, and it had a balance of zero so I destroyed it. But now I found out that it was one of two required keys for a 1000 btc trust! How do I recover the trust with just one key?!"
2423 2011-07-24 17:15:03 Juslex has joined
2424 2011-07-24 17:17:19 <jrmithdobbs> is requiring n sigs really the best way to do this?
2425 2011-07-24 17:17:43 <jrmithdobbs> i've been playing with shamir's secret sharing algo
2426 2011-07-24 17:18:09 <jrmithdobbs> wouldn't a more robust way to implement be to use something similar?
2427 2011-07-24 17:18:30 <jrmithdobbs> that way you could have an escrow with 5 parties involved but only 2 required for sign off
2428 2011-07-24 17:18:34 <jrmithdobbs> eg
2429 2011-07-24 17:18:43 koleg has joined
2430 2011-07-24 17:18:47 <sipa> well, secret sharing solves the problem by making the real key hidden cryptographically from the users
2431 2011-07-24 17:18:49 <gmaxwell> N sigs doesn't require any party to ever have seen all the sigs.
2432 2011-07-24 17:19:01 <jrmithdobbs> oh good point
2433 2011-07-24 17:19:01 <gmaxwell> s/sigs/keys/
2434 2011-07-24 17:19:18 <sipa> multisigs relies on the convention that the world will only accept the transaction if enough signatures are present
2435 2011-07-24 17:19:25 <jrmithdobbs> someone would have had to have seen the shared secret at one point to create the shares
2436 2011-07-24 17:19:53 <jrmithdobbs> is there a way to blind that?
2437 2011-07-24 17:20:03 <sipa> i suppose it's possible
2438 2011-07-24 17:20:28 <gmaxwell> jrmithdobbs: yes, kinda. but it's unblinded on send.
2439 2011-07-24 17:20:30 osmosis has joined
2440 2011-07-24 17:20:42 <sipa> BlueMatt, jrmithdobbs, gmaxwell: what about just enforcing a limit of "only keep the N most recent addresses" ?
2441 2011-07-24 17:20:49 <gmaxwell> (go look at how ECDH works)
2442 2011-07-24 17:20:57 BlueMatt has quit (Remote host closed the connection)
2443 2011-07-24 17:21:13 <gmaxwell> sipa: I think there should be two queues, nodes you've connected to and nodes you haven't. N most on each.
2444 2011-07-24 17:21:27 <gmaxwell> sipa: otherwise some attacker can randomly flush your addnodes.
2445 2011-07-24 17:21:32 <sipa> good idea
2446 2011-07-24 17:21:47 <gmaxwell> (in fact, more than two might make sense for this reason, but two should be the strict minimum)
2447 2011-07-24 17:22:03 <jrmithdobbs> gmaxwell: but on send it doesn't matter any more since the data becomes worthless since it's been used
2448 2011-07-24 17:22:11 koleg has quit (Client Quit)
2449 2011-07-24 17:22:13 MrTiggr has quit (Ping timeout: 240 seconds)
2450 2011-07-24 17:22:24 BlueMatt has joined
2451 2011-07-24 17:22:37 <jrmithdobbs> gmaxwell: assuming it could be blinded, this would also add plausible deniability to *who* actually signed off
2452 2011-07-24 17:22:43 <gmaxwell> e.g. one queue per 'network' you've heard the addresses from. (/16 v4, /32 v6 .. or something smarter in the future) ::shrugs::
2453 2011-07-24 17:22:57 MrTiggr has joined
2454 2011-07-24 17:23:05 <gmaxwell> oh I don't know how to do the blinding for partials. ... only for all required.
2455 2011-07-24 17:23:15 <gmaxwell> Better to just hide who owns the keys.
2456 2011-07-24 17:23:19 <sipa> gmaxwell: in that case, each queue should be quite small then
2457 2011-07-24 17:23:33 <gmaxwell> There is probably a protocol you can do external to bitcoin to blind who wholds which keys.
2458 2011-07-24 17:23:37 <jrmithdobbs> gmaxwell: the blinding doesn't need to be done on partials, the sharing for re-combining could be done out of band
2459 2011-07-24 17:23:52 <jrmithdobbs> gmaxwell: so spending requires the shared secret, not the shares
2460 2011-07-24 17:24:07 <gmaxwell> oh well, you can just hash lock a txn.
2461 2011-07-24 17:24:16 <jrmithdobbs> right it simplifies the whole thing
2462 2011-07-24 17:24:30 <BlueMatt> sipa: well then you can get ddosed into dropping all valid addresses
2463 2011-07-24 17:24:38 <BlueMatt> sipa: have to keep some old ones...
2464 2011-07-24 17:24:48 <jrmithdobbs> no longer requires IsStandard changes, etc
2465 2011-07-24 17:25:02 <sipa> BlueMatt: gmaxwell just suggested using (at least) two separate queues, one for nodes you've connected to yourself, one for others
2466 2011-07-24 17:25:13 <jrmithdobbs> wait no, for that you'd still have to have an un-blinded person create the txn
2467 2011-07-24 17:25:17 <jrmithdobbs> hmmm
2468 2011-07-24 17:25:19 <BlueMatt> yea, that sounds cool
2469 2011-07-24 17:25:22 <sipa> agree
2470 2011-07-24 17:25:25 <BlueMatt> maybe limit the addresses someone can send to something like 50 instead of 1000 as well?
2471 2011-07-24 17:25:58 <BlueMatt> not amount sent, but actually drop a connection (or I guess ignore everything after) you get the first 50 or 100 addresses
2472 2011-07-24 17:26:03 <gmaxwell> sipa: perhaps one list for nodes you've talked to and then N seperate queuses where N=[8-16] based on a hash of the neighbor network addresses... so you wouldn't have thousands so they could be pretty big.
2473 2011-07-24 17:26:38 <gmaxwell> BlueMatt: before that nodes need to be smarter about which ones they give.
2474 2011-07-24 17:26:57 <sipa> i wouldn't *drop* nodes sending you too many
2475 2011-07-24 17:27:02 <sipa> just ignore them
2476 2011-07-24 17:27:06 <sipa> at least for now
2477 2011-07-24 17:27:08 <BlueMatt> that too, but you arent ever going to use >100 addresses
2478 2011-07-24 17:27:17 <iddo> is it still possible to modify this contracts infrastructure ? or it's hard because all the miners already follow the protocol as it's already described?
2479 2011-07-24 17:27:20 <BlueMatt> well you can drop ones that send >1k, as regular nodes shouldnt
2480 2011-07-24 17:27:27 <gmaxwell> sure you will.. when the first 100 were dialup users who have gone offline.
2481 2011-07-24 17:27:30 <senseles> just make the client calculate which nodes are the best most regular
2482 2011-07-24 17:27:32 <sipa> iddo: contracts simply use the script functionality
2483 2011-07-24 17:27:36 <sipa> iddo: there is nothing hardcoded
2484 2011-07-24 17:27:43 <BlueMatt> gmaxwell: no, you use dnsseed before that
2485 2011-07-24 17:27:56 <senseles> use like the top 50 and then drop anyone who tries to spam more than 50 addresses to you
2486 2011-07-24 17:28:39 <senseles> probably would be best to advertise whatever nodes the user has been connected to the longest
2487 2011-07-24 17:28:45 <senseles> some sort of connection time counter for each
2488 2011-07-24 17:28:53 NickelBot has quit (Read error: Connection reset by peer)
2489 2011-07-24 17:28:53 knotwork has quit (Read error: Connection reset by peer)
2490 2011-07-24 17:29:46 <jrmithdobbs> BlueMatt: hey i have a proposal for the wallet crypto stuff, very minor change
2491 2011-07-24 17:29:53 <iddo> but the SIGHASH flags cannot be extended ?
2492 2011-07-24 17:29:58 <BlueMatt> jrmithdobbs: shoot
2493 2011-07-24 17:30:02 <gmaxwell> BlueMatt: the network should not depend on DNSSeed.
2494 2011-07-24 17:30:32 <gmaxwell> We've been able to answer "omg IRC = centraliztion" with 'it's only for bootstrapping', that shouldn't change
2495 2011-07-24 17:30:39 <BlueMatt> well, ok so dont drop after 50 unless you have a good selection of addr.dat stuff already
2496 2011-07-24 17:30:39 <sipa> agree
2497 2011-07-24 17:30:45 <gmaxwell> there is no harm in learning and remembering a couple thousand addresses...
2498 2011-07-24 17:30:57 <jrmithdobbs> BlueMatt: can we add an nNumFactors and change vchCryptedKey and vchSalt to be vectors of vectors of chars?
2499 2011-07-24 17:30:59 <gmaxwell> sounds okay too.
2500 2011-07-24 17:31:21 <BlueMatt> jrmithdobbs: why?
2501 2011-07-24 17:31:32 Marf has quit (Read error: Connection reset by peer)
2502 2011-07-24 17:31:37 NickelBot has joined
2503 2011-07-24 17:31:41 <jrmithdobbs> BlueMatt: to allow n-factor auth of the master key
2504 2011-07-24 17:31:43 <jrmithdobbs> s/of/for/
2505 2011-07-24 17:31:49 <BlueMatt> jrmithdobbs: you can already do that
2506 2011-07-24 17:32:01 <BlueMatt> thats what the additionstuff vch is for
2507 2011-07-24 17:32:08 <BlueMatt> I mean yea, your thing is probably a bit cleaner
2508 2011-07-24 17:32:21 <BlueMatt> but its not worth redoing it now
2509 2011-07-24 17:32:27 <BlueMatt> it works fine currently
2510 2011-07-24 17:32:31 <jrmithdobbs> ya storage inside those without being multi-dimensional is kind of tedious
2511 2011-07-24 17:32:38 <jrmithdobbs> is why i was asking
2512 2011-07-24 17:32:44 <BlueMatt> its not that hard
2513 2011-07-24 17:33:15 <sipa> gmaxwell: so, 1024 last addresses you've connected to yourself in the past, and 16 queues of the 1024 last addresses you've heard about based on ...
2514 2011-07-24 17:33:34 <sipa> hash(groupid(address-of-node-you-got-it-from)) ?
2515 2011-07-24 17:34:16 Marf has joined
2516 2011-07-24 17:34:22 <jrmithdobbs> sipa / gmaxwell: that sounds awesome. then turning on ipv6 isn't nearly as dangerous
2517 2011-07-24 17:34:27 <BlueMatt> gmaxwell: my point was not that you shouldnt accept a ton, but that 1k is overkill for each new connection, plus it severely hampers fill-your-addr.dat attacks
2518 2011-07-24 17:34:52 <BlueMatt> yea, storing 10k, no problem, soring more, sure, but there is a point where it gets to be overkill
2519 2011-07-24 17:35:22 <gmaxwell> sipa: when a node falls off the last 1024 it should get moved into a random group. (or perhaps its duplicated there already)
2520 2011-07-24 17:36:13 <gmaxwell> sipa: the node should also _not_ prefer the 1024-known-good group except perhaps for the first connection or two.
2521 2011-07-24 17:36:32 <BlueMatt> no, not the first, only if its failing
2522 2011-07-24 17:36:33 <jrmithdobbs> BlueMatt: i guess you're right, i'll revist when i have time to patchwork the code for the specific use-case i have in mind ;p
2523 2011-07-24 17:36:36 <sipa> also: 17*1024 addresses * (16 byte IP + 2 byte port + 8 byte services + 4 byte time = 30 bytes) = 510 KiB
2524 2011-07-24 17:36:57 <jrmithdobbs> (*patchwork into bitcoin i mean)
2525 2011-07-24 17:37:01 <Juslex> Hello. My company does offer ICT services that I'd like to sell to people who uses Bitcoin. Right now we use Freshbooks for billing our customers. Is there possibility to develop some kind of Bitcoin API for Freshbooks (http://developers.freshbooks.com/)? That would be awesome feature, I think.
2526 2011-07-24 17:37:32 <gmaxwell> the whole current preference logic makes me feel uncomfortable... but I figure it'll get fixed when we add node rotation.
2527 2011-07-24 17:38:27 MobiusL has quit (Remote host closed the connection)
2528 2011-07-24 17:38:29 <jrmithdobbs> BlueMatt: was just thinking that since it's not in the wild yet it would be the time to bring it up to try and prevent storage-format incompat in the future
2529 2011-07-24 17:38:30 <BlueMatt> wait, we actually have preference logic right now?
2530 2011-07-24 17:38:31 <jrmithdobbs> BlueMatt: ;p
2531 2011-07-24 17:38:39 <gmaxwell> BlueMatt: kinda!
2532 2011-07-24 17:38:40 <jrmithdobbs> BlueMatt: yes, very awful preference logic
2533 2011-07-24 17:39:00 <jrmithdobbs> (not so awful if irc is disabled)
2534 2011-07-24 17:39:01 <BlueMatt> jrmithdobbs: na, its really not hard at all to just shove ints or whatever you want in a vch, its just a blind char string
2535 2011-07-24 17:39:23 <gmaxwell> Also, the groupid filtering gives me hives. I can prevent you from connecting to node X by connecting to you from address Y in the same /16 ... which could create some interesting attacks.
2536 2011-07-24 17:39:31 <BlueMatt> gmaxwell: the whole nLastSeen stuff doesnt really count...
