Timestamps are in GMT/BST.
[0:05] <toad_> nextgens: AVAILABILITY
[0:16] * hjubal (n=hjubal@) Quit ("Leaving")
[0:27] * colione (n=colione@) Quit ("This computer has gone to sleep")
[0:37] * timmy2chk (n=Violent@) has joined #freenet
[0:42] * s14318572 (n=chatzill@) Quit ("Chatzilla 0.9.76 [Firefox 2.0/2006101023]")
[0:46] * colione (n=colione@) has joined #freenet
[0:52] * NullAcht15 (n=NullAcht@) Quit (Remote closed the connection)
[1:48] * toad_ (n=toad@) Quit (Remote closed the connection)
[1:58] * sanity (n=ian@) Quit ()
[2:09] * timmy2chk (n=Violent@) Quit ("what remains of my violence")
[2:48] * Xosy (n=newbie@) has joined #freenet
[2:48] <Xosy> what does HTL stand for, in freenet?
[2:49] * sanity (n=ian@) has joined #freenet
[2:49] * ChanServ sets mode +o sanity
[2:51] <Xosy> what does HTL stand for, in freenet?
[2:57] * Xosy (n=newbie@) Quit ()
[2:59] * protok0l (n=sodkghks@) has joined #freenet
[2:59] * protok0l (n=sodkghks@) has left #freenet
[2:59] * ferrox is now known as ferrox__
[3:00] * ferrox__ is now known as ferrox
[3:15] * phrosty (n=phrosty@) has joined #freenet
[3:21] * Geoff_Whittingto (n=chatzill@) has joined #freenet
[3:22] * MikeW (n=em@) Quit ()
[3:42] * dermoth (n=dermoth@) has joined #freenet
[4:09] * dystopia (n=ph33r@) Quit ()
[4:39] * jblack (n=jblack@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * OctobersDark (n=October@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * maaxel (i=polaris@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * Harik (i=harik@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * fridim (n=fridim@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * maelztrom (i=nobody@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * Tritman (i=Tritman@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * Geoff_Whittingto (n=chatzill@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * JustMe (i=JustMe_@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * Bombe (n=droden@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * Boefje (i=1234@) Quit (sterling.freenode.net irc.freenode.net)
[4:39] * WildZeck (n=zeck@) Quit (sterling.freenode.net irc.freenode.net)
[4:40] * Boefje (i=1234@) has joined #freenet
[4:42] * Bombe (n=droden@) has joined #freenet
[4:44] * Geoff_Whittingto (n=chatzill@) has joined #freenet
[4:44] * JustMe (i=JustMe_@) has joined #freenet
[4:44] * Harik (i=harik@) has joined #freenet
[4:44] * maaxel (i=polaris@) has joined #freenet
[4:44] * jblack (n=jblack@) has joined #freenet
[4:44] * OctobersDark (n=October@) has joined #freenet
[4:44] * fridim (n=fridim@) has joined #freenet
[4:44] * maelztrom (i=nobody@) has joined #freenet
[4:44] * Tritman (i=Tritman@) has joined #freenet
[4:44] * WildZeck (n=zeck@) has joined #freenet
[4:44] -ChanServ- You do not have channel operator access to [#freenet]
[4:44] * ChanServ sets mode -o FreenetLogBot
[4:48] * Geoff_Whittingto (n=chatzill@) Quit (Read error: 110 (Connection timed out))
[4:51] [freenode-connect VERSION]
[5:13] * Geoff_Whittingto (n=chatzill@) has joined #freenet
[5:17] * Guest214 (n=Guest214@) has joined #freenet
[5:37] * Iceman_B (n=Ice@) Quit ("Fate-chan is hotter than Nanoha-chan ^_^")
[5:41] * Iceman_B (n=Ice@) has joined #freenet
[5:56] * Geoff_Whittingto (n=chatzill@) Quit ("Chatzilla 0.9.76 [Firefox 2.0/2006101023]")
[6:19] * awatake (n=bob@) has joined #freenet
[6:22] <awatake> Hi all. I just had a severe hard drive failure and lost my node data. So I have started "littlenodie1" back up. I had about 30 peers in my list...
