Timestamps are in GMT/BST.
[0:01] * piranha (~piranha@192.220.223.46) has joined #freenet
[0:26] * TLF (francisco@128.Red-81-40-116.pooles.rima-tde.net) has joined #freenet
[1:07] * TLF (francisco@128.Red-81-40-116.pooles.rima-tde.net) Quit (Read error: 104 (Connection reset by peer))
[1:07] * TLF (francisco@128.Red-81-40-116.pooles.rima-tde.net) has joined #freenet
[1:20] * TLF (francisco@128.Red-81-40-116.pooles.rima-tde.net) Quit ("http://www.vitaliano.esp.cc || Adi?s a todos y a todas y que les vaya bien.")
[1:20] * TLF (francisco@128.Red-81-40-116.pooles.rima-tde.net) has joined #freenet
[1:50] * ysothoth (~ysothoth@68-174-22-46.si.rr.com) Quit (Read error: 104 (Connection reset by peer))
[2:51] <mikeDOTd> http://www.freenetproject.org/snapshots/freenet-latest.jar is build 5085, not 5086
[3:31] * chotchki (~chotchki@61-61.73-24.tampabay.rr.com) has joined #freenet
[3:52] * Pascal666 (Pascal@c-24-13-17-191.client.comcast.net) has joined #Freenet
[3:52] * Pascal (Pascal@c-24-13-17-191.client.comcast.net) Quit (Read error: 104 (Connection reset by peer))
[3:52] * Pascal666 is now known as Pascal
[3:55] * chotchki (~chotchki@61-61.73-24.tampabay.rr.com) Quit ("Leaving")
[4:00] * piranha (~piranha@192.220.223.46) Quit (Read error: 113 (No route to host))
[4:01] * piranha (~piranha@192.220.223.34) has joined #freenet
[4:13] * BrianR (brianr@2002:1893:ac33:0:0:0:0:1) Quit (Client Quit)
[4:41] * Pascal666 (Pascal@c-24-13-17-191.client.comcast.net) has joined #Freenet
[4:41] * Pascal (Pascal@c-24-13-17-191.client.comcast.net) Quit (Read error: 104 (Connection reset by peer))
[4:41] * Pascal666 is now known as Pascal
[5:28] * Ash-Foxeus (Hal-9000@aap55.neoplus.adsl.tpnet.pl) has joined #FREENET
[5:34] * Ash-Fox (Hal-9000@aad77.neoplus.adsl.tpnet.pl) Quit (Read error: 60 (Operation timed out))
[5:45] * nicktastic (~nicktasti@c-67-163-227-74.client.comcast.net) Quit (Remote closed the connection)
[5:56] * mazzanet (~mazzanet@203-173-44-61.dyn.iinet.net.au) Quit ("brb rebooting")
[6:10] * TLF (francisco@128.Red-81-40-116.pooles.rima-tde.net) has left #freenet
[6:12] * ShaunMacPherson (~ShaunMac@Sudbury-HSE-ppp3976660.sympatico.ca) has left #freenet
[6:14] * Hory (~Miranda@81.196.25.110) has joined #FreeNet
[6:15] * mazzanet (~mazzanet@203-173-44-61.dyn.iinet.net.au) has joined #freenet
[6:52] * guido^pe (~unknown@dsl-213-023-245-203.arcor-ip.net) has joined #freenet
[7:16] * TLF (francisco@79.Red-81-40-113.pooles.rima-tde.net) has joined #freenet
[7:24] * TLF (francisco@79.Red-81-40-113.pooles.rima-tde.net) Quit ("http://www.vitaliano.esp.cc || Adi?s a todos y a todas y que les vaya bien.")
[7:30] * guido^pe (~unknown@dsl-213-023-245-203.arcor-ip.net) Quit (Remote closed the connection)
[7:31] * guido^pe (~unknown@dsl-213-023-245-203.arcor-ip.net) has joined #freenet
[7:43] * guido^pe (~unknown@dsl-213-023-245-203.arcor-ip.net) Quit (Remote closed the connection)
[7:46] * guido^pe (~unknown@dsl-213-023-245-203.arcor-ip.net) has joined #freenet
[7:47] * nicktastic (~nicktasti@c-67-163-227-74.client.comcast.net) has joined #freenet
[7:50] * sylvinus (~sylvinus@ca-metz-4-201.w80-8.abo.wanadoo.fr) Quit (Remote closed the connection)
[7:53] * guido^pe (~unknown@dsl-213-023-245-203.arcor-ip.net) Quit (Remote closed the connection)
[7:58] <Iakin> Anyone else besides me that experiences connection problems with the latest unstable?
[7:58] <hirvox> no, 65 open connections
[7:58] <Iakin> No matter how long time the node is up for it wont establish any outbound connections
[8:00] <hirvox> 29 inbound and 36 outbound
[8:00] <hirvox> build 60168
[8:01] * TLF (francisco@16.Red-81-40-116.pooles.rima-tde.net) has joined #freenet
[8:01] <Iakin3> From my log:
[8:01] <Iakin3> 14:17:14 Caught java.io.IOException: Already closed tcp/connection: CLOSED,freenet.transport.tcpConnection@18297fe setting timeout on tcp/connection: CLOSED,freenet.transport.tcpConnection@18297fe -
[8:01] <Iakin3> 14:18:14 waited more than 120000ms in NIOIS.read() tcp/connection: 4626>128.119.82.146:7831,freenet.transport.tcpConnection@16535bf:freenet.support.io.NIOInputStream@1ab9dac- closing java.lang.Exception: debug
[8:01] <Iakin3> 14:18:15 waited more than 120000ms in NIOIS.read() tcp/connection: 4646>219.98.119.126:3148,freenet.transport.tcpConnection@9036e:freenet.support.io.NIOInputStream@1c059f6- closing java.lang.Exception: debug
[8:01] <Iakin3> 14:18:40 waited more than 120000ms in NIOIS.read() tcp/connection: 4729>211.132.77.194:39852,freenet.transport.tcpConnection@14b1fda:freenet.support.io.NIOInputStream@1183871- closing java.lang.Exception: debug
[8:01] <Iakin3> Connections open (Inbound/Outbound/Limit) 0 (0/0/1000)
[8:01] <Iakin3> bbl
[8:08] * guido^pe (~unknown@dsl-213-023-245-203.arcor-ip.net) has joined #freenet
[8:08] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) Quit (Remote closed the connection)
[8:11] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) has joined #freenet
[8:12] * guido^pe (~unknown@dsl-213-023-245-203.arcor-ip.net) Quit (Remote closed the connection)
[8:19] * piranha (~piranha@192.220.223.34) Quit (Read error: 113 (No route to host))
[8:28] * Iakin (Iakin@as25-4-6.kp.g.bonet.se) Quit (sterling.freenode.net irc.freenode.net)
[8:28] * pwk_ (~phweak@dsl-62-3-71-168.zen.co.uk) Quit (sterling.freenode.net irc.freenode.net)
[8:28] * verl (~verlverl@h165n2fls33o877.telia.com) Quit (sterling.freenode.net irc.freenode.net)
[8:31] * pwk_ (~phweak@dsl-62-3-71-168.zen.co.uk) has joined #freenet
[8:40] * nicktastic (~nicktasti@c-67-163-227-74.client.comcast.net) Quit ()
[9:10] * TLF (francisco@16.Red-81-40-116.pooles.rima-tde.net) Quit ("http://www.vitaliano.esp.cc || Adi?s a todos y a todas y que les vaya bien.")