2537 2011-07-24 17:39:54 <jrmithdobbs> BlueMatt: ya but some of what i'm looking at can have variable length output and no good separator so would require encoding to store like that, it is just an ease thing though, you're right
2538 2011-07-24 17:40:10 <BlueMatt> gmaxwell: should probably be examined more, I mean something like 50% of internet traffic is on just a handful of /8s
2539 2011-07-24 17:40:12 MobiusL has joined
2540 2011-07-24 17:40:12 <gmaxwell> (get a proxy in every /16 that random_pool uses, then connect to everyone via thoses IPs... really agressively, thus depriving the pool of inbounds, and you fill slots so it can't get outbounds)
2541 2011-07-24 17:40:14 <BlueMatt> or client traffic I should say
2542 2011-07-24 17:40:28 briareus has quit (Ping timeout: 240 seconds)
2543 2011-07-24 17:40:31 <gmaxwell> Yea. It's a pretty weak metric.
2544 2011-07-24 17:40:55 <gmaxwell> I'd rather it just be applied to outbounds.. e.g. don't consider what inbounds you have.
2545 2011-07-24 17:41:14 <BlueMatt> wait, oh god I forgot about that
2546 2011-07-24 17:41:20 <BlueMatt> yea, it absolutely should just be applied to outbounds
2547 2011-07-24 17:41:31 <gmaxwell> I've been running a patch that does that.
2548 2011-07-24 17:41:38 <BlueMatt> for sybil, and other similar attack-scenarios it should be assumed that all inbound connections are evil
2549 2011-07-24 17:42:01 <gmaxwell> If we had better group-ids than /16s... then we could randomly pick a group then randomly pick a node. right now that would be dumb due to the imbalance of allocations.
2550 2011-07-24 17:42:20 <gmaxwell> but yea, not letting inbounds influnce that would be good.
2551 2011-07-24 17:42:25 <sipa> how would you improve the group ids?
2552 2011-07-24 17:42:49 <gmaxwell> sipa: by shipping some kind of perfect hash that maps addresses into ASNs perhaps. Not sure. The challenge is making it small and simple, not making one.
2553 2011-07-24 17:42:52 <BlueMatt> you could do rdns stuff, but that would only be a supplement, and isnt very reliable
2554 2011-07-24 17:43:23 <gmaxwell> IT's pretty easy to snapshot the internet routing table and make a static map...
2555 2011-07-24 17:43:43 <BlueMatt> shouldnt depend on a possibly-out-of-date map though...
2556 2011-07-24 17:44:36 <BlueMatt> also, with ipv6 rollout speeding up, it becomes out of date quicker
2557 2011-07-24 17:45:14 <sipa> gmaxwell: ASN?
2558 2011-07-24 17:45:27 <BlueMatt> atonymous system number
2559 2011-07-24 17:45:48 <BlueMatt> essentially the isp
2560 2011-07-24 17:45:51 <gmaxwell> BlueMatt: a map from 10 years ago wouldn't be worse than using /16s. :)
2561 2011-07-24 17:45:53 thefinn93 has joined
2562 2011-07-24 17:46:04 <BlueMatt> heh, thats probably true
2563 2011-07-24 17:46:18 <gmaxwell> v6 is a point though... but in that case /32s are actually not that terrible.
2564 2011-07-24 17:47:23 MrTiggr has quit (Ping timeout: 276 seconds)
2565 2011-07-24 17:47:33 <BlueMatt> whats min iana allocation?
2566 2011-07-24 17:47:36 <BlueMatt> a /32 iirc
2567 2011-07-24 17:47:56 <gmaxwell> Yes.
2568 2011-07-24 17:48:09 <BlueMatt> sorry, not iana, ripe or other
2569 2011-07-24 17:48:16 koleg has joined
2570 2011-07-24 17:48:26 <BlueMatt> arin is what I was looking for
2571 2011-07-24 17:48:29 <gmaxwell> well there are some special /48s but they don't give them out easily.
2572 2011-07-24 17:48:48 TheZagitta has joined
2573 2011-07-24 17:48:53 <BlueMatt> yea, Im not concerned about "really hard to get" allocations
2574 2011-07-24 17:49:03 Zagitta has quit (Ping timeout: 258 seconds)
2575 2011-07-24 17:49:05 <gmaxwell>  /16 is pretty bad.. in some spots one /16 will cover a dozen providers... and at the same time MIT has 256 /16s. :)
2576 2011-07-24 17:49:16 <gmaxwell> (about v4, obviously)
2577 2011-07-24 17:49:29 <BlueMatt> yea v4 is just a pain
2578 2011-07-24 17:49:49 rynx has joined
2579 2011-07-24 17:49:52 <jrmithdobbs> gmaxwell: the problem with /32s is all of he.net comes out of /32 and since something like 80% of ipv6 traffic goes to/from that ...
2580 2011-07-24 17:49:54 <BlueMatt> though yea, smarter group stuff based on allocation tables would probably work pretty well
2581 2011-07-24 17:49:56 <sipa> gmaxwell: what about adding let's say a 64-bit flag to each address, each bit N of which corresponds to "heard this address recently from someone whose group id hashes to N"
2582 2011-07-24 17:50:14 <sipa> and then prioritize by number of bits set in that flag
2583 2011-07-24 17:50:33 <jrmithdobbs> gmaxwell: but looked at the math yesterday and how they allocate tunnels makes it hard to reduce the size limit for their specific /32 because you can easily end up with 2 different /33 assignments on the same account
2584 2011-07-24 17:50:48 <TuxBlackEdo> hey so the pools give work to the miners, would the miners be able to check if the work meets minimum difficulty and simply not send that result to the pool, but rather a own private solo pool?
2585 2011-07-24 17:51:07 <gmaxwell> sipa: I like that, but it requires that peers tend to flood a lot of addresses to keep it current.
2586 2011-07-24 17:51:08 <sipa> TuxBlackEdo: since they need to decide who the reward goes to before starting to mine, no
2587 2011-07-24 17:51:18 <TuxBlackEdo> ah
2588 2011-07-24 17:51:31 <TuxBlackEdo> that makes complete sense
2589 2011-07-24 17:51:45 <sipa> they could decide the withold the found hash
2590 2011-07-24 17:51:49 <TuxBlackEdo> it just hashes the receiving address with sha256
2591 2011-07-24 17:51:51 <TuxBlackEdo> right
2592 2011-07-24 17:51:55 <TuxBlackEdo> but that wouldnt help anyone
2593 2011-07-24 17:52:02 <sipa> indeed, that's just vandalism
2594 2011-07-24 17:52:26 <jrmithdobbs> anyone have access to some i5/i7 hardware and/or xeon 3500/5500 hardware that can run some timing tests for me?
2595 2011-07-24 17:52:34 <gmaxwell> TuxBlackEdo: "withholding attack" ... killer against PPS pools, but otherwise not interesting.
2596 2011-07-24 17:52:56 <b4epoche_> jrmithdobbs:  normally but not today
2597 2011-07-24 17:53:11 <TuxBlackEdo> jrmithdobbs: i got an i7
2598 2011-07-24 17:53:12 <BlueMatt> jrmithdobbs: i7 920 ocd here?
2599 2011-07-24 17:53:32 <sipa> gmaxwell: hmm, you could make that timeout a day or so
2600 2011-07-24 17:53:35 <BlueMatt> i7 laptop too, but thats another class entirely
2601 2011-07-24 17:53:39 <jrmithdobbs> BlueMatt: cool should just take a second of your time, give me a sec to write up build instructions
2602 2011-07-24 17:53:50 <jrmithdobbs> well, hopefully about 30-60 seconds ;p
2603 2011-07-24 17:54:20 <BlueMatt> doing timing tests for a new crypto key derivation algo?
2604 2011-07-24 17:54:27 <jrmithdobbs> not new, no
2605 2011-07-24 17:54:37 <BlueMatt> well, scrypt or something?
2606 2011-07-24 17:54:38 <sipa> scrypt?
2607 2011-07-24 17:54:41 <jrmithdobbs> ya scrypt
2608 2011-07-24 17:54:49 <jrmithdobbs> but not using it as a hash
2609 2011-07-24 17:55:08 <diki> worst thread ever -> http://forum.bitcoin.org/index.php?topic=31153.0
2610 2011-07-24 17:55:27 <diki> these guys havent even the slightest idea of how to create a db structure
2611 2011-07-24 17:55:51 <diki> they dont even know what sizes the fields to be...
2612 2011-07-24 17:55:54 <diki> why did they even bother
2613 2011-07-24 17:56:34 <diki> it's just plain and simply wrong....
2614 2011-07-24 17:57:30 koleg has quit (Quit: KVIrc 4.0.4 Insomnia http://www.kvirc.net/)
2615 2011-07-24 17:57:35 <sipa> diki: then tell that to the people in that thread
2616 2011-07-24 17:58:04 <jrmithdobbs> BlueMatt: http://home.jrbobdobbs.org/mith/backup-spec/shamirs-poc.sh
2617 2011-07-24 17:58:45 <jrmithdobbs> BlueMatt: could you build and run that through a few times, I just need an idea of real time for running on newer archs than core2duo ;p (script spits it out)
2618 2011-07-24 17:59:17 RAM2012 has joined
2619 2011-07-24 17:59:35 <jrmithdobbs> BlueMatt: also if you don't have a good entropy source for /dev/random that'll make it capable of spitting out at least 128bytes changing that to urandom is fine for this test ;p
2620 2011-07-24 17:59:37 <BlueMatt> arg, scrypt isnt even in package manager...
2621 2011-07-24 17:59:55 <jrmithdobbs> BlueMatt: doesn't matter you need to patch it to make this work anyways
2622 2011-07-24 17:59:59 <BlueMatt> yea, I need to get a hardware random thing
2623 2011-07-24 18:00:14 makomk has quit (Ping timeout: 250 seconds)
2624 2011-07-24 18:00:27 <jrmithdobbs> BlueMatt: because the upstream source forces passphrase input from the tty
2625 2011-07-24 18:00:35 <jrmithdobbs> hate shit that does that
2626 2011-07-24 18:01:02 <BlueMatt> does it make sense to anyone that the spec for the hardware crypto stuff that everyone uses uses hash algos as its random generator instead of a dumb antenna and a->d converter?
2627 2011-07-24 18:01:31 <BlueMatt> youd think it would be cheaper to just get an a->d converter and only use the LSB than a hasher in hardware
2628 2011-07-24 18:01:38 <jrmithdobbs> BlueMatt: i've been thinking about getting one of those entropy key things to play with
2629 2011-07-24 18:01:52 <jrmithdobbs> but the gf might kill me if i buy any more crypto toys right now, ha
2630 2011-07-24 18:01:56 br00t has joined
2631 2011-07-24 18:01:56 <jrmithdobbs> (just bought 6 yubikeys)
2632 2011-07-24 18:02:03 <gmaxwell> The entropy keys make me sad.
2633 2011-07-24 18:02:08 <BlueMatt> why the hell do you need 6?
2634 2011-07-24 18:02:14 <jrmithdobbs> project i'm working on
2635 2011-07-24 18:02:26 <gmaxwell> IIRC they have a design flaw. They run the FIPS test on the data, and then discard blocks that fail.
2636 2011-07-24 18:02:29 <gmaxwell> :(
2637 2011-07-24 18:02:37 <gmaxwell> makes me wonder what else they did wrong.
2638 2011-07-24 18:02:53 <jrmithdobbs> got a source on that?
2639 2011-07-24 18:03:31 <jrmithdobbs> seems like discarding the data would actually make the output LESS random instead of keeping some kind of cache of failures and then marking itself bad after a certain threshhold in a timeframe
2640 2011-07-24 18:04:44 <gmaxwell> oh maybe I'm confusing its behavior for rngd.
2641 2011-07-24 18:04:51 <gmaxwell> jgarzik: you wrote that didn't you?
2642 2011-07-24 18:05:08 <BlueMatt> jrmithdobbs: I have to go, Ill run your test when I get back if you dont mind
2643 2011-07-24 18:05:29 <sipa> gmaxwell: how often is that test expected to fail on truly random data?
2644 2011-07-24 18:06:16 knotwork has joined
2645 2011-07-24 18:06:27 <jrmithdobbs> BlueMatt: sure man, when you got time
2646 2011-07-24 18:06:34 <gmaxwell> I thought it was pretty common. like 1:1000.
2647 2011-07-24 18:07:24 <gmaxwell> jrmithdobbs: one possible better thing to do is to collect 40k bits (instead of the 20k fips tests run on) then test half. if it passes, xor the halves and output
2648 2011-07-24 18:07:28 SerajewelKS has quit (Ping timeout: 240 seconds)
2649 2011-07-24 18:07:46 <gmaxwell> if it fails save the data, and try another one. just keep xoring more in until you get a pass.
2650 2011-07-24 18:07:55 <jrmithdobbs> gmaxwell: that'd cut the data rate down to 16Kb/s from 32Kb/s
2651 2011-07-24 18:07:57 <jrmithdobbs> though
2652 2011-07-24 18:08:12 <jrmithdobbs> it's a tradeoff i guess
2653 2011-07-24 18:08:17 <lfm> VIA CPUs have a different hw rng
2654 2011-07-24 18:08:49 <gmaxwell> In any case the later nist docs don't recommend that procedure anymore.
2655 2011-07-24 18:09:13 <jrmithdobbs> gmaxwell: honestly, i need to do some better randomness testing, but haveged makes /dev/random on my atom d525 under not-much load have ~1MB/2s of /dev/random available
2656 2011-07-24 18:09:43 <jrmithdobbs> http://haveged.sourceforge.net/
2657 2011-07-24 18:10:01 <jrmithdobbs> it's, unfortunately, gplv3 versions of the original havege code :(
2658 2011-07-24 18:10:12 rynx has quit (Quit: Leaving)
2659 2011-07-24 18:10:18 <phantomcircuit> lol microsoft ripped off havege
2660 2011-07-24 18:10:21 <gmaxwell> yea, .. too bad the linux kernel rnd is basically broken.