[6:23] <awatake> if any of you have littlenode1 in your reference list, please delete it and add this new one: http://code.bulix.org/wujh66-23601?raw
[6:23] <awatake> thank you. :)
[6:28] * nextgens sets mode +o fridim
[6:28] * nextgens sets mode -o fridim
[6:28] * nextgens sets mode +o FreenetLogBot
[7:08] * awatake (n=bob@) Quit ("Leaving")
[7:29] * __themoog__ (i=OK_Compu@) Quit ("Leaving")
[7:39] * JustMe (i=JustMe_@) Quit ()
[8:16] * sbc (n=sbc@) has joined #freenet
[8:18] * _ph00 (n=z@) Quit ("Leaving")
[8:19] * _ph00 (n=z@) has joined #freenet
[8:19] * _ph00 (n=z@) Quit (Client Quit)
[8:39] * whiterabbit (n=Whiterab@) has joined #freenet
[8:45] * sbc (n=sbc@) Quit ("Ex-Chat")
[9:08] * sandos_ (n=sandos@) has joined #freenet
[9:10] * sanity (n=ian@) Quit ()
[9:10] * IMCensored1 (n=KMIntern@) has joined #freenet
[9:18] * sandos (n=sandos@) Quit (Read error: 110 (Connection timed out))
[9:28] * anonymouse (n=a@) Quit ()
[9:30] * CommanderSky (n=Nelio_To@) has joined #freenet
[9:35] * JFlesch_ (n=jflesch@) has joined #freenet
[9:40] * JFlesch__ (n=jflesch@) has joined #freenet
[9:44] * JFlesch_ (n=jflesch@) Quit (Read error: 60 (Operation timed out))
[9:47] * JFlesch (n=jflesch@) Quit (Read error: 110 (Connection timed out))
[10:11] * JFlesch (n=jflesch@) has joined #freenet
[10:16] * JFlesch (n=jflesch@) Quit (Remote closed the connection)
[10:16] * JFlesch (n=jflesch@) has joined #freenet
[10:16] * JFlesch__ (n=jflesch@) Quit (Read error: 60 (Operation timed out))
[10:19] * toad_ (n=toad@) has joined #freenet
[10:19] * ChanServ sets mode +o toad_
[10:22] * JFlesch_ (n=jflesch@) has joined #freenet
[10:26] * JFlesch (n=jflesch@) Quit (Read error: 60 (Operation timed out))
[10:38] * phrosty (n=phrosty@) Quit (Read error: 104 (Connection reset by peer))
[10:38] * phrosty (n=phrosty@) has joined #freenet
[10:38] * CommanderSky (n=Nelio_To@) Quit ("Client exiting")
[10:38] * JFlesch (n=jflesch@) has joined #freenet
[10:44] * JFlesch (n=jflesch@) Quit (Remote closed the connection)
[10:44] * JFlesch (n=jflesch@) has joined #freenet
[10:47] * JFlesch_ (n=jflesch@) Quit (Read error: 110 (Connection timed out))
[10:49] * JFlesch (n=jflesch@) Quit (Remote closed the connection)
[10:49] * JFlesch (n=jflesch@) has joined #freenet
[10:55] * JFlesch (n=jflesch@) Quit (Remote closed the connection)
[10:56] * JFlesch (n=jflesch@) has joined #freenet
[10:58] * docgonzo (n=Chatuser@) has joined #freenet
[10:58] * docgonzo (n=Chatuser@) has left #freenet
[11:14] * phrosty (n=phrosty@) Quit (Read error: 104 (Connection reset by peer))
[11:15] * phrosty (n=phrosty@) has joined #freenet
[11:17] * subhuman_ (n=subhuman@) has joined #freenet
[11:39] * subhuman_ (n=subhuman@) has left #freenet
[11:41] * railk (n=railk@) has joined #freenet
[11:43] * oCc (n=nodeOne@) has joined #freenet
[11:52] * oCc (n=nodeOne@) Quit (Read error: 104 (Connection reset by peer))
[12:01] * railk (n=railk@) Quit (Remote closed the connection)
[12:14] * railk (n=railk@) has joined #freenet
[12:19] <nextgens> hi
[12:20] <toad_> nextgens: hi
[12:20] <toad_> nextgens: haven't finished the updater changes yet, but they should go in today
[12:21] <toad_> after that i have some changes to block transfer (back compatible; remove unnecessary pings) and the unmatched packets tracker (keep them for an hour only)
[12:22] <toad_> and binary blobs sometime soon
[12:25] <nextgens> ok
[12:25] * nextgens will resume his work on FEC for x86_64 then :)
[12:25] <toad_> and after that, the containers rewrite
[12:26] <nextgens> no update over mandatory while you're at binary blobs ?