[9:36] * jay (jcl@ool-18bf6dac.dyn.optonline.net) has joined #freenet
[9:41] * greycat (rfc1413@wooledge.org) has joined #freenet
[10:00] * nicktastic (~nicktasti@c-67-163-227-74.client.comcast.net) has joined #freenet
[10:05] * jay (jcl@ool-18bf6dac.dyn.optonline.net) Quit ("Client exiting")
[10:19] * jmercier (~jmercier@4.10.164.111) has joined #freenet
[10:19] * verl (~verlverl@h165n2fls33o877.telia.com) has joined #freenet
[10:32] * kork (~kork@2001:618:400:0:0:0:d945:a58f) has joined #freenet
[10:33] <kork> hi
[10:39] * kork (~kork@2001:618:400:0:0:0:d945:a58f) Quit ("leaving")
[10:43] * guido^pe (~unknown@dsl-213-023-245-203.arcor-ip.net) has joined #freenet
[10:44] <KenMan> bye ?
[10:50] * hilo (~none@24-155-81-84.ip.grandenetworks.net) has joined #freenet
[10:50] * hilo (~none@24-155-81-84.ip.grandenetworks.net) has left #freenet
[11:44] <toad_> hi
[11:46] <toad_> hmm http://www.eff.org/IP/P2P/p2p_copyright_wp.php
[11:47] <toad_> looking pretty good, although advice point 1 is dubious
[11:48] <toad_> |merely notifying Xerox that one of its photocopiers is being misused at a neighborhood Kinko's would not suddenly make Xerox a contributory infringer?Xerox has no ability to disable, repossess, or remotely control its photocopiers once they have been sold.
[11:48] <toad_> nice analogy
[11:49] <iip_i2p> <Newsbyte> toad, you there?
[11:50] <toad_> yeah
[11:50] <nicktastic> oh god
[11:50] <toad_> nicktastic: hmm?
[11:50] <nicktastic> nothing
[11:50] * toad_ will get back to work soon
[11:52] <toad_> #3 is a strong argument against automatic updates on a production network :|
[11:52] <toad_> we can do it on a testnet though
[11:53] <iip_i2p> <Newsbyte> did I hear the word 'testnetwork'?
[11:53] <nicktastic> toad_: Is http://freenetproject.org/snapshots/freenet-latest.jar supposed to be 5086?
[11:53] <toad_> nicktastic: yes
[11:53] <nicktastic> Hmm
[11:53] * toad_ checks
[11:53] <iip_i2p> <Newsbyte> so, toad, we're a day further...what's the verdict on the new build?
[11:53] <toad_> nicktastic: yes
[11:54] <iip_i2p> <Newsbyte> 8085, that is
[11:54] <toad_> #4 is good...
[11:54] <toad_> Newsbyte: I don't know
[11:54] <iip_i2p> <Newsbyte> that good?
[11:54] <iip_i2p> <Newsbyte> ;-)
[11:54] <toad_> #5 we should do more of
[11:54] <toad_> #6 - well duh!
[11:54] <iip_i2p> <Newsbyte> what was #3 again?
[11:55] <toad_> Newsbyte: no automatic update
[11:55] <nicktastic> Ah, so it is
[11:55] <toad_> "Better to sell stand-alone software products than on-going services. "
[11:56] <iip_i2p> <Newsbyte> what about my suggestion of more rapidly moving to other nodes then those in seednodes.ref?
[11:56] <toad_> #6 - difficult, beyond keeping Frost out of the mainline distro
[11:56] <toad_> Newsbyte: well, don't we do that already?
[11:56] <toad_> the main thing is to get announcements working
[11:56] <iip_i2p> <Newsbyte> uh? NO automatic update? I thought one would say just the oposite. I mean, you don't think newbies will constantly change/update stuff manually?
[11:56] <toad_> which is on the todo list
[11:57] <toad_> Newsbyte: legally it appears to be better not to have an auto update because then you can't be made to abuse it
[11:57] <toad_> #8 - :)
[11:58] <iip_i2p> <Newsbyte> I don't know to what extent. Can you test it out on your mini-testnetwork what it does, if you make it faster move to other nodes that aren't in the seednodes.ref?
[11:58] <toad_> #9 - need to check the wininstaller
[11:58] <nicktastic> Well, the point is "At a minimum, users should always retain the ability to decline any update. Control should rest in the end-user's hands, not the software vendor's."
[11:58] <toad_> nicktastic: well yeah..
[11:58] <toad_> Newsbyte: atm, announcements don't work :|
[11:59] <toad_> #10 repeats: NO AUTOMATIC UPDATING !!
[11:59] <iip_i2p> <Newsbyte> could you elaborate on the legal reason for not auto-updating a bit?
[11:59] <toad_> #11 is very nasty
[11:59] <toad_> Newsbyte: read the URL
[12:00] <toad_> search for "3. Better"
[12:00] <iip_i2p> <Newsbyte> what url?
[12:00] <toad_> http://www.eff.org/IP/P2P/p2p_copyright_wp.php
[12:01] <toad_> i'm reasonably careful about #11... still it can be difficult
[12:01] <toad_> no customer support if they openly admit bootlegging copyrighted files
[12:01] <KenMan> you know, if toad held a private key, and signed "messages" that announced the latest stable version, then they could be shown to the user
[12:01] <KenMan> but the current "at least 3 nodes" or whatever logic it is seems sufficient
[12:01] <toad_> #12 is good
[12:01] <nicktastic> Isn't that kind of hokey in the case of Freenet, where new builds are frequently manditory?
[12:02] <toad_> nicktastic: that's a concern
[12:02] <toad_> you can however keep running old builds
[12:02] <toad_> you'll just have to fork the network... :|
[12:02] <nicktastic> toad_: Whee
[12:02] <nicktastic> ;)
[12:02] <KenMan> it would simply ensure that users are shown authenticated "latest version" messages
[12:03] <iip_i2p> <Newsbyte> are you referring to the 'Contributory infringement' ?