2661 2011-07-24 18:10:34 <phantomcircuit> their RNG is 100% based on the havege concept
2662 2011-07-24 18:10:48 <lfm> huh?
2663 2011-07-24 18:10:52 <gmaxwell> the pool is only 4kbit, and since most drivers don't add entropy anymore anything that uses it constantly runs out
2664 2011-07-24 18:10:54 <phantomcircuit> audio-entropyd
2665 2011-07-24 18:11:11 <phantomcircuit> works pretty well as long as you have static
2666 2011-07-24 18:11:17 <gmaxwell> phantomcircuit: the haveged implementation (in terms of feeding the pool) is much better than anything else.
2667 2011-07-24 18:11:28 <phantomcircuit> gmaxwell, yes i said they stole the concept
2668 2011-07-24 18:11:32 <lfm> you can add entropy from a user level pgm too
2669 2011-07-24 18:11:40 <jrmithdobbs> gmaxwell: ya, haveged makes /dev/random usable again
2670 2011-07-24 18:11:48 <gmaxwell> Not the measurement, but the fact that it dribbles it in.
2671 2011-07-24 18:11:50 <jrmithdobbs> i was pleasantly surprised
2672 2011-07-24 18:12:25 <gmaxwell> Other tools write 100k of data at a time then sleep for 60 seconds... meanwhile the pool is 4kbits.. it fills it and the pool is drained by second two of that sleep.
2673 2011-07-24 18:12:26 <lfm> gmaxwell: thats not broken, its just slow without harware assist
2674 2011-07-24 18:12:56 <jrmithdobbs> gmaxwell: like I said, i need to do better randomness testing, but i'd almost be comfortable using /dev/urandom for strong-crypto stuff on a machine using haveged
2675 2011-07-24 18:13:13 <gmaxwell> lfm: It's broken: I wasted youdon'tfuckingwanttoknow how many hours trying to figure out why ssh had intermittent one second hangs on connect to a busy webserver.
2676 2011-07-24 18:13:24 <gmaxwell> ... only to discover the machine was running out of entropy.
2677 2011-07-24 18:13:27 <lfm> /dev/urandom is good enuf for almost anything we do.
2678 2011-07-24 18:13:40 <jrmithdobbs> lfm: not without entropy sources it's not
2679 2011-07-24 18:13:43 <gmaxwell> Because _only_ the timer added any. So it was _constantly_ dry.
2680 2011-07-24 18:13:47 <jrmithdobbs> lfm: ESPECIALLY on tasks that run at boot
2681 2011-07-24 18:13:57 <gmaxwell> well, bitcoin is fine, I agree...
2682 2011-07-24 18:13:59 <jrmithdobbs> like, auto-generating ssh host keys
2683 2011-07-24 18:14:06 <gmaxwell> we don't need enough data for it to matter, much.
2684 2011-07-24 18:14:21 <lfm> gmaxyou just need some more entropy sources to feed into it. you cant just make them up.
2685 2011-07-24 18:14:33 TheZimm has quit (Read error: Connection reset by peer)
2686 2011-07-24 18:14:41 <gmaxwell> lfm: Linux _used_ to get entropy from disks and the nic.
2687 2011-07-24 18:14:50 <gmaxwell> But most drivers don't contribute anymore.
2688 2011-07-24 18:14:55 <jrmithdobbs> lfm: you missed his comments about modern drivers not feeding the entropy properly
2689 2011-07-24 18:15:00 <jrmithdobbs> ya that
2690 2011-07-24 18:15:02 <Eliel> gmaxwell: doesn't that count as a bug almost?
2691 2011-07-24 18:15:09 <lfm> gmaxmaybe cuz they were not good sources
2692 2011-07-24 18:15:10 <gmaxwell> It was intentional.
2693 2011-07-24 18:15:19 TheZimm has joined
2694 2011-07-24 18:15:20 <gmaxwell> Combined with the very small pool (4kbit) it's a real problem.
2695 2011-07-24 18:15:26 <jrmithdobbs> whois gmax|canary anyways, breaking tab complete
2696 2011-07-24 18:15:27 <jrmithdobbs> heh
2697 2011-07-24 18:15:44 <lfm> gmaxwell: simple enuf to get some user level pgms that add entropy
2698 2011-07-24 18:15:48 <jrmithdobbs> gmaxwell: if that's your client change it's name to canary|gmax ;p
2699 2011-07-24 18:15:51 <gmaxwell> jrmithdobbs: dunno why it breaks ... it's later in order and hasn't talkd lately.
2700 2011-07-24 18:16:24 <gmaxwell> lfm: yes— using even worse entropy sources than the nic/io layer.
2701 2011-07-24 18:16:25 <jrmithdobbs> gmaxwell: doesn't always, at least not for me, but when you've been inactive for a bit and that client drops/rejoins it starts breaking again ;p
2702 2011-07-24 18:16:47 <jrmithdobbs> lfm: hence the mention of haveged I made
2703 2011-07-24 18:17:08 <jrmithdobbs> it makes /dev/random as useful, if not moreso, as it used to be
2704 2011-07-24 18:17:15 <gmaxwell> yea, haveged is the only good one— the other userlevel programs don't work right.
2705 2011-07-24 18:17:30 <gmaxwell> (as mentioned, they dump tons of data at a time, which does you no good)
2706 2011-07-24 18:18:08 <lfm> gmaxwell: depends on the pgm. some use pretty good sources from the real world
2707 2011-07-24 18:19:27 <gmaxwell> in any case, I'd bitch less if it were possible to increase the pool size.
2708 2011-07-24 18:19:34 <gmaxwell> making it a whole page would be a vast improvement.
2709 2011-07-24 18:19:41 <gmaxwell> it used to be adjustable.
2710 2011-07-24 18:19:56 <gmaxwell> But someone found an exploit in the adjustment api. (0_o) so it was just removed.
2711 2011-07-24 18:20:13 <lfm> gmaxwell: a larger pool wont change the rate of entropy gathering
2712 2011-07-24 18:20:31 asuk has joined
2713 2011-07-24 18:21:06 <Eliel> lfm: yes, but if the average need is small enough, it prevents it running dry even if the need is very bursty.
2714 2011-07-24 18:21:20 <gmaxwell> lfm: Yes. But it would reduce the number of times the device runs dry and blocks due to bursts... and it would make userspace apps more effective because they wouldn't have to run many times a second to keep it full.
2715 2011-07-24 18:21:36 <jrmithdobbs> gmaxwell: don't you think, even with the FIPS check (which their documentation does say it does just as you described) that so long as the entropykey isn't the *only* etropy source (eg, combined with haveged) it's still useful?
2716 2011-07-24 18:21:50 <gmaxwell> oh sure.. it's useful.
2717 2011-07-24 18:22:00 <jrmithdobbs> it could be better
2718 2011-07-24 18:22:14 <gmaxwell> looking at the docs I think it actually does the right things with the FIPS check.
2719 2011-07-24 18:22:19 <jrmithdobbs> but considering the next lowest cost hw generator i can find is quite literally 10x the cost ...
2720 2011-07-24 18:22:20 <gmaxwell> it's rngd that doesn't.
2721 2011-07-24 18:22:58 <jrmithdobbs> but ya, haveged alone has made all my weird kerberos/ssh hangs magically vanish
2722 2011-07-24 18:23:01 <jrmithdobbs> ;p
2723 2011-07-24 18:23:09 <gmaxwell> oh you were seeing weird hangs too?
2724 2011-07-24 18:23:11 <lfm> I think the main problem is people useing /dev/random when /dec/yrandom is all they need.
2725 2011-07-24 18:23:25 <lfm> yrand -> urand
2726 2011-07-24 18:23:26 imsaguy has joined
2727 2011-07-24 18:23:26 imsaguy has quit (Changing host)
2728 2011-07-24 18:23:26 imsaguy has joined
2729 2011-07-24 18:23:35 <jrmithdobbs> gmaxwell: ya, i never investigated and just noticed they stopped once i put haveged on everything
2730 2011-07-24 18:23:42 <gmaxwell> You'd think that "running out of entropy" should be BOFH excuse calender item....
2731 2011-07-24 18:23:49 makomk has joined
2732 2011-07-24 18:23:51 <jrmithdobbs> lol
2733 2011-07-24 18:24:03 <gmaxwell> lfm: yea great, good luck auditing everything and changing it out.
2734 2011-07-24 18:24:36 <lfm> gmaxwell: did you look at ssh to see if it could be cahnged?
2735 2011-07-24 18:24:52 <gmaxwell> It didn't have a configuration option.
2736 2011-07-24 18:25:05 TheZimm has quit (Read error: Connection reset by peer)
2737 2011-07-24 18:25:12 <iddo> can the contracts infrastructure be used where in 1st step two users lock same amount of their own BTC, so that only both of them can release the combined amount, without involving trust?
2738 2011-07-24 18:25:15 <gmaxwell> Plus I didn't know what _else_ might be using /dev/random.
2739 2011-07-24 18:25:19 <jrmithdobbs> lfm: i disagree, i think more things should be using /dev/random instead of /dev/urandom
2740 2011-07-24 18:25:23 <phantomcircuit> gmaxwell, you can set ssh to use urandom instead of random
2741 2011-07-24 18:25:24 <jrmithdobbs> but i'm paranoid
2742 2011-07-24 18:25:32 thefinn93 has quit (Ping timeout: 276 seconds)
2743 2011-07-24 18:25:40 TheZimm has joined
2744 2011-07-24 18:25:43 <jrmithdobbs> lfm: s/things/more things/
2745 2011-07-24 18:25:45 <sipa> things should also not read more from /dev/random than necessary
2746 2011-07-24 18:25:49 koleg has joined
2747 2011-07-24 18:25:56 molecular has joined
2748 2011-07-24 18:26:16 <sipa> if you're generating a key with a security level of 256 bits, don't read more than 32 bytes of /dev/random data
2749 2011-07-24 18:26:24 <gmaxwell> sipa: for example, urandom users may drain you to zero.. then you need /dev/random for getnewaddress and you block for a while waiting for the pool to refill.
2750 2011-07-24 18:26:37 <TuxBlackEdo> is it totally safe to switch out wallet.dat files?
2751 2011-07-24 18:26:57 <jrmithdobbs> TuxBlackEdo: fairly but it fucks up the usefulness of the bdb txn logs
2752 2011-07-24 18:26:59 <sipa> does anyone know how much random data openssl actually reads from /dev/random when generating a key?
2753 2011-07-24 18:27:19 danbri has quit (Read error: Connection timed out)
2754 2011-07-24 18:27:23 <phantomcircuit> sipa, 256 bits
2755 2011-07-24 18:27:35 <jrmithdobbs> sipa: think it's partially configurable depending on what you setup in the rng code before generating
2756 2011-07-24 18:27:48 <jrmithdobbs> i'd have to look
2757 2011-07-24 18:27:54 <gmaxwell> I wonder if I can block it forever by having an agressive thing that reads 64 bits at a time from /dev/urandom. :)
2758 2011-07-24 18:28:00 danbri has joined
2759 2011-07-24 18:28:03 <BlueMatt> jrmithdobbs: hey, ok what files do I have to modify here?
2760 2011-07-24 18:28:17 <jrmithdobbs> BlueMatt: just main.c in scrypt as described in the comments
2761 2011-07-24 18:28:27 <BlueMatt> sorry, that wasnt clear ;)
2762 2011-07-24 18:28:33 <BlueMatt> or Im blind
2763 2011-07-24 18:28:34 <BlueMatt> I dont know
2764 2011-07-24 18:28:39 Joric has joined
2765 2011-07-24 18:28:43 <sipa> how many fingers am i showing?
2766 2011-07-24 18:28:44 <cjdelisle> wait isn't /dev/urandom the one that is just a prng and doesn't block?
2767 2011-07-24 18:28:45 <jrmithdobbs> BlueMatt: line 117/118 change the readpass call's last args to NULL, NULL, 0
2768 2011-07-24 18:28:57 <jrmithdobbs> cjdelisle: yes
2769 2011-07-24 18:29:42 <cjdelisle> IMO the "quality" of your random doesn't really matter as long as you don't make a really big *cough* debian mistake
2770 2011-07-24 18:29:57 <phantomcircuit> cjdelisle, it *does* block
2771 2011-07-24 18:30:02 <phantomcircuit> but usually not for long
2772 2011-07-24 18:30:14 <jrmithdobbs> phantomcircuit: hey didn't you just get some awesome desktop with an i7?
2773 2011-07-24 18:30:33 <phantomcircuit> nah
2774 2011-07-24 18:30:36 <jrmithdobbs> i5?
2775 2011-07-24 18:30:40 <phantomcircuit> i just have an i3-2100
2776 2011-07-24 18:30:42 <TuxBlackEdo> amd seperon 140 over here
2777 2011-07-24 18:30:45 <phantomcircuit> it's fast enough though
2778 2011-07-24 18:30:48 sgornick has quit (Quit: Leaving.)
2779 2011-07-24 18:30:56 <jrmithdobbs> oh ok
2780 2011-07-24 18:31:05 <phantomcircuit> i miss being able to buy 1U racks with dual xeons on ebay
2781 2011-07-24 18:31:06 <phantomcircuit> :(
2782 2011-07-24 18:31:07 <jrmithdobbs> phantomcircuit: i need timing data on i5/i7 you are useless ;p
2783 2011-07-24 18:31:15 <cosurgi> tcatm: I like when bitomat - my favourite exchange is on third place. But now with PLN sorted alphabetically I must scroll all the way down there. And I hate it :)
2784 2011-07-24 18:31:16 <jrmithdobbs> <3
2785 2011-07-24 18:31:18 sgornick has joined
2786 2011-07-24 18:31:35 <TuxBlackEdo> i got an i7 720qm
2787 2011-07-24 18:31:41 <cosurgi> tcatm: btw - bitomat disappeared totally.