[12:26] <toad_> and another 5000 bugfixes and minor features :)
[12:26] <nextgens> :)
[12:26] <toad_> nextgens: yeah, of course, that's the main reason for blobs
[12:26] * toad_ wonders about binary blobs format ... it should be easy to combine two binary blob files ...
[12:27] * nextgens logs onto servalan
[12:28] <toad_> bbiab
[12:29] * railk (n=railk@) Quit (Remote closed the connection)
[12:31] * Guest214 (n=Guest214@) Quit ("PJIRC forever!")
[12:32] * oCc (n=nodeOne@) has joined #freenet
[12:32] * oCc (n=nodeOne@) Quit (Remote closed the connection)
[12:33] * oCc (n=nodeOne@) has joined #freenet
[12:34] * oCc (n=nodeOne@) Quit (Client Quit)
[12:34] * oCc (n=nodeOne@) has joined #freenet
[12:52] * whiterabbit (n=Whiterab@) Quit ("KVIrc 3.2.5 Anomalies http://www.kvirc.net/")
[13:06] * CIA-12 (i=cia@) Quit (Client Quit)
[13:11] * railk (n=railk@) has joined #freenet
[13:11] * geeku is now known as geekU
[13:14] * nextgens reboots
[13:14] * CIA-14 (i=cia@) has joined #freenet
[13:20] * railk (n=railk@) Quit (Remote closed the connection)
[13:21] <nextgens> we seems to have less than 1access per second to the datastore
[13:21] * dermoth (n=dermoth@) Quit (Read error: 104 (Connection reset by peer))
[13:21] * dermoth_ (n=dermoth@) has joined #freenet
[13:21] <nextgens> that's the value I get here at least ...
[13:21] <nextgens> does anyone else get an higher one ?
[13:22] <nextgens> a significantly higher one I mean
[13:25] * greycat (i=rfc1413@) has joined #freenet
[13:33] * rguerra (n=rguerra@) has joined #freenet
[13:35] * JFlesch (n=jflesch@) Quit (Read error: 54 (Connection reset by peer))
[13:35] * nextgens investigates the segfault
[13:38] * CIA-14 (i=cia@) Quit (Client Quit)
[13:40] <nextgens> :<
[13:40] <nextgens> it works with gij but not with sun's jvm
[13:41] * JFlesch (n=jflesch@) has joined #freenet
[13:43] * sbc (n=sbc@) has joined #freenet
[13:43] <nextgens> hmmm
[13:44] <nextgens> I've got problems understanding how that can possibly happen
[13:46] * JFlesch (n=jflesch@) Quit (Remote closed the connection)
[13:49] * CIA-14 (i=cia@) has joined #freenet
[13:51] * JFlesch (n=jflesch@) has joined #freenet
[13:54] * OctobersDark_ (n=October@) has joined #freenet
[13:56] <toad_> nextgens: that's bizarre
[13:56] * JFlesch_ (n=jflesch@) has joined #freenet
[13:56] * JFlesch (n=jflesch@) Quit (Remote closed the connection)
[13:56] * OctobersDark (n=October@) Quit (Nick collision from services.)
[13:57] <toad_> nextgens: what segfault?
[13:57] * OctobersDark_ is now known as OctobersDark
[13:57] * CIA-14 (i=cia@) Quit (Client Quit)
[13:57] * OctobersDark_ (n=October@) has joined #freenet
[13:58] * OctobersDark_ (n=October@) Quit (Client Quit)
[13:58] * CIA-14 (i=cia@) has joined #freenet
[14:02] * CIA-14 (i=cia@) Quit (Client Quit)
[14:02] * JFlesch (n=jflesch@) has joined #freenet
[14:05] * CIA-14 (i=cia@) has joined #freenet
[14:05] * JFlesch_ (n=jflesch@) Quit (Read error: 104 (Connection reset by peer))
[14:05] * JFlesch (n=jflesch@) Quit (Read error: 54 (Connection reset by peer))
[14:06] <nextgens> toad_> on the native FEC code for x86_64
[14:06] <nextgens> the C test runs fine
[14:06] <nextgens> the java one with JNI works only on gij
[14:08] * CIA-14 (i=cia@) Quit (Client Quit)
[14:08] * JFlesch (n=jflesch@) has joined #freenet
[14:09] * CIA-14 (i=cia@) has joined #freenet
[14:10] <toad_> nextgens: JNI definitely works with gij? it's not falling back to the pure java version?