[12:03] <toad_> Newsbyte: it talks about both in the advice section at the bottom
[12:04] <KenMan> we are implicitly (fully) responsible if we build a gun, and then someone chooses to use it to shoot someone...
[12:04] <nicktastic> haha
[12:04] <iip_i2p> <Newsbyte> one worth of caution: the page only covers USA law
[12:05] <KenMan> the interesting thing in some of that INDUCE act is that it implies P2P developer-people seek capital gain from their 'infringing' activities
[12:05] <KenMan> actually, i recall this from the MaryBeth Copyright office testimony
[12:06] <iip_i2p> <Newsbyte> for instance, they keep saying copies in RAM are infringing too, but IIRC, under european law, this does not count as copy-infringement
[12:06] <KenMan> Ms MaryBeth seemed to focus on companies developing P2P, and indirectly-implied (to me) that open source P2P developers would be non-infringing.
[12:07] <KenMan> Of course, that is simply because OS slipped her radar :(
[12:08] <KenMan> but the angle they claimed to be focused on was removing the profit-incentive from encouraging users to infringe (C)'s ...
[12:09] <nicktastic> Newsbyte, Its often best to plan for the worst case
[12:10] <KenMan> which is why they are flawed yet again. The goal is simply to reduce (C) infringement , using bandaids, duct tape, or whatever tools are at their disposal (like generation of poor legislation)
[12:11] <iip_i2p> <Newsbyte> true, but reading it further, I don't see anything alarming in the analysis on the page. Freenet makes a good example of 'anarchistic' P2P system, as he says courts will probably rule such systems can not be held liable
[12:12] <KenMan> no, but then the '(C) police' must focus on attacking the users. And if the anarchistic systems make it too hard, we can save money by just outlawing the enabling technology.
[12:12] <iip_i2p> <Newsbyte> that said, let's be realistic; no newbie is going to manually update the stuff like we die-hards are, and the risk of an auto-update is, quite frankly, very remote
[12:12] <iip_i2p> <Newsbyte> so, I would be against it, at least untill 1.0 comes out, and then maybe we can just distribute it in Freenet itself
[12:15] <iip_i2p> <Newsbyte> the courts have already ruled you can not outlaw something because of copyrightinfringements, when it has substantial other use, btw
[12:16] <iip_i2p> <Newsbyte> ofcourse, china&co could (and probably will)
[12:33] <iip_i2p> <Newsbyte> me notices everyone agrees in silence ;-)
[12:44] <toad_> the stuff on that page isn't alarming
[12:44] <toad_> INDUCE is alarming
[12:44] <toad_> <KenMan> the interesting thing in some of that INDUCE act is that it implies P2P developer-people seek capital gain from their 'infringing' activities
[12:45] <toad_> sadly, Freenet Project Inc will need to obtain cash to pay me
[12:45] <toad_> :)
[12:45] <toad_> <iip_i2p> <Newsbyte> for instance, they keep saying copies in RAM are infringing too, but IIRC, under european law, this does not count as copy-infringement - maybe so, but if you transfer it that will be
[12:46] <toad_> <iip_i2p> <Newsbyte> the courts have already ruled you can not outlaw something because of copyrightinfringements, when it has substantial other use, btw - no, they've ruled that current legislation means you can't. you can always pass new legislation.
[12:47] <toad_> and we'll never get a constitutional guarantee in the EU, because our constitutions are so weak
[12:47] <toad_> in the US they'll find a way to work around the constitution if it comes up
[12:47] <toad_> e.g. the failed challenge to Sonny Bono
[12:47] <toad_> brb
[12:48] <nicktastic> Again, always plan for the worst case
[12:48] <toad_> nicktastic: worst case: INDUCE passes, we relocate FP.org to the EU
[12:49] <toad_> the EU passes something similar because they believe it will enhance their competitiveness rather than crippling them for generations to come, through some software-patent-esque doublespeak
[12:49] <toad_> and we can no longer pay a full time developer
[12:49] <toad_> and we have to make development reasonably anonymous
[12:50] <toad_> then things get nasty.. but only if we can make the bloody thing WORK
[12:50] <toad_> :)
[12:50] <nicktastic> toad_: One step at a time ;)
[12:50] <toad_> :)
[12:50] <toad_> brb
[12:53] <toad_> bbiab, ~~20 minutes
[12:54] <iip_i2p> <Newsbyte> the argument of 'you can always pass new legislation' is a moot one. They can pass laws that outlaw computers and the internet too. With that reasoning, everything can be outlawed.
[12:57] <iip_i2p> <Newsbyte> and I don't think it's correct about 'transferring' it. A lot of ISPs use cache to make the Net work faster, and there too, it gets 'transferred'...yet, it was exactly for that reason they decided caching wasn't an infringement. If it were, then the Net would become almost unworkable. At least that was their argument, and the EU has folled them in that reasoning.
[13:01] * jmercier (~jmercier@4.10.164.111) Quit ("Leaving")
[13:05] * guido^pe (~unknown@dsl-213-023-245-203.arcor-ip.net) Quit (Remote closed the connection)
[13:22] <toad_> |They can pass laws that outlaw computers and the internet too. - arguably that's what INDUCE does...
[13:23] <toad_> okay, to work...
[13:27] * toad_ is not going to set up new version control server yet
[13:27] * toad_ will just work on traffic analysis etc today
[13:39] <toad_> <lostlogic> why on earth are 35 threads opening connection?
[13:39] <toad_> ummm
[13:39] <toad_> because they're needed?
[13:39] <toad_> <mikeDOTd> RNF: 58 backed off nodes, yay! - out of how many?
[13:41] <toad_> <Pascal> toad: I wonder if L66-L72 are also in the wrong order - I don't get it...
[13:42] <toad_> Writing 699 on freenet.node.states.data.SendData@5819cf
[13:42] <toad_> Writing 447 on freenet.node.states.data.SendData@7b6c3e
[13:43] <toad_> Writing 161 on freenet.node.states.data.SendData@16c97e6
[13:43] <toad_> Writing 2 on freenet.node.states.data.SendData@7b6c3e - hmm, must be end of file
[13:43] <toad_> Writing 601 on freenet.node.states.data.SendData@c247e0
[13:44] <toad_> the overwhelming majority are 699
[13:44] <toad_> Writing 301 on freenet.node.states.data.SendData@d8cd70
[13:47] <iip_i2p> <Newsbyte> feel free to begin working on that testnetwork any time now, toad!