2788 2011-07-24 18:31:53 <jrmithdobbs> TuxBlackEdo: running a posix OS?
2789 2011-07-24 18:32:04 <BlueMatt> jrmithdobbs: scrypt isng using 100% cpu, Im assuming something is broken? or is that some kind of scrypt thing?
2790 2011-07-24 18:32:10 <BlueMatt> jrmithdobbs: in any case real:44.923s
2791 2011-07-24 18:32:11 <tcatm> cosurgi: yes, I'm currently syncing some exchanges because the API format changed
2792 2011-07-24 18:32:14 <BlueMatt> encode
2793 2011-07-24 18:32:19 <TuxBlackEdo> jrmithdobbs: windoze
2794 2011-07-24 18:32:22 <BlueMatt> jrmithdobbs: decode 7.281s real
2795 2011-07-24 18:32:35 <jrmithdobbs> awesome, similar numbers to the atom d525
2796 2011-07-24 18:32:43 <phantomcircuit> tcatm, also i propose bitmarket be put in a different category as it isn't *really* an exchange
2797 2011-07-24 18:33:01 <cosurgi> tcatm: this sorting could be nice if it was optional. Like a clickable column title "currency" or such (or a stupid checkbox), and when clicked it would sort by USD,PLN.GBP etc...
2798 2011-07-24 18:33:20 <jrmithdobbs> BlueMatt: that's weird that encode took 44 seconds, could you do like 10 sequential runs and just throw it up on pastebin? ;p
2799 2011-07-24 18:33:24 <BlueMatt> jrmithdobbs: what is it depending on? no way is an i7 getting similar to an atom, unless that atom has some insane ram speds
2800 2011-07-24 18:33:46 <jrmithdobbs> BlueMatt: does a bunch of weird memory expansion stuff
2801 2011-07-24 18:33:51 <tcatm> cosurgi: thanks for feedback. I'll consider that :)
2802 2011-07-24 18:33:57 <jrmithdobbs> BlueMatt: and my atom is using ddr3 so comparable
2803 2011-07-24 18:34:07 <tcatm> phantomcircuit: but it's a market, right?
2804 2011-07-24 18:34:14 <BlueMatt> jrmithdobbs: what speed ddr3?
2805 2011-07-24 18:34:22 <jrmithdobbs> i forget tbqh
2806 2011-07-24 18:34:30 <BlueMatt> dmidecode
2807 2011-07-24 18:34:42 <phantomcircuit> tcatm, it's more like bitcoin-otc than anything else you have listed
2808 2011-07-24 18:35:01 <phantomcircuit> each trade is actually a trade between completely different people
2809 2011-07-24 18:35:27 <tcatm> phantomcircuit: it's the same on mtgox/britcoin/...
2810 2011-07-24 18:35:32 <jrmithdobbs> BlueMatt: 1066Mhz
2811 2011-07-24 18:35:52 <phantomcircuit> tcatm, bitmarket doesn't hold funds afaik
2812 2011-07-24 18:35:57 <phantomcircuit> maybe im wrong about that?
2813 2011-07-24 18:35:58 <BlueMatt> well then yea, I mean mine is only 1600, it has nice when I got it...need to upgrade that now
2814 2011-07-24 18:36:06 <phantomcircuit> no im not
2815 2011-07-24 18:36:09 <cjdelisle> The diff with bitmarket is when a trade is executed, there's no guarantee you won't get ripped off.
2816 2011-07-24 18:36:15 <tcatm> phantomcircuit: it holds bitcoins. the other half of the trade is done directly
2817 2011-07-24 18:37:11 <cjdelisle> btw I assume you know this: Interesting ports on static-ip-188-138-101-169.inaddr.intergenia.de (188.138.101.169): 27007/tcp closed flexlm7
2818 2011-07-24 18:37:32 <phantomcircuit> ok so they're an escrow service for one side of the transaction with a web of trust built in
2819 2011-07-24 18:37:45 <BlueMatt> jrmithdobbs: http://pastebin.com/raw.php?i=8Kskb5BA
2820 2011-07-24 18:38:07 <tcatm> cjdelisle: yes, the streaming daemon is currently disabled to prevent flooding in -market
2821 2011-07-24 18:38:28 <cjdelisle> ahh
2822 2011-07-24 18:38:29 <phantomcircuit> you're still reliant on them being good arbitrators
2823 2011-07-24 18:38:45 <tcatm> phantomcircuit: if -otc offered reliable trades.json I would list them, too
2824 2011-07-24 18:38:55 <jrmithdobbs> BlueMatt: do you still see 30+ s if you switch to /dev/urandom? you must be running out of entropy because my atom does the encode in similar times to the decode
2825 2011-07-24 18:38:57 <phantomcircuit> fair enough
2826 2011-07-24 18:39:18 <jrmithdobbs> BlueMatt: apt-get install haveged; for better entropy ;p (it's in debian anyways, i assume it's in ubuntu)
2827 2011-07-24 18:39:20 <cjdelisle> I have a bot running too and I managed to prevent flooding (by trial and kline)  any idea when it should be back?
2828 2011-07-24 18:39:22 <BlueMatt> jrmithdobbs: no thats urandom, I have a felling if its ram limited then yea, I mean 1066 and 1600 are pretty damn close
2829 2011-07-24 18:39:30 Firefly777 has joined
2830 2011-07-24 18:39:48 * BlueMatt needs to get faster ram
2831 2011-07-24 18:39:50 <jrmithdobbs> BlueMatt: that's crazy, like i said the atom is doing encodes at 3-9s just like the decodes
2832 2011-07-24 18:39:52 dr_win has joined
2833 2011-07-24 18:40:06 <BlueMatt> oh 9s encodes
2834 2011-07-24 18:40:09 <BlueMatt> mmm dont know then
2835 2011-07-24 18:40:10 <TuxBlackEdo> is there any way to download blkindex.dat and blk0001.dat from one of you?
2836 2011-07-24 18:40:19 <BlueMatt> TuxBlackEdo: bitcoin.bluematt.me
2837 2011-07-24 18:40:34 <TuxBlackEdo> thanks
2838 2011-07-24 18:40:35 <jrmithdobbs> either way that's good news, scrypt is pretty awesome ;p
2839 2011-07-24 18:40:37 <TuxBlackEdo> :)
2840 2011-07-24 18:40:58 <BlueMatt> jrmithdobbs: no thats messed up, an i7 taking 4x the time of an atom? thats not scrypt, thats me
2841 2011-07-24 18:41:07 <BlueMatt> so...invalid test
2842 2011-07-24 18:41:20 <jgarzik> gmaxwell: yes, wrote rngd
2843 2011-07-24 18:41:20 <cjdelisle> I wrote a scrypt clone in java
2844 2011-07-24 18:41:32 <jrmithdobbs> BlueMatt: ya but the decode times being similar indicate that even if something wasn't screwed up (I blame ubuntu) they should be similar
2845 2011-07-24 18:41:34 <BlueMatt> jrmithdobbs: maybe yours is doing crypto offloading onto processor?
2846 2011-07-24 18:41:45 <BlueMatt> true
2847 2011-07-24 18:41:52 <sipa> interesting: before the pubkey is used on the network, the transaction has 160 bit security level - once it is revealed, only 128 bit left
2848 2011-07-24 18:41:57 <lfm> get a VIA c7 or Nano, their hw rng is super fast
2849 2011-07-24 18:42:16 <jrmithdobbs> BlueMatt: i see similar numbers on core2duo as well
2850 2011-07-24 18:42:27 <jrmithdobbs> BlueMatt: and scrypt isn't offloadable, that's the point
2851 2011-07-24 18:42:31 <BlueMatt> jrmithdobbs: ah, well then yea, it would be expected
2852 2011-07-24 18:44:07 <jrmithdobbs> BlueMatt: have someone getting me numbers on i5 should shed some light
2853 2011-07-24 18:45:12 Juslex has quit (Ping timeout: 252 seconds)
2854 2011-07-24 18:45:18 erus` has joined
2855 2011-07-24 18:45:51 <jgarzik> wow, somebody just added a lot of hashing power
2856 2011-07-24 18:46:47 <sipa> ;;bc,stats
2857 2011-07-24 18:46:49 <gribble> Current Blocks: 137843 | Current Difficulty: 1690906.2047244 | Next Difficulty At Block: 139103 | Next Difficulty In: 1260 blocks | Next Difficulty In About: 1 week, 1 day, 6 hours, 27 minutes, and 0 seconds | Next Difficulty Estimate: 1791189.57139948
2858 2011-07-24 18:46:50 <sipa> ;;bc,nethash
2859 2011-07-24 18:46:56 <jrmithdobbs> BlueMatt: btw, you sure you changed both instances of /dev/random to /dev/urandom
2860 2011-07-24 18:46:56 <gribble> 13054.810905858203
2861 2011-07-24 18:48:30 * TheZagitta lost contact to his remote box while messing with secondary lan card *damnit*
2862 2011-07-24 18:49:06 <gmaxwell> sipa: hm? thats not obvious to me. Why does it lose it
2863 2011-07-24 18:49:23 TheZagitta is now known as Zagitta
2864 2011-07-24 18:49:43 <jrmithdobbs> fucked up route
2865 2011-07-24 18:49:43 <Tril> sipa: your graphs are scaling wrong for me
2866 2011-07-24 18:49:50 <sipa> gmaxwell: EC has a security level of n/2 bits
2867 2011-07-24 18:50:07 <sipa> gmaxwell: if you know the pubkey, it only takes 2^128 steps to find the private key
2868 2011-07-24 18:50:13 <sipa> "only"
2869 2011-07-24 18:50:19 <jrmithdobbs> heh, was about to say
2870 2011-07-24 18:50:26 <gmaxwell> dur. okay.
2871 2011-07-24 18:50:37 <jgarzik> sipa: IPv6 support... nice!
2872 2011-07-24 18:50:50 <sipa> :)
2873 2011-07-24 18:50:50 <gmaxwell> Yea, it's moot because 2^128 operations aren't doable on purely physical grounds.
2874 2011-07-24 18:50:51 <jgarzik> sipa: does it handle the common case of "configured, but not working" IPv6?
2875 2011-07-24 18:51:05 <jgarzik> sipa: usually you must detect routing errors and such upon connect
2876 2011-07-24 18:51:16 <sipa> jgarzik: if a routable ipv4 address is available, it will use that
2877 2011-07-24 18:51:16 <Dagger2> that sounds like a job for the system administrator to me
2878 2011-07-24 18:51:23 <sipa> jgarzik: as local address
2879 2011-07-24 18:51:40 Juslex has joined
2880 2011-07-24 18:51:49 <gmaxwell> 6to4 behind nat...
2881 2011-07-24 18:51:52 <sipa> so in that case, nobody will even know you support ipv6, and if you see an ipv6 address yourself the connect will fail/timeout
2882 2011-07-24 18:52:13 <BlueMatt> jrmithdobbs: oh random is there twice?
2883 2011-07-24 18:52:17 <BlueMatt> no I only changed one instance
2884 2011-07-24 18:52:18 RAM2012 has quit (Ping timeout: 260 seconds)
2885 2011-07-24 18:52:24 <jrmithdobbs> BlueMatt: ya, once in the dd and once in the ./shamirs call
2886 2011-07-24 18:52:30 <gmaxwell> e.g. you get a 6to4 address... but it doesn't actually work because there is some moron firewall in the v4 path.
2887 2011-07-24 18:52:32 <jrmithdobbs> (the first one)
2888 2011-07-24 18:52:45 <jgarzik> sipa: "if a routable..."  sounds like it will never use IPv6 then?  most will be dual stack for a long time.
2889 2011-07-24 18:53:23 <jrmithdobbs> BlueMatt: once i got the guy using i5 (ubuntu with 1033Mhz ddr3) i'm seeing similar numbers to the atom
2890 2011-07-24 18:53:29 <sipa> jgarzik: it will "use" ipv6 in the sense that its listening socket will be v6, and it will attempt ipv6 connections to other nodes
2891 2011-07-24 18:53:36 <BlueMatt> jrmithdobbs: ah, well that sure sped things up, 4.484 for decode and 4.772 encode
2892 2011-07-24 18:53:56 <jrmithdobbs> BlueMatt: ya it varies between 3-9s from testing i've done, cool
2893 2011-07-24 18:54:02 <BlueMatt> jrmithdobbs: well then it proceeds to get slower as it goes on
2894 2011-07-24 18:54:06 <BlueMatt> max 7s atm
2895 2011-07-24 18:54:39 <jgarzik> sipa: even on dual stack, with working ipv4 _and_ working ipv6?
2896 2011-07-24 18:54:56 br00t has quit (Quit: Leaving)
2897 2011-07-24 18:55:25 <sipa> jgarzik: if you have a dual stack with fully reachable ipv4 and ipv6 addresses, it's clearly preferably to advertize using your ipv4 address
2898 2011-07-24 18:55:45 <sipa> well, maybe both should be advertized
2899 2011-07-24 18:56:48 Incitatus has quit (Ping timeout: 240 seconds)
2900 2011-07-24 18:56:59 wardearia has quit (Ping timeout: 252 seconds)
2901 2011-07-24 18:57:39 <sipa> maybe we need a vector of addrLocalHosts
2902 2011-07-24 18:57:43 osmosis has quit (Ping timeout: 240 seconds)
2903 2011-07-24 18:58:26 <gmaxwell> sipa: advertise both please.