[13:48] <iip_i2p> <Newsbyte> :-)
[13:51] * jmercier (~jmercier@4.10.164.111) has joined #freenet
[13:53] <toad_> Newsbyte: please don't be annoying for no good reason
[13:55] <iip_i2p> <Newsbyte> what? a testnetwork is no good reason?
[13:57] <toad_> well right now we have more immediate problems
[13:58] <toad_> e.g. how to minimize the impact of traffic padding
[13:58] <iip_i2p> <Newsbyte> is that a new problem?
[13:58] <toad_> and merging the CPU usage reduction via native code stuff to stable
[13:58] <toad_> Newsbyte: yes, as of yesterday
[13:59] <nicktastic> Surely there's a client or script that will ignore lines matching certain regexps...
[13:59] <toad_> we didn't have any until yesterday
[13:59] <toad_> traffic padding, that is
[13:59] <toad_> okay, fixing SendData will be difficult...
[13:59] <iip_i2p> <Newsbyte> with 8085, you mean?
[13:59] * toad_ works on removing RequestInterval instead...
[13:59] <toad_> 5085
[14:00] <iip_i2p> <Newsbyte> right
[14:01] <iip_i2p> <Newsbyte> nick: maybe so, but where to find them?
[14:01] <iip_i2p> <Newsbyte> so how comes traffic padding has become a bigger problem then it was before?
[14:02] <toad_> Newsbyte: because we weren't traffic padding at all until 5085
[14:02] <toad_> and now the issue is how to minimize the impact of it
[14:02] <toad_> 5085's traffic padding is pretty heavy duty
[14:03] <toad_> hmm, interesting... wildly varying but generally pretty good success rates on unstable...
[14:06] <iip_i2p> <Newsbyte> hmm.. I must misunderstand 'padding'
[14:06] <iip_i2p> <Newsbyte> padding: A thin, cushionlike mass of soft material used to fill, to give shape, or to protect against jarring, scraping, or other injury
[14:07] <iip_i2p> <Newsbyte> that seems of little relevance
[14:07] <toad_> Newsbyte: when we send a packet, we round its size up to the next multiple of 256 bytes, in order to prevent an attacker from establishing what kind of message(s) are contained (or worse!)
[14:07] <toad_> this was explained on my posts yesterday
[14:07] <toad_> to ALL THREE mailing lists
[14:07] <iip_i2p> <Newsbyte> well, for some reason, I seem to miss some posts
[14:08] <toad_> you mean you don't read them ? ;)
[14:08] * nicktastic laughs
[14:08] <iip_i2p> <Newsbyte> like, I never saw your response to me on the webthingy maillist
[14:08] <iip_i2p> <Newsbyte> no, they don't appear
[14:08] <nicktastic> Look harder, its there
[14:09] <iip_i2p> <Newsbyte> if it is, it's a) not in a thread under my post b)not with the same title
[14:09] <nicktastic> a) yes it is b) yes it is
[14:09] <toad_> webthingy?
[14:09] <toad_> oh, the archive?
[14:09] <nicktastic> Mailing list archives, I'd assume
[14:10] <iip_i2p> <Newsbyte> yes
[14:10] <iip_i2p> <Newsbyte> that's how I follow the discussions, since I'm not subscribed
[14:11] <iip_i2p> <Newsbyte> but I understand padding in IT terms now :-) It's basically the same system as zeroknowledge used
[14:12] <nicktastic> Try using the mailman archives instead of the gmane archives
[14:12] <iip_i2p> <Newsbyte> or even JAP, if I'm not mistaken
[14:12] <nicktastic> newsbyte, that is
[14:14] <iip_i2p> <Newsbyte> ermm...right...and the url would be....(?)
[14:14] <nicktastic> http://freenet.sourceforge.net/index.php?page=lists
[14:15] <nicktastic> Click on the list, click on Archives
[14:17] <iip_i2p> <Newsbyte> darn, since when did tthat work again?
[14:17] <toad_> Newsbyte: yeah, make all your packets the same size
[14:17] <toad_> it works?
[14:17] <toad_> I don't think I made it work...
[14:17] <iip_i2p> <Newsbyte> it used to be non-working
[14:17] <toad_> maybe ian did
[14:17] <iip_i2p> <Newsbyte> yeah, I see a lo of posts now
[14:17] <toad_> ah
[14:17] <toad_> the links go to the gmane archive :)
[14:18] <iip_i2p> <Newsbyte> well, yeah, but not directly klciking the arcives
[14:18] <iip_i2p> <Newsbyte> how comes the one shows it, but the other doesn't?
[14:19] <nicktastic> Who knows
[14:19] * tee2547 (~nobody@p50832813.dip0.t-ipconnect.de) has joined #freenet
[14:20] * emacsen (~serge@pcp09296048pcs.arlngt01.va.comcast.net) has joined #freenet
[14:20] <iip_i2p> <Newsbyte> wait a sec...even there I don't see my latest posts...only an old one
[14:22] <iip_i2p> <Newsbyte> 'compromise on testnetwork' is the latest I see of me
[14:24] <nicktastic> a) perhaps you should subscribe b) it may take time for messages to make it in to the archive
[14:24] * nicktastic &
[14:25] <iip_i2p> <Newsbyte> a)doubtfull b)it should appear on the archives too, that's what it's therefor, after all
[14:26] <iip_i2p> <Newsbyte> heh...k, nm, found it
[14:28] * emacsen (~serge@pcp09296048pcs.arlngt01.va.comcast.net) Quit ("Download Gaim: http://gaim.sourceforge.net/")
[14:30] <iip_i2p> <Newsbyte> well...at least the response of Ian, still can't see toads'
[14:30] <toad_> hmmm
[14:31] <toad_> okay... the most popular content source sends 699 packets (we have 35,246 of them)
[14:31] <toad_> i suppose it might not be 1.. maybe they just all have the same setting
[14:31] <nicktastic> Either you're expecting more than one response, or you're blind
[14:32] <toad_> the second most popular sends 161 byte packets
[14:32] <toad_> other popular sizes are 140, 21, 322, 447, 601
[14:33] * toad_ is going to check that the 699 length chunks come from different hosts
[14:33] <toad_> nicktastic: huh?
[14:33] <nicktastic> toad_: Not you
[14:34] <iip_i2p> <Newsbyte> /me is contemplating the possibility of blindness, but choses for the first option
[14:35] <iip_i2p> <Newsbyte> toad; are you looking for some standard chunks you can use?
[14:36] <toad_> Newsbyte: hmm?
[14:36] <toad_> what've you got?
[14:36] <iip_i2p> <Newsbyte> ?
[14:36] <iip_i2p> <Newsbyte> it was a gneric question, not an offer
[14:36] <toad_> i was looking for proof that the currennt system is a problem
[14:37] <iip_i2p> <Newsbyte> I meant, are you looking for some determined chunks (amount of bytes) so you can fasten the process
[14:37] <iip_i2p> <Newsbyte> I dunno: JAP uses the same system, and there it seems to work
[14:38] <toad_> yeah, it will
[14:38] <toad_> I need to fix the trailer chunk size too
[14:38] <toad_> currently trailer chunk size can probably be used as a node signature
[14:38] <toad_> although many nodes will share the same sig
[14:38] <toad_> but it's the sig of the DATA SOURCE NODE !
[14:39] <iip_i2p> <Newsbyte> I dunno: the smaller the better, but since it always involves some overhead, at a given point, the overhead outweighs the datatransport the chunk itself is capable of
[14:39] <toad_> well, I want one trailer chunk to fit with headers into one low level chunk
[14:40] <toad_> err padding chunk
[14:40] <iip_i2p> <Newsbyte> and a trailerchunk is 256?
[14:40] <toad_> yeah
[14:40] <toad_> soon it will probably be 128 though
[14:40] <toad_> well it can be multiples of that
[14:41] <toad_> hopefully most of the time it will be, but that's the minimum
[14:42] <iip_i2p> <Newsbyte> hmpf...it still means a message of 21 will be put in a chunk of 256...that's pretty much waste
[14:42] <toad_> or 128
[14:42] <toad_> but sadly at the moment there are hardly any 21 size messages
[14:42] <toad_> if messages were smaller we'd use smaller chunks
[14:42] <toad_> also, there's 24 bytes TCP overhead
[14:43] <toad_> and probably 16 bytes link overhead
[14:43] <toad_> PER PACKET
[14:43] <iip_i2p> <Newsbyte> I don't follow that. The problem is rather that they aren't biger (if the minimum chunk is going to be 128/256 anyway)
[14:43] <toad_> huh?
[14:44] <toad_> observe:
[14:44] <toad_> Raw messages:
[14:44] <toad_> Accepted ~~ 72-82 depends on RequestInterval
[14:44] <toad_> QueryRejected ~~ 135-136, depends on RequestInterval and Reason and HopsToLive
[14:44] <toad_> QueryRestarted ~~ 87-88, depends on RequestInterval
[14:44] <toad_> DataNotFound ~~ 99-104, depends on RequestInterval, TimeSinceQuery
[14:44] <toad_> DataReply ~~ 325-326
[14:44] <toad_> InsertRequest ~~ 147-156
[14:44] <toad_> StoreData ~~ 34kB
[14:44] <toad_> these are all including RequestInterval
[14:44] <toad_> which I'm in process of removing
[14:44] <toad_> and excluding headers, which add 3 bytes
[14:44] <toad_> (because the above are 1 byte off and headers are actually 4 bytes)
[14:47] <iip_i2p> <Newsbyte> ermm...well, guess it's because I'm no coder I understand badly, but in JAP, they fill the chunks that are to 'small' to fit an entire standard chunk up with garbage...what are you going to do? Add a bunch of little (usefull) chunks untill you have 128 or 256?
[14:47] <toad_> yeah, you pad it with garbage, if you need to
[14:47] <toad_> so if you have a 12 byte MRIPacketMessage, or a 6 byte trailerkillmessage, and nothing else to send, yes, you get wastage
[14:48] <iip_i2p> <Newsbyte> well, then, the problem are those chunks that are way too small to fill up a standard chunk
[14:48] <toad_> Newsbyte: our current messages aren't
[14:48] <toad_> see above
[14:49] <toad_> i'd use 128 bytes if it wasn't for DataReply and InsertRequest and QueryRejected
[14:49] <iip_i2p> <Newsbyte> those exceed the standardchunk
[14:49] <toad_> those exceed 128 bytes
[14:49] <iip_i2p> <Newsbyte> so why not cut them up like you do with the trailers
[14:50] <toad_> DataReply exceeds 256
[14:50] <toad_> that's not the point
[14:50] <toad_> we could use 1 byte padding chunks
[14:50] <toad_> then a DataReply would ALWAYS take up so many chunks that it's obvious what it is :)
[14:52] <iip_i2p> <Newsbyte> I don't see why...if you have a bunch of chunks, and two of them is for a datareply, then how is annyone going to know it IS that, and not something else?
[14:53] <toad_> Newsbyte: well, that's why it's not too disastrous that the padding size is 256
[14:53] <toad_> but I'd prefer that a DR fit within a padding chunk
[14:54] <toad_> if it was 8 bytes you could probably still tell the difference between different messages
[14:54] <toad_> if it was 4 bytes, you certainly could
[14:54] <toad_> if it was 16, then maybe
[14:54] <iip_i2p> <Newsbyte> well, you could go for 512, but what about the smaller chunks, then?
[14:54] <toad_> exactly
[14:54] <iip_i2p> <Newsbyte> maybe you should go for a set of standardchunks
[14:55] <toad_> eeek
[14:55] <iip_i2p> <Newsbyte> 64, 128, 256 and 512?
[14:55] <toad_> well at the moment it's 256, 512, 768, 1024, 1280 ...
[14:55] <toad_> going off to infinity, or at least to the MTU
[14:55] <iip_i2p> <Newsbyte> 1280??
[14:55] <toad_> 1024 + 256 = 1280
[14:56] <iip_i2p> <Newsbyte> yeah, but I mean; why? are there chunks that are that big?
[14:56] <toad_> i think so
[14:56] <toad_> if you have lots of messages queued, you'll use a bigger packet
[14:56] <toad_> even without padding, busy links are less vulnerable
[14:56] <toad_> because of that
[14:56] <iip_i2p> <Newsbyte> I thought big stuff, like movies, were cut up in smaller chunks...so what could possibly be 1280?
[14:57] <toad_> ummm, lots of separate messages bundled together into a single packet to save on overhead and minimize exposure
[14:58] <toad_> freenet's done that ever since we started muxing
[14:58] <iip_i2p> <Newsbyte> well, maybe one should only bundle them untill they reach a certain treshold?
[14:58] <toad_> sure
[14:58] <toad_> there is a packet size limitt
[14:58] <toad_> it depends on the bandwidth of the host
[14:58] <toad_> therein lies the problem, as it's also used to decide how big chunks to send - and they're then kept intact down the chain
[14:58] <toad_> trailer chunks i mean
[14:59] <iip_i2p> <Newsbyte> ok, so...what if you set a limit of max 512, also for bundling, and then you can have a far smaller set of standard chunks
[14:59] <toad_> why is it better to send two 512 byte packets than one 1024 byte packets?