2904 2011-07-24 18:58:36 <jrmithdobbs> ya and it should be setable which ones to use via commandline, my public node has a routable ipv4 and 4 ipv6 addresses
2905 2011-07-24 18:58:41 <Tril> multihoming on v4 and v6 is important
2906 2011-07-24 18:59:02 <gmaxwell> Don't want to have v4/v6 network splits.
2907 2011-07-24 18:59:10 <jrmithdobbs> good point
2908 2011-07-24 18:59:23 <sipa> i don't think that's an issue, as v6 clients can connect to v4 ones
2909 2011-07-24 18:59:37 <sipa> but yes, advertizing all local addresses is probably the way to go
2910 2011-07-24 18:59:39 Joric has quit ()
2911 2011-07-24 18:59:43 <gmaxwell> if they have working v4. But you might not: for exaple, onioncat nodes.
2912 2011-07-24 18:59:56 <jrmithdobbs> sipa: so long as the ones advertised can be restricted
2913 2011-07-24 19:00:06 <jrmithdobbs> by user
2914 2011-07-24 19:00:07 <gmaxwell> though I suppose onioncat nodes should be handled specially in any case.
2915 2011-07-24 19:01:06 <gmaxwell> (onioncat maps the entire tor hidden service address space (80bits) into IPv6, I've been assuming that how bitcoin would eventually have proper hidden service support.)
2916 2011-07-24 19:02:00 <sipa> IPv4 is now effectively handled as a mapped address space
2917 2011-07-24 19:02:11 <sipa> shouldn't be a problem to add another mapped range
2918 2011-07-24 19:04:37 RAM2012 has joined
2919 2011-07-24 19:06:07 <iddo> can the contracts infrastructure be used where in 1st step two users lock same amount of their own BTC, so that only both of them can release the combined amount, without involving trust?
2920 2011-07-24 19:06:31 <gmaxwell> iddo: sure.
2921 2011-07-24 19:06:42 <iddo> how...?
2922 2011-07-24 19:06:44 <gmaxwell> iddo: A and B send 1 BTC each to {a,b}
2923 2011-07-24 19:07:06 <iddo> what is {a,b} ?
2924 2011-07-24 19:07:22 <gmaxwell> iddo: multisig with both of their keys.
2925 2011-07-24 19:08:09 <iddo> do you dont even need these contract scripts?
2926 2011-07-24 19:08:17 <iddo> how to create this multisig ?
2927 2011-07-24 19:08:35 <iddo> s/do/so
2928 2011-07-24 19:08:36 <gmaxwell> You need a regular escrow transaction. Via the patch that I sent you that you declined to read the docs on?
2929 2011-07-24 19:08:36 <sipa> iddo: there is no "contracts infrastructure"
2930 2011-07-24 19:08:49 <sipa> contracts (as satoshi called them) are just an interesting application
2931 2011-07-24 19:09:19 <sipa> each transaction output has a script that determines which input is allowed to spend it
2932 2011-07-24 19:09:21 <iddo> ok i'm not sure how to call the contract concept that's on the wiki
2933 2011-07-24 19:09:31 <sipa> even regular transactions as they are used now
2934 2011-07-24 19:09:35 Incitatus has joined
2935 2011-07-24 19:09:37 <gmaxwell> thats doing far more than you need.
2936 2011-07-24 19:09:55 <TuxBlackEdo> thanks bluematt
2937 2011-07-24 19:09:55 <gmaxwell> For what you're describing a simple multisig escrow transaction works.
2938 2011-07-24 19:10:46 <iddo> gmaxwell: multisig is explained on https://github.com/groffer/bitcoin/blob/dc2dfbab6a0f75070fc3b962da4eb2967e9659df/doc/README_escrow.txt  ?
2939 2011-07-24 19:10:48 <gmaxwell> A,B pay {a,b} (two keys required) each. Then a and b each are only willing to sign a txn that spends both of those inputs, to whatever output they've agreed on.
2940 2011-07-24 19:10:51 <TuxBlackEdo> why would it say my bitcoin address is 1BabB?
2941 2011-07-24 19:11:03 <TuxBlackEdo> hmm
2942 2011-07-24 19:11:38 <TuxBlackEdo> in my receiving addresses it shows my address i always use, but in "your bitcoin address" it shows a different one
2943 2011-07-24 19:11:40 <gmaxwell> TuxBlackEdo: you imported a key?
2944 2011-07-24 19:11:54 <iddo> gmaxwell: when i said not involving trust, i meant that A cannot destroy B's coin without also destroying his own coin
2945 2011-07-24 19:12:00 RazielZ has quit (Quit: Leaving)
2946 2011-07-24 19:12:15 <TuxBlackEdo> i just imported the 400mb block chain
2947 2011-07-24 19:13:00 <gmaxwell> iddo: ah, no, the only way I can think to do that is to add a third party.
2948 2011-07-24 19:13:16 <TuxBlackEdo> In my address book it says under receiving address: "1ME91.." which is the address i am familiar with, but under "Your bitcoin address" it shows a different one
2949 2011-07-24 19:13:31 wardearia has joined
2950 2011-07-24 19:13:45 <iddo> gmaxwell: what's the scenario with third party?
2951 2011-07-24 19:14:40 <gmaxwell> you sign to A,B,C ...and require 2 of 3. C is a disinterested third party to breaks the tie if a or b try to screw the other.
2952 2011-07-24 19:15:11 osmosis has joined
2953 2011-07-24 19:15:21 <iddo> both A and B need to broadcast something, no?
2954 2011-07-24 19:16:43 <gmaxwell> yes, a send secured to any 2 of a,b,c.
2955 2011-07-24 19:17:27 <jrmithdobbs> i wonder if i could trivially modify haveged for os x
2956 2011-07-24 19:17:35 <jrmithdobbs> i have a feeling no
2957 2011-07-24 19:17:52 <iddo> and B sends the same of his?
2958 2011-07-24 19:18:09 <iddo> and C can return them their coins?
2959 2011-07-24 19:18:44 huk has joined
2960 2011-07-24 19:18:47 <b4epoche_> can someone explain why CS people use entropy when they mean random?
2961 2011-07-24 19:19:15 <gmaxwell> iddo: yes. C can cooperate with A or B if the other doesn't.
2962 2011-07-24 19:19:37 <Tril> TuxBlackEdo: click the address and press OK in receiving addresses, it will change Your address. but Your Address will update again to a new one if you receive a payment to it.
2963 2011-07-24 19:20:18 amiller has quit (Remote host closed the connection)
2964 2011-07-24 19:20:18 devrandom has quit (Write error: Broken pipe)
2965 2011-07-24 19:20:21 random_cat has quit (Remote host closed the connection)
2966 2011-07-24 19:21:48 <TuxBlackEdo> Tril: you are right that fixed the problem? Any idea on where it got this address from? I was running rescan because I imported bluematt's blockchain, and an address that I have never seen before popped up as my receiving address, but it wasn't listed in my "address book" anywhere
2967 2011-07-24 19:21:57 <TuxBlackEdo> s/?/.
2968 2011-07-24 19:22:04 <Tril> sipa: horizontal scale on every graph is the same (back to 1/1/09) at http://bitcoin.sipa.be/
2969 2011-07-24 19:22:24 Burgundy has joined
2970 2011-07-24 19:22:55 <Tril> TuxBlackEdo: bitcoin generates new receiving addresses for various reasons, it's normal
2971 2011-07-24 19:23:08 random_cat has joined
2972 2011-07-24 19:23:18 <Tril> TuxBlackEdo: they should all be in address book
2973 2011-07-24 19:24:35 <iddo> when you have 5 coins and send 2, the other 3 go into address that doesn't appear in your address book, right?
2974 2011-07-24 19:27:05 devrandom has joined
2975 2011-07-24 19:27:08 <Eliel> b4epoche_: entropy is more specific term I guess.
2976 2011-07-24 19:27:46 <Eliel> there are often more than one way to interpret the word random
2977 2011-07-24 19:27:53 <jrmithdobbs> b4epoche_: because it's too easy to confused prng output with
2978 2011-07-24 19:27:56 <jrmithdobbs> b4epoche_: basically
2979 2011-07-24 19:28:17 <b4epoche_> jrmithdobbs ??
2980 2011-07-24 19:28:33 <jrmithdobbs> b4epoche_: re: entropy vs random
2981 2011-07-24 19:29:06 <b4epoche_> so entropy is used for pseudo-random?
2982 2011-07-24 19:29:14 <jrmithdobbs> b4epoche_: mostly when you're talking about "random" numbers in cs/crypto related discussion you mean the output of (hopefully) a crypto-strong prng
2983 2011-07-24 19:29:41 <iddo> one way to look at entropy is how many random bit flips are needed
2984 2011-07-24 19:29:43 <jrmithdobbs> b4epoche_: whereas entropy is used to describe the (hopefully) "real" random numbers used to seed said prng
2985 2011-07-24 19:30:27 <b4epoche_> iddo:  I understand what entropy is…  it's just that I don't see how the definitition applies to how people in CS use it
2986 2011-07-24 19:30:56 <sipa> b4epoche_: how do you mean?
2987 2011-07-24 19:31:04 <sipa> b4epoche_: for example, if i have a variable that is 0 50% of the time, and 1 50% of the time, it has 1 bit of entropy
2988 2011-07-24 19:31:25 <iddo> i guess it's just more elaborate on the exact randomness behavior
2989 2011-07-24 19:31:47 <sipa> if it's 10% to be 0, and 90% to be one, it has 0.468 bit of entropy
2990 2011-07-24 19:31:53 Daviey has quit (Excess Flood)
2991 2011-07-24 19:32:28 <sipa> because a file with all such variable (independently) could be compressed to use one 0.468 bit per file bit
2992 2011-07-24 19:32:31 <iddo> meaning you need to do 0.468 coin flips to have this kind of random behavior :)
2993 2011-07-24 19:32:53 <sipa> sorry, 0.468 file bits per variable
2994 2011-07-24 19:32:57 <b4epoche_> but you guys don't use it as the number of different ways a system can meet imposed constraints?
2995 2011-07-24 19:33:18 <iddo> hmm
2996 2011-07-24 19:33:21 <sipa> if it's uniform, that's correct
2997 2011-07-24 19:33:45 <sipa> if i have a variable that can take 16 different values, and each is as likely as the rest, it has an entropy of 4 bits
2998 2011-07-24 19:34:07 * b4epoche_ understands it from physics (statistical mechanics) and is trying to see the analogy in CS/IS
2999 2011-07-24 19:34:22 <TuxBlackEdo> Tril: you know whats weird, the Your Receiving Bitcoin Address was not in the address book, then i selected it and no mention of that address was ever given again
3000 2011-07-24 19:34:29 <sipa> b4epoche_: well for one thing it's exactly the same formula
3001 2011-07-24 19:34:35 random_cat is now known as base37
3002 2011-07-24 19:34:41 <sipa> sum(P(x)*log(P(x))
3003 2011-07-24 19:34:46 larsivi has joined
3004 2011-07-24 19:35:27 <b4epoche_> sipa:  Yea, I know a little about the IS stuff…
3005 2011-07-24 19:35:45 <b4epoche_> but in reading about HAVEGE I'm not seeing the connection
3006 2011-07-24 19:35:54 Stellar has joined
3007 2011-07-24 19:36:13 <sipa> can you paste/link/... some text that uses the word entropy in a way you don't think is correct?
3008 2011-07-24 19:36:30 <sipa> Tril: graphs fixed
3009 2011-07-24 19:36:37 <TuxBlackEdo> so will bitcoin be the new facebook/twitter/google of currency?
3010 2011-07-24 19:36:37 <b4epoche_> e.g.:  http://www.issihosts.com/haveged/
3011 2011-07-24 19:37:06 <TuxBlackEdo> i know i could ask in #bitcoin, but i wanna ask #bitcoin-dev
3012 2011-07-24 19:37:08 <sipa> b4epoche_: right, so they talk about the entropy of the pool?
3013 2011-07-24 19:37:40 <sipa> if you start from a pool of 4096 bits, and they are all unknown and independent, it is said to have 4096 bits of entropy, ok?
3014 2011-07-24 19:38:23 koleg has quit (Read error: Connection reset by peer)
3015 2011-07-24 19:38:49 <sipa> now, i tell you the first 2048 bits of the pool, and perform a deterministic operation to it, to scramble all bits again
3016 2011-07-24 19:39:02 <sipa> but i don't use any truly random data source
3017 2011-07-24 19:39:07 <b4epoche_> Yea, I guess this one makes more sense to me:  http://packages.debian.org/sid/haveged
3018 2011-07-24 19:39:23 <sipa> in that case, you can (theoretically) determine which 2^2048 states my pool is in
3019 2011-07-24 19:39:28 Daviey has joined
3020 2011-07-24 19:39:37 <sipa> so it only has 2048 bits of entropy left
3021 2011-07-24 19:39:53 <Eliel> TuxBlackEdo: I definitely am hoping it is :)
3022 2011-07-24 19:39:56 <sipa> sorry, you can theoretically limit the possible states my pool is in to 2^2048
3023 2011-07-24 19:39:59 <b4epoche_> I guess it's main the acronym that was confusing me since this "harvesting randomness for the system entropy pool" makes sense
3024 2011-07-24 19:40:22 sgornick1 has joined
3025 2011-07-24 19:40:29 <b4epoche_> s/main/mainly
3026 2011-07-24 19:40:33 eastender has joined
3027 2011-07-24 19:41:09 <b4epoche_> specifically:  "Entropy Gathering" seems very strange terminology
3028 2011-07-24 19:41:54 <iddo> they just mean gathering truly random bits?
3029 2011-07-24 19:42:16 <b4epoche_> do you also use "partition function" for something?