[14:59] * toad_ -s
[15:00] <iip_i2p> <Newsbyte> well, for what we said earlier, to stop bw-waste when small packets have to be send in big chunks
[15:00] <toad_> ummm
[15:00] * tee2547 (~nobody@p50832813.dip0.t-ipconnect.de) Quit ("I'm going away")
[15:00] <toad_> we're mixed up over packets and messages here
[15:00] <toad_> messages are something like one trailer chunk, or an Accepted message
[15:00] <iip_i2p> <Newsbyte> within big standard chunks, I mean
[15:01] <toad_> packets are the blocks we send out to the network
[15:01] <toad_> packets are 256, 512, 768... max packet size
[15:01] <toad_> messages can be any size
[15:02] <iip_i2p> <Newsbyte> hmm...ok, but...messages can be limited to a certain size too, right?
[15:02] <toad_> hmm?
[15:02] <toad_> well FNP messages are limited to 64kB iirc
[15:03] <toad_> in practice StoreData is the biggest
[15:03] <toad_> that won't fit in the MTU so is normally sent in pieces
[15:03] <toad_> but it's probably obvious to an observer that a StoreData is being transferred
[15:04] <iip_i2p> <Newsbyte> ok, well...what if you did that for all messages?
[15:04] <toad_> did what?
[15:04] <toad_> we do
[15:04] <toad_> the stack won't let us send really big packets
[15:04] <toad_> if we try we end up sending a chunk then waiting for it to go
[15:04] <toad_> then sending another chunk
[15:04] <toad_> but we have no control over where it fragments it
[15:05] <iip_i2p> <Newsbyte> well, what's the problem then? If we make all messages in fairly little standard chunks, the same size as we use for trailers and the lot, then how is anyone going to determine what chunk is for what?
[15:07] <toad_> hmmm
[15:07] <toad_> I got a DNF... does that mean this code hasn't broken anything really badly?
[15:08] <toad_> Newsbyte: well as I showed above... if we see 5 chunks going into a node, and 5 chunks going out to the next node, regularly, then what are we going to conclude?
[15:08] <iip_i2p> <Newsbyte> is that a rethorical question? ;-)
[15:08] <toad_> I mean has it broken it worse than it was?
[15:09] <iip_i2p> <Newsbyte> dunno
[15:09] <iip_i2p> <Newsbyte> will a node in reality not send a flow of chunks, toad?
[15:10] <toad_> Newsbyte: thee point in the above scenario is that the trailer chunk size being sent is five times the padding size
[15:10] <iip_i2p> <Newsbyte> it continiously receives and sends stuff, the chunks will be flying around, I don't think it's that easy to determine a fixed pattern when there is a continious flow
[15:12] <toad_> okay, it's working.. i'm committing...
[15:12] <iip_i2p> <Newsbyte> so, let's say 5 times 128...when a node spits out dozens of chunks a minute, then the only thing remaining for making traffic analysis next to impossible, is to put some random time in the flowing of the chunks, or am I wrong?
[15:12] <toad_> Newsbyte: that was my assumption too until I discovered the recent data on traffic analysis
[15:13] <iip_i2p> <Newsbyte> and that data indicated...?
[15:13] <toad_> that on low traffic links you could see the message types and deduce some of their contents
[15:13] <toad_> especially if you have some collusion
[15:14] <iip_i2p> <Newsbyte> how comes?
[15:14] <toad_> (until I put padding in)
[15:16] <iip_i2p> <Newsbyte> and with padding, you think it's still possible to determine, because of fixed burst of chunks that are telltale signs of some messages?
[15:18] * toad_ hmm. SSKvsCHK asking for scaling support in fproxy so that you can use IFRAMEs and still scale the images... yuck
[15:18] <toad_> Terminating corrupt mux connection
[15:18] <toad_> hrrm
[15:19] <toad_> okay... Accepted is now 50 bytes inc headers
[15:20] <toad_> QueryRejected is 104
[15:20] <iip_i2p> <Newsbyte> add randomn send garbage chunks, then, that will solve it
[15:21] <toad_> QueryRestarted is 56
[15:22] <toad_> DNF is 70
[15:22] <toad_> InsertRequest is 122
[15:23] <toad_> of course DataReply and StoreData are still huge
[15:23] <iip_i2p> <Newsbyte> Connections open (Inbound/Outbound/Limit) 13 (0/13/200)
[15:23] <toad_> InsertRequest is in fact 126, soryr
[15:23] <iip_i2p> <Newsbyte> Transfers active (Transmitting/Receiving) 8 (3/5)
[15:23] <toad_> so I can probably cut the padding interval to 128 bytes...
[15:23] <iip_i2p> <Newsbyte> Data waiting to be transferred 445 Bytes
[15:23] <iip_i2p> <Newsbyte> Total amount of data transferred 22 MiB
[15:24] <iip_i2p> <Newsbyte> 22 in a few hours? time to shut my node down again ;-)
[15:24] <toad_> umm, 22MB is a big problem for you?!?
[15:26] <iip_i2p> <Newsbyte> personnaly, I have no problems with a tenfold of that
[15:26] <iip_i2p> <Newsbyte> if my b/w was unlimited ;-)
[15:26] <toad_> okay, that's irritating...
[15:27] <toad_> VoidPacketMessage takees up 4 bytes headers
[15:27] * Hory (~Miranda@81.196.25.110) Quit ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
[15:27] <toad_> so we can't set it to 128 bytes yet
[15:27] <toad_> we could set it to 130 though
[15:28] <iip_i2p> <Newsbyte> or you could try to reduce the 4 bytes
[15:28] <toad_> can't really, that's fixed
[15:29] <toad_> by the format
[15:29] <toad_> 2 bytes for the size, 2 bytes for the type
[15:29] <toad_> 126 + 12 = 138... that fits an InsertRequest and an MRIPacketMessage...
[15:30] <toad_> but a datareply would probably be 3 of them, which is undesirable...
[15:31] <toad_> ouch
[15:31] <toad_> DataReply for a KSK/SSK/SVK is 1300 bytes
[15:31] <toad_> DataReply for a CHK is ~ 300 bytes
[15:31] <iip_i2p> <Newsbyte> 22 in 2 hours, say 10 an hour, x24 is 240 a day is 7200 a month...you are right toad, that's doable, on a didicated machine. (alas, I'm runing/doingd/ling a lot of other stuff on this machine.
[15:32] <toad_> gzip only cuts it to 777 bytes...
[15:32] <toad_> that's doable on any machine without a fascist ISP
[15:32] <iip_i2p> <Newsbyte> depending on what else you were d/ling, no doubt? ;-)
[15:34] <iip_i2p> <Newsbyte> don't worry, toad, my new ISP is far better then the former. Maybe not fanfuckingtastic in a toads' viewpoint, but still twice as good as it used to be.
[15:34] <iip_i2p> <Newsbyte> with my former, 7200 would mean I would be at 2/3 of my limit already...
[15:35] <iip_i2p> <Newsbyte> ermm...3/4 rather
[15:36] <toad_> strange how ISPs vary across europe
[15:36] <toad_> download limits are pretty rare in the UK
[15:36] <toad_> at least the very-low-cost-cable provider doesn't limit...
[15:36] <iip_i2p> <Newsbyte> it's rather normal here
[15:36] <toad_> neither does AOL
[15:37] <iip_i2p> <Newsbyte> I actually don't think you find some that doesn't set a limit...unless very expensive ones
[15:37] <toad_> surprising really.. you'd expect germany to have good connectivity
[15:38] <iip_i2p> <Newsbyte> you'd expect any western country in europe to have good connectivity, but I don't think there lays the problem
[15:38] <iip_i2p> <Newsbyte> they just want to make money
[15:39] <toad_> well yeah but britain is renowned for its rampant capitalism at the cost of the consumer, the environment, and everyone else
[15:39] <toad_> and yet we have good llargely unlimited if a bit slow bandwidth
[15:39] <toad_> (you can't get uplink over 256kbps.. unless you pay a LOT more)
[15:40] <iip_i2p> <Newsbyte> well...i don't understand how they do it. If a dude uses 500 gig a month, isn't anyone going to react in the uk (isp)?
[15:40] <toad_> 500 gig a month iss posible?
[15:41] <iip_i2p> <Newsbyte> especially with cable, the b/w has to be shared with the poeple on the same local ring, so someone using exesive b/w would do so to the detrimence of the other users
[15:41] <toad_> 256kbps uplink = 85GB/mo up ... freenet node running both ways would be 170GB/mo, in theory...
[15:41] <iip_i2p> <Newsbyte> at least, that's what my former ISP gave as excuse ;-)
[15:41] <toad_> well, recently I've throttledd mine because of the KenManHack experiment
[15:41] <toad_> but I never had any problems
[15:42] <toad_> bbl
[15:43] <iip_i2p> <Newsbyte> 85GB/mo would be way over my max, even with this ISP ;-)
[15:43] <iip_i2p> <Newsbyte> that said, this one is not really controlling/enforcing strict limits, like my other
[16:14] <leexgx> probly not the right place but how do i find out what my proxy server is
[16:18] <greycat> leexgx: ask your network admin
[16:39] <leexgx> ?
[16:39] <leexgx> ask NTL
[16:39] <leexgx> be lucky if my ISP knows what an proxy server is
[16:39] <greycat> huh?
[16:39] <greycat> what makes you think you *have* a proxy server if you don't have a local network administrator?
[16:40] <leexgx> i know i do
[16:40] <greycat> how?
[16:40] <leexgx> if i could spell the world i want to say i could tell you
[16:40] <leexgx> grr
[16:40] <leexgx> trasparent proxy server
[16:40] <leexgx> i know i have one
[16:41] <greycat> if it's transparent then there's not a damned thing you can do about it. It's your default gateway, usually.
[16:44] * Tril (tril@bespin.org) has joined #freenet
[16:44] <leexgx> hmm
[16:44] <leexgx> i got an list any way
[16:47] * Tril (tril@bespin.org) has left #freenet
[16:48] <toad_> hmmm
[16:48] <leexgx> just for me emule
[16:48] <toad_> okay, how to fix SendData...
[16:48] <leexgx> testing this web proxy download
[16:49] <leexgx> seems like an good idea
[16:50] <toad_> read is sync
[16:50] <toad_> not async
[16:50] <toad_> so this is pretty easy
[16:51] <toad_> all I have to do is set the buffer size to 242 bytes...
[16:52] <toad_> and aggregate reads...
[16:55] <toad_> right
[16:55] <toad_> always read up to buffer size (or the length left in the file)
[16:55] <toad_> and make the buffer size small
[16:55] <toad_> and we're done
[16:55] <toad_> woah, as easy as that...
[16:57] * toad_ suspects getting binary presentation going won't be quite so easy
[16:57] <toad_> first I have to get in contact with thelema
[16:57] <toad_> and he's been away for aages
[16:58] <toad_> (2 days!)
[17:01] <toad_> hmmm
[17:01] <toad_> tricky one...
[17:02] <toad_> how to cloak the downstream chunk size while still allowing to send more than one chunk's worth of data at once?
[17:06] <toad_> hmmm
[17:07] <toad_> it's safest to only send one chunk per packet from each stream... because then they are hard to trace down.. but not only is it slower, the attacker knows that a packet containing lots of chunks must come from multiple sources...
[17:11] <leexgx> you get it working fine
[17:12] <toad_> huh?
[17:12] <leexgx> dono
[17:12] <leexgx> just intresting
[17:13] <leexgx> keep talking if you want
[17:13] <toad_> :)
[17:14] * toad_ restarts node with new changes..
[17:14] <leexgx> lol
[17:14] <toad_> hrrm?
[17:15] <leexgx> be intresting if some one made an p2p app for freenet
[17:15] <toad_> lol
[17:15] <leexgx> well when it gets faster
[17:15] <toad_> but what you might mean is Frost, http://jtcfrost.sf.net/
[17:15] <leexgx> or well more users
[17:15] <toad_> Freenet IS a P2P app
[17:15] <toad_> it's just not exactly a Napster clone
[17:15] <toad_> Frost is a little closer
[17:16] <toad_> although it's more like USENET
[17:16] <toad_> but it does have a searchable file library associated with each board
[17:16] <toad_> I suppose that's more like a BBS, really...
[17:17] * Iakin (Iakin@as25-4-6.kp.g.bonet.se) has joined #freenet
[17:17] <toad_> hi Iakin!
[17:17] <leexgx> any thing that goes up cant be pulled down tho
[17:17] <leexgx> so it say heh
[17:17] <leexgx> so posting stuff that some one want to stop cant do it
[17:17] <greycat> Murphy disagrees, though.
[17:17] <toad_> Iakin: do you think the new CPU usage reduction code is sufficiently mature to merge to stable?
[17:18] <greycat> Murphy's Law of Freenet: if you want it to stay in the network, it will fall out immediately. If you fucked up and want it to be eradicated from the face of the universe, it will live forever.
[17:18] <leexgx> lol
[17:18] <leexgx> that figers
[17:18] <leexgx> but more then latey that ant ture
[17:18] <leexgx> well for free sites any way
[17:19] <leexgx> i say for files yes
[17:19] <toad_> leexgx: you mean for files they never stay?
[17:20] <toad_> and for freesites they often stay?