3030 2011-07-24 19:43:06 <sipa> b4epoche_: in the sense that it is an operation that increases the number of possible states an attacker can know the entropy pool is in, i guess it's correct but a bit far fetched
3031 2011-07-24 19:43:40 <b4epoche_> I mean, is P in your entropy equation called a partition function like in mechanics?
3032 2011-07-24 19:43:58 <sipa> heh, no, probability
3033 2011-07-24 19:44:25 <iddo> what is the partition operation?
3034 2011-07-24 19:44:49 <b4epoche_> oh, sorry, I wasn't referring to the equation you wrote.
3035 2011-07-24 19:46:46 freakazoid has joined
3036 2011-07-24 19:48:37 <b4epoche_> well, I guess I kinda did…  I mean, is there a partition function that helps define the entropy (and probability)?
3037 2011-07-24 19:49:01 spirals has quit (Changing host)
3038 2011-07-24 19:49:01 spirals has joined
3039 2011-07-24 19:49:16 <b4epoche_> I'm guessing not because it seems like you guys don't use the term…  although I suspect maybe you have a different name for it.
3040 2011-07-24 19:49:33 base37 is now known as random_cat
3041 2011-07-24 19:49:34 <b4epoche_> or maybe it just isn't an important quantity in CS
3042 2011-07-24 19:49:45 <sipa> i'm not sure what it is in mechanics
3043 2011-07-24 19:50:14 <iddo> yeah i now looked at http://en.wikipedia.org/wiki/Partition_function_(statistical_mechanics)
3044 2011-07-24 19:50:15 <sipa> b4epoche_: entropy = sum(P(x)*log(P(x)), foreach x in states)/log(0.5) bits
3045 2011-07-24 19:50:25 <iddo> never seen it before..
3046 2011-07-24 19:50:58 shLONG has joined
3047 2011-07-24 19:51:02 shLONG has quit (Client Quit)
3048 2011-07-24 19:51:42 <b4epoche_> actually, I think it's probably not important because you don't have an analogy to energy
3049 2011-07-24 19:52:14 bitcoinbulletin has quit (Remote host closed the connection)
3050 2011-07-24 19:52:19 <sipa> no, indeed
3051 2011-07-24 19:52:24 <b4epoche_> P(x) = 1/Z exp(-\beta E(x))
3052 2011-07-24 19:52:35 <b4epoche_> Z is the partition function
3053 2011-07-24 19:52:41 <b4epoche_> E energy
3054 2011-07-24 19:54:18 <b4epoche_> okay, I think I see…  you always have a uniform distribution of 'energy' (which I think would be analogous to the bits)
3055 2011-07-24 19:54:20 bitcoinbulletin has joined
3056 2011-07-24 19:55:03 pimpinganiteasy has joined
3057 2011-07-24 19:57:55 E-sense has joined
3058 2011-07-24 19:58:08 <jrmithdobbs> oh hi journal paywall, ugh
3059 2011-07-24 19:58:16 <jrmithdobbs> anyone have this article? http://www.sciencenews.org/view/feature/id/5669/title/Take_a_Chance ;p
3060 2011-07-24 20:00:58 citiz3n has joined
3061 2011-07-24 20:04:12 ByronJohnson has quit (Read error: Operation timed out)
3062 2011-07-24 20:04:25 <[Tycho]> "The material you are trying to access is available to subscribers only"
3063 2011-07-24 20:04:36 osmosis has quit (Ping timeout: 246 seconds)
3064 2011-07-24 20:04:41 <jrmithdobbs> hence paywall comment
3065 2011-07-24 20:11:34 <TuxBlackEdo> gmaxwell might
3066 2011-07-24 20:12:26 <jrmithdobbs> doesn't look like it matters, it's in reference to a dead project anyways
3067 2011-07-24 20:13:05 <TuxBlackEdo> isnt it crazy awesome that one our guys got slashdotted the other day? http://science.slashdot.org/story/11/07/22/2254204/Release-of-33GiB-of-Scientific-Publications
3068 2011-07-24 20:15:17 <jrmithdobbs> o hey http://www.barcoding.si.edu/BackgroundPublications/ScienceNews_Dec4th_2004.pdf
3069 2011-07-24 20:15:52 <jrmithdobbs> nm, not a complete copy :(
3070 2011-07-24 20:16:50 ByronJohnson has joined
3071 2011-07-24 20:20:25 Beccara has joined
3072 2011-07-24 20:21:59 <justmoon> jrmithdobbs, http://m.je/4015789.pdf
3073 2011-07-24 20:22:26 <justmoon> jrmithdobbs, let me know when you have downloaded it so I can delete it
3074 2011-07-24 20:22:51 maikmerten has quit (Remote host closed the connection)
3075 2011-07-24 20:23:03 <jrmithdobbs> justmoon: thanks, got it
3076 2011-07-24 20:28:37 <Zagitta> Will NICs automatically go down to 100mbit from 1gbit if the cable can't support it for whatever reason?
3077 2011-07-24 20:29:35 <luke-jr> Zagitta: sometimes
3078 2011-07-24 20:31:04 <Zagitta> luke-jr: ah okay thanks, was just wondering because i can't get my server to run with 1gbit although both NICs supports it and it's a direct cable, it's 15-20m long though which might explain it
3079 2011-07-24 20:31:48 datagutt has quit (Quit: Computer has gone to sleep.)
3080 2011-07-24 20:32:33 spirals has quit (Read error: Operation timed out)
3081 2011-07-24 20:34:02 cenuij has joined
3082 2011-07-24 20:34:03 cenuij has quit (Changing host)
3083 2011-07-24 20:34:03 cenuij has joined
3084 2011-07-24 20:35:08 viggi_ is now known as viggi
3085 2011-07-24 20:35:13 ahbritto has quit (Ping timeout: 240 seconds)
3086 2011-07-24 20:35:17 ahbritto_ has quit (Ping timeout: 250 seconds)
3087 2011-07-24 20:39:27 <upb> Zagitta: try disabling that auto negotiating stuff _ÖP
3088 2011-07-24 20:39:29 <upb> ?:P
3089 2011-07-24 20:40:09 <b4epoche_> jrmithdobbs:  you still need it?
3090 2011-07-24 20:40:28 <jrmithdobbs> b4epoche_: nah found it, and it was anticlimatic
3091 2011-07-24 20:40:36 <WakiMiko_> guys i have a question: once all coins have been generated, can there still be blocks without transactions?
3092 2011-07-24 20:40:41 <jrmithdobbs> b4epoche_: thanks to justmoon ;p
3093 2011-07-24 20:40:43 <b4epoche_> jrmithdobbs:  that's the problem I have with that...
3094 2011-07-24 20:40:44 TheAncientGoat has quit (Ping timeout: 276 seconds)
3095 2011-07-24 20:40:59 <WakiMiko_> that is, will there still be a coinbase tx (just with output 0)?
3096 2011-07-24 20:41:05 <jrmithdobbs> b4epoche_: ended up just being a 3 page article on what rngs are
3097 2011-07-24 20:41:10 <jrmithdobbs> was like, um, awesome
3098 2011-07-24 20:41:12 * b4epoche_ most papers are anti-climatic
3099 2011-07-24 20:41:46 sanity has quit (Remote host closed the connection)
3100 2011-07-24 20:41:47 <Zagitta> upb: huh?
3101 2011-07-24 20:41:57 <b4epoche_> "Boy, I'd be pissed if I had to pay for this!" crosses my mind a lot
3102 2011-07-24 20:42:12 larsivi has quit (Ping timeout: 252 seconds)
3103 2011-07-24 20:43:38 <jrmithdobbs> b4epoche_: heh, from another channel where the topic originally came up: 15:27 <@mith> and ya, glad i didn't fuckin pay for that
3104 2011-07-24 20:45:12 <b4epoche_> I remember frequently being pissed in the old days walking back from the library after finding what looked like an interesting article in an online database
3105 2011-07-24 20:45:31 fnord0 has joined
3106 2011-07-24 20:45:44 sytse has quit (Read error: Connection reset by peer)
3107 2011-07-24 20:45:46 <b4epoche_> signal to noise ratio is so low, I've about given up on 'scholarly articles'
3108 2011-07-24 20:45:50 sytse has joined
3109 2011-07-24 20:48:21 ahbritto has joined
3110 2011-07-24 20:48:39 zeropointo has joined
3111 2011-07-24 20:48:58 ahbritto_ has joined
3112 2011-07-24 20:49:59 nameless has quit (Remote host closed the connection)
3113 2011-07-24 20:51:05 <nanotube> WakiMiko_: yes, blocks will continue to be generated. coinbase will just have fees, no creation bounty.
3114 2011-07-24 20:51:16 soap has quit (Read error: Connection reset by peer)
3115 2011-07-24 20:52:17 Bachfischer has joined
3116 2011-07-24 20:53:26 <WakiMiko_> nanotube: my point is that right now there can be blocks with ONLY the generation transaction e.g. http://blockexplorer.com/b/137854 . is that still possible with all coins generated? will there be a coinbase tx just without an output?
3117 2011-07-24 20:53:34 <WakiMiko_> or will the output be set to 0?
3118 2011-07-24 20:53:52 spirals has joined
3119 2011-07-24 20:53:53 <WakiMiko_> if there is no coinbase tx there is no merkle root
3120 2011-07-24 20:55:45 T_X has left ()
3121 2011-07-24 20:56:18 <mtrlt> the output will be the fees
3122 2011-07-24 20:56:47 <mtrlt> the collected fees.
3123 2011-07-24 20:56:55 <WakiMiko_> fees are input - output of collected transactions
3124 2011-07-24 20:57:10 <WakiMiko_> however im talking about a block WITHOUT collected transactions
3125 2011-07-24 20:57:11 <justmoon> WakiMiko_, correct, if there are no transactions you'd have a coinbase with zero output
3126 2011-07-24 20:57:13 <WakiMiko_> like the one i linked
3127 2011-07-24 20:57:24 <mtrlt> hmpf
3128 2011-07-24 20:57:32 <mtrlt> yea i didn't read carefully enough
3129 2011-07-24 20:58:18 <WakiMiko_> ok, so blocks will always have a coinbase tx, even though it doesnt anything any more
3130 2011-07-24 20:58:53 <justmoon> WakiMiko_, yes a valid block has to have a coinbase tx even if its value is zero
3131 2011-07-24 20:59:55 <WakiMiko_> thats what i thought, thx
3132 2011-07-24 21:02:39 amiller has joined
3133 2011-07-24 21:02:55 danbri has quit (Read error: Connection timed out)
3134 2011-07-24 21:03:33 danbri has joined
3135 2011-07-24 21:04:25 jimon has quit (Excess Flood)
3136 2011-07-24 21:04:34 jimon has joined
3137 2011-07-24 21:07:25 Bachfischer has quit (Quit: Linkinus - http://linkinus.com)
3138 2011-07-24 21:08:39 Stellar has quit (Ping timeout: 246 seconds)
3139 2011-07-24 21:11:28 p0s has quit (Remote host closed the connection)
3140 2011-07-24 21:11:30 <nanotube> WakiMiko_: if there are no transactions, it'll be 0.
3141 2011-07-24 21:11:39 <nanotube> (and if block bounty has reached 0)
3142 2011-07-24 21:11:54 <nanotube> bah i see justmoon already covered it hehe
3143 2011-07-24 21:12:57 <WakiMiko_> <3
3144 2011-07-24 21:14:14 <nanotube> :)
3145 2011-07-24 21:14:20 <luke-jr> so I don't get this math
3146 2011-07-24 21:14:26 <copumpkin> omg maff
3147 2011-07-24 21:14:27 <luke-jr> I bought 50 BTC at $13.85010
3148 2011-07-24 21:14:32 <luke-jr> I sold it at $14.00999
3149 2011-07-24 21:14:36 <luke-jr> how the *%&*# did I lose $7?
3150 2011-07-24 21:14:49 <luke-jr> (no trade fees involved)
3151 2011-07-24 21:15:02 <nanotube> luke-jr: did you withdraw any btc in between? heh
3152 2011-07-24 21:15:17 <luke-jr> nanotube: yes, but I'm accounting for that
3153 2011-07-24 21:15:25 <luke-jr> I'm comparing the difference, not the total
3154 2011-07-24 21:15:48 molecular has quit (Ping timeout: 255 seconds)
3155 2011-07-24 21:16:11 <luke-jr> hmm
3156 2011-07-24 21:16:17 <nanotube> well, post your transaction log (de-personal-datified if you wish) and i can take a look heh
3157 2011-07-24 21:16:22 <luke-jr> it appears I somehow bought 51 BTC, I wonder if that would do it
3158 2011-07-24 21:16:32 * luke-jr also wonders why MtGox would trade 51 when he inputs 50
3159 2011-07-24 21:17:08 <justmoon> I always found mtgox to be a strange and mystical place
3160 2011-07-24 21:17:09 wardearia has quit (Ping timeout: 255 seconds)
3161 2011-07-24 21:17:55 <nanotube> hehe
3162 2011-07-24 21:18:03 <nanotube> are you sure you didn't typo, luke-jr ?