[17:20] <toad_> IMHO frost's files don't stay because they were never inserted in the first place
[17:22] <leexgx> yes
[17:22] <Iakin3> toad: Yes.. rpobably
[17:23] <Iakin3> Has worked fine for me for quite a while
[17:23] <toad_> even with my recent changes?
[17:24] <leexgx> not tryed inserting an file in 6 months
[17:24] <leexgx> :)
[17:25] <Iakin3> toad: Well.. since yesterday
[17:25] <Iakin3> I am currently wrestling with this though:
[17:25] <Iakin3> Connections open (Inbound/Outbound/Limit) 0 (0/0/1000)
[17:25] <Iakin3> Uptime: 0 days, 9 hours, 27 minutes
[17:29] <toad_> Iakin3: strange...
[17:29] <toad_> works fine here...
[17:29] <toad_> the web interface is fully responsive?
[17:29] <toad_> what's the obvious cause? i mean are they all timing out?
[17:30] <toad_> ummm
[17:30] <toad_> i'm connected to iakin.poweruser.org:26828
[17:30] <toad_> is that a different node?
[17:31] <toad_> hmm
[17:31] <toad_> one of my peers just booted me
[17:39] <leexgx> resone
[17:42] <toad_> okay, /me cuts padding interval to 160 bytes
[17:42] <toad_> thus a DataReply easily fits in two chunks, and everything else (except StoreData) fits in one
[18:08] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) Quit (Remote closed the connection)
[18:18] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) has joined #freenet
[18:20] <toad_> hmmm
[18:20] <toad_> i think i've reproduced iakin's bug
[18:27] <leexgx> crash your pc
[18:27] <toad_> Iakin's bug was that his node couldn't open any connections...
[18:36] * toad_ starts up nodes with what is hopefully a fix...
[18:56] <toad_> freenet.AuthenticationFailedException: Against peer (null) @ 127.0.0.1:48111 - R
[18:56] <toad_> emote does not posess the private key to the public key it offered (terminal)
[18:56] * toad_ boggles
[19:00] <oierw> nice. your own computer is compromised
[19:09] * MikeW (Mike@159-134-233-242.as1.cbr.castlebar.eircom.net) has joined #freenet
[19:09] <MikeW> toad_!
[19:10] <MikeW> libjbigi :o)
[19:10] <MikeW> wrong clipboard
[19:10] <MikeW> "major vulnerability of Freenet to passive traffic analysis, relating to packet sizes."
[19:12] <oierw> yeah. you could tell what kind of message was being sent because the packet sizes weren't being kept at a uniform size
[19:13] <MikeW> yeah, I know, its a possible attack, but isn't very likely
[19:14] <oierw> agreed. but still a vulnerability
[19:16] <MikeW> and nice to see freenet using jbigi too :o)
[19:16] <toad_> okay
[19:16] <toad_> something's messed up with group C
[19:17] <toad_> MikeW: the scary thing about that was that packet sizes could be influenced by fixing the chain ID at the other end
[19:17] <toad_> but generally being able to find the types of messages is very bad
[19:17] <MikeW> nice that you could take away good ideas from I2P, perhaps stop by more often, heh
[19:18] <toad_> and it was a practical attack, I did it myself
[19:18] <toad_> on my own node
[19:18] <toad_> ;)
[19:18] <MikeW> you mean being able to get the type of the data?
[19:18] <toad_> maybe someday I2P will take our http filter
[19:18] <MikeW> or figuring our who sent it?
[19:18] <MikeW> chain = tunnel?
[19:21] <toad_> MikeW: chain ID
[19:21] <toad_> roughly a tunnel yeah
[19:21] <toad_> but for a request, before the data is got
[19:21] <toad_> there are nasty issues with packet size on trailing field transfers too
[19:22] <MikeW> good that these issues came up so early
[19:22] <MikeW> I shall try freenet again sometime soon, if I can ever get it working
[19:23] <MikeW> never played nicely on 56k with winsock screwed by sp2. anyway, reformat this weekend, I'll try an unstable then
[19:23] <toad_> hmmm
[19:23] <toad_> this is crazy
[19:23] <toad_> group C signatures DO NOT VERIFY
[19:24] <toad_> but we've been using group C for years and they did
[19:24] <toad_> they don't verify ... EVEN IF WE TURN OFF NATIVE BIGINT SUPPORT !
[19:26] <oierw> toad_: go back a couple of days and see if they still verify
[19:26] <toad_> hmmm
[19:26] <toad_> my fault...
[19:26] <toad_> now it works, even with native
[19:26] <toad_> wieerf
[19:26] <toad_> dd
[19:27] <toad_> they verify... they always verify...
[19:27] <toad_> but they don't in situ
[19:27] <toad_> WTF?
[19:28] <toad_> must be signing the wrong thing...
[19:29] <toad_> but has any of this code actually changed recently??
[19:29] <toad_> maybe a problem in tcpconn?
[19:29] <oierw> could it have something to do with breaking up the packets?
[19:29] * oierw is just thinking outloud. I doubt that could be the problem
[19:39] * Ash-Foxeus is now known as Ash-Fox
[19:48] <toad_> hmmm ok i have some leads
[19:48] <toad_> but i'm not there yet
[19:48] <toad_> and I'm not going to work on it any more until monday
[19:48] <toad_> hahahahahahaha
[19:52] <oierw> argh!
[19:52] <oierw> :)
[19:54] <iip_i2p> <lucky> i move for toad to be named the greatest briton ever!
[19:54] <toad_> hahahahahahahahahahahahahahahahahahahahahahahahahahaha !
[19:55] <toad_> lucky: eh?
[19:55] <toad_> not working on sundays is not a british national thing
[19:55] <MikeW> not a british national party thing?
[19:55] * toad_ (toad@82-32-16-91.cable.ubr03.azte.blueyonder.co.uk) Quit (Remote closed the connection)
[19:57] <leexgx> not working sat an sunday is my thing here
[19:58] <MikeW> oooh, he didn't like that!
[20:22] * Ash-Fox (Hal-9000@aap55.neoplus.adsl.tpnet.pl) Quit (Remote closed the connection)
[20:26] * Ash-Fox (Hal-9000@aap55.neoplus.adsl.tpnet.pl) has joined #FREENET
[20:27] * jmercier (~jmercier@4.10.164.111) Quit (Read error: 110 (Connection timed out))
[21:25] * MikeW (Mike@159-134-233-242.as1.cbr.castlebar.eircom.net) Quit ()
[22:04] * Pascal is now known as Pascal666
[23:10] * proteusguy (proteusguy@66-90-229-20.dyn.grandenetworks.net) has joined #freenet
These logs were automatically created by Jay Oliveri with his gimp hapi on irc.freenode.net.