3163 2011-07-24 21:18:10 <nanotube> i mean 1 is right next to 0... hehe
3164 2011-07-24 21:18:13 <luke-jr> pretty sure, but I can't prove it
3165 2011-07-24 21:18:16 <luke-jr> no, 1 is shift-0
3166 2011-07-24 21:18:16 BlueMattBot has quit (Ping timeout: 258 seconds)
3167 2011-07-24 21:18:16 dvide has quit ()
3168 2011-07-24 21:19:13 <luke-jr> http://luke.dashjr.org/education/tonal/keyboard/dvorak-tonal.png
3169 2011-07-24 21:19:19 BlueMatt has quit (Ping timeout: 260 seconds)
3170 2011-07-24 21:20:30 BlueMatt has joined
3171 2011-07-24 21:20:35 <nanotube> luke-jr: haha nice
3172 2011-07-24 21:21:04 <nanotube> so that's how you produce the smiley faces
3173 2011-07-24 21:21:19 <justmoon> so 50 is shift-5, 0 - you probably held shift a split second too long
3174 2011-07-24 21:21:42 <justmoon> HAppens to me all the time
3175 2011-07-24 21:22:03 <nanotube> any reason why you have the card suites on qjkx ? doesn't strike me as a very useful set of symbols to have "easy at hand"
3176 2011-07-24 21:22:15 <nanotube> some with the stuff on gcrl
3177 2011-07-24 21:22:17 <WakiMiko_> what layout is that?
3178 2011-07-24 21:22:21 <justmoon> nanotube, clearly you don't play IRC poker very often :D
3179 2011-07-24 21:22:35 <nanotube> justmoon: haha indeed i do not.
3180 2011-07-24 21:22:39 <WakiMiko_> oh i totally missed the link
3181 2011-07-24 21:22:40 <WakiMiko_> nvm
3182 2011-07-24 21:22:49 Stellar has joined
3183 2011-07-24 21:23:01 <nanotube> WakiMiko_: dvorak-tonal ;) well, at least it's got the dvorak in it :)
3184 2011-07-24 21:23:51 <luke-jr> justmoon: held?
3185 2011-07-24 21:24:12 <luke-jr> nanotube: couldn't think of more useful symbols, and I wanted the heart :p
3186 2011-07-24 21:24:14 <luke-jr> ♥
3187 2011-07-24 21:24:45 <justmoon> luke-jr, does your shift work differently than mine?
3188 2011-07-24 21:24:55 <luke-jr> justmoon: does yours work different from normal?
3189 2011-07-24 21:25:09 <luke-jr> oh, I see what you mean
3190 2011-07-24 21:25:22 <luke-jr> I suppose it's not impossible
3191 2011-07-24 21:25:24 <justmoon> luke-jr, anything that I type while I hold it down is in CAPS
3192 2011-07-24 21:26:02 <justmoon> luke-jr, so if i type a Captital letter and I don't let go of shift quickly enough, the next letter will be a CApital as well
3193 2011-07-24 21:27:16 <justmoon> oh, just saw your last two messages, nvm
3194 2011-07-24 21:31:01 eastender2 has joined
3195 2011-07-24 21:33:01 eastender has quit (Ping timeout: 264 seconds)
3196 2011-07-24 21:34:50 Clipse has quit (Read error: Connection reset by peer)
3197 2011-07-24 21:35:12 spirals_ has joined
3198 2011-07-24 21:35:14 mmoya has joined
3199 2011-07-24 21:35:59 Cusipzzz has joined
3200 2011-07-24 21:36:19 spirals_ has quit (Client Quit)
3201 2011-07-24 21:36:58 BlueMattBot has joined
3202 2011-07-24 21:37:27 denisx has quit (Quit: denisx)
3203 2011-07-24 21:37:44 lolwat` has joined
3204 2011-07-24 21:39:12 spirals has quit (Ping timeout: 255 seconds)
3205 2011-07-24 21:40:53 devon_hillard has quit (Read error: Connection reset by peer)
3206 2011-07-24 21:41:13 <Zagitta> Off topic: I threw 4 scsi drives (2x37gb 2x36gb) in a software 0 raid today for the heck of it... the 2 37gb in raid gives 140 mb/s read and write while all of them together only gives 160mb/s :3 is that *normal* ?
3207 2011-07-24 21:44:25 <cjdelisle> sounds like you're hitting a bottleneck on your motherboard
3208 2011-07-24 21:45:17 <cjdelisle> you have a fast scsi controller?
3209 2011-07-24 21:45:26 Titeuf_87 has quit (Read error: Connection reset by peer)
3210 2011-07-24 21:46:32 sabalaba has joined
3211 2011-07-24 21:48:11 <luke-jr> I was just recently told Linux's softraid is a joke in ##Linux
3212 2011-07-24 21:48:28 <Zagitta> well it's a fujitsu siemens rx100 server
3213 2011-07-24 21:49:25 <Zagitta> and yeah i should probably just run a HW raid
3214 2011-07-24 21:49:46 <cjdelisle> I would be blaming the processor at this point
3215 2011-07-24 21:50:21 <cjdelisle> I find I hit limits around 120MB/s playing with data on the processor even if it's cached in ram
3216 2011-07-24 21:50:22 <Zagitta> might be it as it's just 2x old P4 based xeon
3217 2011-07-24 21:50:40 <phantomcircuit> Zagitta, use lvm stripping
3218 2011-07-24 21:51:21 Stellar has quit (Ping timeout: 246 seconds)
3219 2011-07-24 21:51:45 spirals has joined
3220 2011-07-24 21:52:56 <Zagitta> phantomcircuit: i'll have a look at that
3221 2011-07-24 21:55:45 sabalaba has quit (Quit: Leaving)
3222 2011-07-24 21:57:27 danbri has quit (Read error: Connection timed out)
3223 2011-07-24 21:58:06 danbri has joined
3224 2011-07-24 22:03:57 danbri has quit (Remote host closed the connection)
3225 2011-07-24 22:05:24 soap has joined
3226 2011-07-24 22:05:24 soap has quit (Changing host)
3227 2011-07-24 22:05:24 soap has joined
3228 2011-07-24 22:07:42 Diablo-D3 has joined
3229 2011-07-24 22:10:04 TD has joined
3230 2011-07-24 22:12:30 lolwat` has quit (Ping timeout: 255 seconds)
3231 2011-07-24 22:13:00 <upb> 24 23:37 < Zagitta> upb: huh? <- for the switch port and server nic
3232 2011-07-24 22:13:45 lolwat` has joined
3233 2011-07-24 22:14:23 <Zagitta> upb: there's no switch in between and as far as i can see the servers nic the advertised mode is 1000mbit full duplex
3234 2011-07-24 22:18:58 danbri has joined
3235 2011-07-24 22:23:52 <JFK911> Zagitta: you doing stripe or mirror?
3236 2011-07-24 22:24:18 <JFK911> seems that linux 2.6 wont stripe reads on a mirrored md normally, you have to lay it out as "raid10"
3237 2011-07-24 22:24:34 <JFK911> otoh windoze mirror does it
3238 2011-07-24 22:24:59 <JFK911> the other problem is that different spindles that get simultaneous access should be on different scsi bus
3239 2011-07-24 22:26:02 larsivi has joined
3240 2011-07-24 22:35:19 <gjs278> linux's softraid is not a joke
3241 2011-07-24 22:35:23 <Zagitta> JFK911: i was doing stripe but i just finished up LVM now and i'm testing performance
3242 2011-07-24 22:35:28 <gjs278> the processor would be the joke in this instance
3243 2011-07-24 22:36:26 <gjs278> linux softraid can easilly do 700mb/s using intel ich10r ports
3244 2011-07-24 22:36:49 zeropointo has quit (Remote host closed the connection)
3245 2011-07-24 22:37:22 <Zagitta> yeah it's all just old hardware, i wasn't expecting anything fancy but 140->160 in difference for 2 extra discs just made me wonder why
3246 2011-07-24 22:37:27 <gjs278> well
3247 2011-07-24 22:37:28 <gjs278> what's your test
3248 2011-07-24 22:37:30 <gjs278> bonnie?
3249 2011-07-24 22:37:42 Beccara has quit (Ping timeout: 255 seconds)
3250 2011-07-24 22:37:50 <Zagitta> i was using the built in one gnome
3251 2011-07-24 22:37:57 <Zagitta> *in gnome
3252 2011-07-24 22:38:08 <gjs278> does that just use hdparm
3253 2011-07-24 22:38:45 <Zagitta> think so, but i switched to iozone to bench lvm
3254 2011-07-24 22:39:28 <Zagitta> should probably have tested the other setup with that aswell to be able to compare but meh
3255 2011-07-24 22:39:57 <gjs278> yeah
3256 2011-07-24 22:40:13 <gjs278> hdparm is never correct
3257 2011-07-24 22:40:18 <gjs278> iozone will be
3258 2011-07-24 22:40:22 <gjs278> bonnie will be
3259 2011-07-24 22:41:34 <Zagitta> now idk how many tests iozone runs but this 10 gig test is taking an awefull long time to complete
3260 2011-07-24 22:42:13 <gjs278> yes it will
3261 2011-07-24 22:42:16 <gjs278> because it is doing a 4k test
3262 2011-07-24 22:42:21 <gjs278> your 4k test will take your whole life
3263 2011-07-24 22:42:28 <gjs278> how much ram does the machine have
3264 2011-07-24 22:42:38 <gjs278> you generally only have to do ram size x2 to get accurate results
3265 2011-07-24 22:42:43 <CIA-103> bitcoin: Jeff Garzik master * ra9ba471 / (13 files):
3266 2011-07-24 22:42:43 <CIA-103> bitcoin: Merge pull request #403 from sipa/cbitcoinaddress
3267 2011-07-24 22:42:43 <CIA-103> bitcoin: keys indexed by address + introduced CBitcoinaddress - https://github.com/bitcoin/bitcoin/commit/a9ba47101a46533ff0418d6868ebc9bf5c889818
3268 2011-07-24 22:43:18 <BlueMatt> jgarzik: yay
3269 2011-07-24 22:43:19 <Dagger2> Zagitta: uhm, the RX100's SCSI controller is U160, which has a maximum throughput of 160 MB/s
3270 2011-07-24 22:43:23 <Dagger2> I think you might be hitting that
3271 2011-07-24 22:43:31 <Dagger2> just possibly.
3272 2011-07-24 22:43:47 <Diablo-D3> lol
3273 2011-07-24 22:43:52 <Diablo-D3> time to throw that controller out
3274 2011-07-24 22:43:54 <Zagitta> well that was one hell of a time waste then lol
3275 2011-07-24 22:43:56 <Diablo-D3> and get a cheap sata/sas controller
3276 2011-07-24 22:43:58 <gjs278> that's probably it
3277 2011-07-24 22:44:17 <Zagitta> Diablo-D3: i got all of the hardware for free so can't really complain
3278 2011-07-24 22:44:40 <gjs278> http://img28.imageshack.us/img28/9094/forcegtvsc300.png
3279 2011-07-24 22:44:57 <gjs278> that's my setup on 2 different raid0s
3280 2011-07-24 22:45:23 <Zagitta> regular discs or ssd's?
3281 2011-07-24 22:45:30 <gjs278> ssds
3282 2011-07-24 22:45:34 <gjs278> the 15k sas drives at work
3283 2011-07-24 22:45:37 <jgarzik> BlueMatt, sipa, tcatm: time to get moving on wallet import/export :)
3284 2011-07-24 22:45:49 <gjs278> like 12 of them together pulls in 900mb/s
3285 2011-07-24 22:46:24 <gjs278> that's in a raid6 though
3286 2011-07-24 22:46:27 <Zagitta> nice, my kingston ssd in my desktop does like 220 mb/s
3287 2011-07-24 22:46:37 <BlueMatt> jgarzik: ipv6 for 0.4?
3288 2011-07-24 22:46:47 <Diablo-D3> Zagitta: yeah, but a cheap sata or sas controller is really cheap
3289 2011-07-24 22:47:08 <Diablo-D3> a lot of enterrprise users with parallel scsi in boxes ripped it out
3290 2011-07-24 22:47:16 <Diablo-D3> assuming they didnt just junk the boxes
3291 2011-07-24 22:47:29 <Diablo-D3> the cost of maintaining parallel scsi hardware is too high
3292 2011-07-24 22:47:29 <gjs278> some hp cards can be reflashed to LSI firmware
3293 2011-07-24 22:47:33 <Diablo-D3> sas drives are fucking cheap
3294 2011-07-24 22:47:51 osmosis has joined
3295 2011-07-24 22:48:05 <JFK911> gjs278: right on, SATA ports work faster than just about any parallel scsi i've used
3296 2011-07-24 22:48:17 <JFK911> ich10 is fast as hell but its bios raid sucks
3297 2011-07-24 22:48:18 DukeOfURL has joined
3298 2011-07-24 22:48:21 <gjs278> yeah
3299 2011-07-24 22:48:23 <JFK911> (it wont stripe reads)
3300 2011-07-24 22:48:46 osmosis has quit (Remote host closed the connection)
3301 2011-07-24 22:48:51 <gjs278> I used ich10r but then moved to an lsi 9260 and then sold that and got 9265
3302 2011-07-24 22:49:05 <jgarzik> BlueMatt: I am an IPv6 fan, so I like it.  There is an unaddressed worry about nodes w/ IPv4 and IPv6 nodes somehow appearing twice.
3303 2011-07-24 22:49:10 <Diablo-D3> [06:44:14] <JFK911> ich10 is fast as hell but its bios raid sucks
3304 2011-07-24 22:49:12 <JFK911> when i did use parallel scsi i had each mirror component on a different chain
3305 2011-07-24 22:49:14 <Diablo-D3> JFK911: because it doesnt have raid
3306 2011-07-24 22:49:22 <Diablo-D3> just use normal swraid in linux
3307 2011-07-24 22:49:23 <Diablo-D3> much faster
3308 2011-07-24 22:49:26 <JFK911> i know it is
3309 2011-07-24 22:49:29 <gjs278> I had to dualboot
3310 2011-07-24 22:49:29 <jgarzik> JFK911: what fscking idiot uses bios raid?
3311 2011-07-24 22:49:34 <gjs278> so ich10 was all I had for a bit
3312 2011-07-24 22:49:35 <Diablo-D3> jgarzik: windows users
3313 2011-07-24 22:49:45 <JFK911> well, @ windoze the softraid is faster
3314 2011-07-24 22:49:51 <JFK911> than linux'
3315 2011-07-24 22:49:55 <Diablo-D3> windows STILL does not support partition mirroring w
3316 2011-07-24 22:49:59 <Diablo-D3> I have no clue why
3317 2011-07-24 22:50:11 <Zagitta> i actually have an extra scsi controller that came with it sitting in one of the pci-X ports but i ripped it out to reduce boot time
3318 2011-07-24 22:50:12 <jgarzik> JFK911: not according to my numbers
3319 2011-07-24 22:50:56 <gjs278> lol
3320 2011-07-24 22:50:57 <AlonzoTG> yeah, that's one of my biggest worries about bitcoin -- protecting the wallet file.
3321 2011-07-24 22:50:59 <gjs278> put that controller back in
3322 2011-07-24 22:51:04 <JFK911> i couldnt get md's mirror to stripe reads without "raid10"
3323 2011-07-24 22:51:22 <JFK911> and then it worked differently in various sits depending on if i had the layout far or not
3324 2011-07-24 22:51:35 <gjs278> if that controller was in pci-x you should be using it, I guaranatee the thing can do better than 160
3325 2011-07-24 22:52:06 <Diablo-D3> well its not like pci-x is magical
3326 2011-07-24 22:52:07 <JFK911> single 7k1000 = 133 read 102 write; d0ze mirror = 279 read 130 write
3327 2011-07-24 22:52:16 <Diablo-D3> the slowest pci-x speed is just 64bit/66mhz pci.
3328 2011-07-24 22:52:59 <JFK911> striping reads in a mirror is only common sense
3329 2011-07-24 22:53:09 <JFK911> idk why linux made it so rude
3330 2011-07-24 22:53:10 <Diablo-D3> JFK911: you dont run raid1 for speed
3331 2011-07-24 22:53:22 <sipa> jgarzik: i think some better way of dealing with address selection/storing may be necessary before ipv6 is merged
3332 2011-07-24 22:53:23 <Diablo-D3> you run it for shit drive proofing.
3333 2011-07-24 22:53:28 <gjs278> lsi says the fastest their card can do in 2.0 express is 2.8gb/s
3334 2011-07-24 22:53:39 <Diablo-D3> gjs278: what width?
3335 2011-07-24 22:53:41 <Diablo-D3> x4?
3336 2011-07-24 22:53:45 <gjs278> x16
3337 2011-07-24 22:53:48 <gjs278> or 8
3338 2011-07-24 22:53:49 <JFK911> Diablo-D3: well, in this instance, MS thought of the obvious
3339 2011-07-24 22:54:19 <Diablo-D3> gjs278: pci-e 2.0 x8 is 32 gbit/sec
3340 2011-07-24 22:54:30 <Zagitta> well it's a pci-x and not just pci card... no clue what controller it is though and i'll also have to reroute the wiring from the front of the case :s normally the hotswap bays are just linked to the mb and this controller was used for an external tape drive or something
3341 2011-07-24 22:54:31 <gjs278> hold on
3342 2011-07-24 22:54:34 <Diablo-D3> 1.0 x8 is half that, 2.0 x4 is half that, 2.0 x16 is twice that, etc
3343 2011-07-24 22:54:35 <gjs278> I'm looking at this stuff right now
3344 2011-07-24 22:54:44 <Diablo-D3> gjs278: they must be quoting the controller's speed
3345 2011-07-24 22:54:55 <gjs278> yeah
3346 2011-07-24 22:55:10 <Diablo-D3> is that 2.8gbyte/sec or gbit?
3347 2011-07-24 22:55:11 <gjs278> the marvell raid on most motherbaords is on pci 1x
3348 2011-07-24 22:55:13 <gjs278> byte
3349 2011-07-24 22:55:17 <Diablo-D3> hah
3350 2011-07-24 22:55:23 <Diablo-D3> thats like enough for a 24 drive controller
3351 2011-07-24 22:55:42 <JFK911> the sata3 marvell chip on my board is hanging off the northbridge
3352 2011-07-24 22:55:54 <JFK911> probably the lesser interfaces are stuck at 1.0 rate
3353 2011-07-24 22:55:58 <gjs278> 2.8 = 22gbit so yeah controller limit for now
3354 2011-07-24 22:56:17 <Diablo-D3> 2.8/24 = 116 each
3355 2011-07-24 22:56:19 <Diablo-D3> which is enough
3356 2011-07-24 22:56:37 <gjs278> people buy them with the 24 port expanders from intel or hp
3357 2011-07-24 22:56:42 <gjs278> so that's usually what happens
3358 2011-07-24 22:56:48 <Diablo-D3> yeah, sas backplanes == nice
3359 2011-07-24 22:56:56 <gjs278> because the card only has 8 ports by default
3360 2011-07-24 23:01:54 <Zagitta> this is the controller: http://imagebin.org/164709 don't know if anyone can recognize it
3361 2011-07-24 23:02:14 <BlueMattBot> Project Bitcoin-Test build #14: STILL FAILING in 11 sec: http://www.bluematt.me/jenkins/job/Bitcoin-Test/14/
3362 2011-07-24 23:02:15 <BlueMattBot> * pieter.wuille: get rid of mapPubKeys
3363 2011-07-24 23:02:15 <BlueMattBot> * pieter.wuille: Use CBitcoinAddress instead of string/uint160
3364 2011-07-24 23:02:15 <BlueMattBot> * pieter.wuille: split off CBase58Data from CBitcoinAddress
3365 2011-07-24 23:02:16 <BlueMattBot> * abuse: Fix bug with accessing vchData[0] when vchData is empty.
3366 2011-07-24 23:02:33 <sipa> BlueMatt: can you fix that, please :)
3367 2011-07-24 23:02:50 <BlueMatt> not until it gets fixed
3368 2011-07-24 23:02:54 RenaKunisaki has quit (Ping timeout: 255 seconds)
3369 2011-07-24 23:02:57 <BlueMatt> well...
3370 2011-07-24 23:03:21 <gjs278> Zagitta you'd haev an easier time just booting with the card and checking inside its bios
3371 2011-07-24 23:03:32 RenaKunisaki has joined
3372 2011-07-24 23:03:47 <sipa> BlueMatt: wait, what's the problem exactly?
3373 2011-07-24 23:03:50 <BlueMatt> sipa: it needs parts of pull 363
3374 2011-07-24 23:03:56 <BlueMatt> the test suite doesnt build
3375 2011-07-24 23:04:06 <Zagitta> gjs278: good point except it's on the roof :P oh well, i'll throw it in tomorrow
3376 2011-07-24 23:04:11 bitcoinbulletin has quit (Remote host closed the connection)
3377 2011-07-24 23:04:14 * AlonzoTG tries to figure out the transaction and block database operations. 
3378 2011-07-24 23:04:42 AStove has quit (Ping timeout: 255 seconds)
3379 2011-07-24 23:05:58 Clipse has joined
3380 2011-07-24 23:06:09 <jgarzik> BlueMatt: current upstream builds fine for me
3381 2011-07-24 23:06:23 <BlueMatt> it does on my desktop, but on the server, it doesnt...
3382 2011-07-24 23:06:25 * AlonzoTG has idea. =0
3383 2011-07-24 23:06:30 <BlueMatt> jgarzik: bitcoin_test not bitcoin
3384 2011-07-24 23:06:39 <BlueMatt> sorry, test_bitcoin
3385 2011-07-24 23:06:42 <AlonzoTG> =P
3386 2011-07-24 23:06:44 <AlonzoTG> =(
3387 2011-07-24 23:06:45 <AlonzoTG> =$
3388 2011-07-24 23:06:53 <AlonzoTG> idea = quick payment protocol.
3389 2011-07-24 23:07:13 <AlonzoTG> When you are at the register at the quickie mart and you try to pay with bitcoins,
3390 2011-07-24 23:07:47 SomeoneWeirdTAFE has joined
3391 2011-07-24 23:07:58 <AlonzoTG> the store would give not only their payment address but also the IP address so that your wallet server can connect directly to the store's IP address and submit the transaction directly so that it can be verified in short order.
3392 2011-07-24 23:08:03 mosimo has quit (Quit: ( www.nnscript.com :: NoNameScript 4.22 :: www.esnation.com ))
3393 2011-07-24 23:08:47 <sipa> AlonzoTG: or using NFC
3394 2011-07-24 23:08:56 <AlonzoTG> ???
3395 2011-07-24 23:09:12 larsivi has quit (Ping timeout: 255 seconds)
3396 2011-07-24 23:09:12 <sipa> google it
3397 2011-07-24 23:09:13 <SuprTiggr> http://en.wikipedia.org/wiki/Information_technology
3398 2011-07-24 23:09:40 <BlueMatt> sipa: if you redo pull 363 without the -I. and relative paths for includes, test_bitcoin works fine
3399 2011-07-24 23:10:18 <sipa> vegard: ping
3400 2011-07-24 23:12:06 <sipa> AlonzoTG: for instant payments, it's possible to carry a bitcoin wallet which does not have any network communication capbilities - it just gets an amount + address to pay to, and creates a transaction that does so after getting your confirmations
3401 2011-07-24 23:12:29 <sipa> could be a smart card, or a mobile phone, or ...
3402 2011-07-24 23:14:21 bitcoinbulletin has joined
3403 2011-07-24 23:16:39 <AlonzoTG> How would it check balances?
3404 2011-07-24 23:16:48 <JFK911> it has the coins
3405 2011-07-24 23:17:13 <AlonzoTG> It was my understanding that coins are all in the form of transaction-outs.
3406 2011-07-24 23:17:18 Incitatus has quit (Ping timeout: 255 seconds)
3407 2011-07-24 23:18:34 <lfm> It might not know about incomming payments till it syncs up with a base
3408 2011-07-24 23:22:04 <luke-jr> AlonzoTG: see wallet protocol on wiki
3409 2011-07-24 23:22:25 <sipa> AlonzoTG: that is correct
3410 2011-07-24 23:22:39 <sipa> AlonzoTG: and since it has coins, it knows those can be spent
3411 2011-07-24 23:23:22 <sipa> you would "charge" it using a trusted system
3412 2011-07-24 23:25:27 zeropointo has joined
3413 2011-07-24 23:27:58 <lfm> it could sync up with just block explorer if you trust it
3414 2011-07-24 23:28:08 spirals has quit (Read error: Connection reset by peer)
3415 2011-07-24 23:28:38 spirals has joined
3416 2011-07-24 23:28:45 spirals has quit (Changing host)
3417 2011-07-24 23:28:45 spirals has joined
3418 2011-07-24 23:29:12 <cjdelisle> I wonder who controls that 5 terrahash mining network that just got flipped on.
3419 2011-07-24 23:29:19 TD has quit (Quit: TD)
3420 2011-07-24 23:29:46 spirals_ has joined
3421 2011-07-24 23:29:59 Incitatus has joined
3422 2011-07-24 23:30:45 <cjdelisle> The way I calculate, it it's all boxes just like mrb's it's 2.1 megawatts of electricity and 6 million $ worth of equipment.
3423 2011-07-24 23:31:10 <lfm> or a botnet
3424 2011-07-24 23:32:17 <sipa> where do you see there is a 5 TH increase?
3425 2011-07-24 23:32:21 <sipa> ;;bc,nethash
3426 2011-07-24 23:32:22 <gribble> 13370.987974554551
3427 2011-07-24 23:32:40 kreal- has quit (Quit: leaving)
3428 2011-07-24 23:32:49 <cjdelisle> it was at 15 TH earlier today according to bitcoincharts.com
3429 2011-07-24 23:32:56 <cjdelisle> and it was at like 10 the other day
3430 2011-07-24 23:33:04 spirals has quit (Ping timeout: 250 seconds)
3431 2011-07-24 23:33:07 <gmaxwell> those numbers are very noisy
3432 2011-07-24 23:33:10 <sipa> it's just a very rough estimate
3433 2011-07-24 23:33:21 <gmaxwell> non-psycho numbers don't support that statement: http://bitcoin.sipa.be/speed-lin-2k.png
3434 2011-07-24 23:34:48 nhodges has quit (Ping timeout: 250 seconds)
3435 2011-07-24 23:35:17 <cjdelisle> ahh indeed the numbers bounce around more than I had thought, I just happened to take samples at "convienent" times.
3436 2011-07-24 23:36:46 freakazoid has left ("Textual IRC Client: http://www.textualapp.com/")
3437 2011-07-24 23:36:49 spirals_ has quit (Ping timeout: 240 seconds)
3438 2011-07-24 23:37:30 kreal- has joined
3439 2011-07-24 23:37:41 nhodges has joined
3440 2011-07-24 23:44:20 nhodges has quit (Ping timeout: 250 seconds)
3441 2011-07-24 23:44:22 osmosis has joined
3442 2011-07-24 23:44:58 nhodges has joined
3443 2011-07-24 23:46:56 erus` has quit (Remote host closed the connection)
3444 2011-07-24 23:49:07 sipa has left ()
3445 2011-07-24 23:52:59 <jrmithdobbs> jgarzik: do you know of a method on freebsd similar to what you use with rngd to feed the linux kernel
3446 2011-07-24 23:53:02 <jrmithdobbs> ?
3447 2011-07-24 23:53:45 <jrmithdobbs> or darwin since that's what i'm *really* interested in
3448 2011-07-24 23:53:58 <jrmithdobbs> i'm looking for a way to feed entropy on os x
3449 2011-07-24 23:57:22 <Zagitta> gj278: if i'm not reading this data wrong i'm only getting like 60mb/s on this LVM :3
3450 2011-07-24 23:57:49 Beccara has joined