[14:10] <nextgens> well, then it's as fast as the C code ;)
[14:11] <nextgens> toad_> I suggest you logon servalan ... do su - nextgens
[14:11] <nextgens> screen -rx
[14:11] <nextgens> I'll show you if you want
[14:11] <toad_> nextgens@servalan:~$ screen -rx
[14:11] <toad_> Cannot open your terminal '/dev/pts/6' - please check.
[14:12] <nextgens> hmm
[14:12] <nextgens> ssh nextgens@servalan then
[14:12] * JFlesch (n=jflesch@) Quit (Read error: 54 (Connection reset by peer))
[14:12] <nextgens> ensure you haven't forgotten the - while 'suing'
[14:13] <nextgens> hmm
[14:13] <nextgens> I've fixed it :))
[14:13] * toad_ is watching
[14:13] <nextgens> YAY
[14:13] <nextgens> now it works on both
[14:14] <nextgens> and the gij version is indeed slow
[14:14] <toad_> what was the bug?
[14:14] <nextgens> a missing cast
[14:14] <nextgens> the c version seems to be a bit faster than jni
[14:14] <nextgens> but I guess that's normal
[14:14] <toad_> why is the gij so much slower?
[14:15] <nextgens> I don't understand see :
[14:15] <toad_> hmm, so the java+jni is 50% slower than the c version?
[14:15] <nextgens> with the c binaries we achieve ~=300MB/s
[14:15] <nextgens> with sun's java : ~=150-200MB/s
[14:15] <nextgens> pure java code ~3MB/s
[14:15] <nextgens> but :
[14:16] <toad_> ah
[14:16] <nextgens> that one is f*cking fast
[14:16] <toad_> 300MB/sec with gij
[14:16] <nextgens> but segfaults on java's sun
[14:16] <nextgens> (using the old jar)
[14:16] <toad_> ah
[14:16] <toad_> what was the difference?
[14:16] <nextgens> dunno
[14:17] <nextgens> well the difference in between both is that one has been misscompilated
[14:17] <nextgens> well, it's segfaulting
[14:17] <nextgens> whereas the other one works
[14:17] <toad_> what code changed?
[14:17] <nextgens> the JNI headers
[14:18] <nextgens> well, we were inputing a jint ...
[14:18] * JFlesch (n=jflesch@) has joined #freenet
[14:18] <nextgens> wich is 4 bits
[14:18] <toad_> 64 bits?
[14:18] <nextgens> and the code was casting it wrongly to a 8 bit void *
[14:18] <toad_> hmmm
[14:18] <nextgens> I've just casted it the right way ... so that it cleans up the remaining bytes
[14:19] <toad_> ok
[14:19] <nextgens> of course that's not "uberoptimized" but at least it works
[14:19] <toad_> is it possible to do memory/cpu tradeoffs with the amount of data we pass in at once?
[14:19] <nextgens> and according to the simple benchs I've done it worths it
[14:20] <nextgens> I don't get what you mean
[14:20] <toad_> we pass an array into the native code, right?
[14:20] <nextgens> yes
[14:20] <toad_> how big?
[14:20] <nextgens> it depends
[14:21] * nextgens hasn't checked how we use it in our code
[14:21] <toad_> how big in the benchmark?
[14:22] <nextgens> we use 1024 bytes it seems
[14:22] * nextgens checks the java test
[14:22] <toad_> what is packetSize passed in as?
[14:23] <toad_> well, at present, in the main code, we use stripe size of 4096 bytes
[14:23] <nextgens> we can addapt benchs if it's what you want
[14:23] <toad_> so we have an array of 192x 4096 bytes, basically
[14:24] <nextgens> ok
[14:24] * nextgens will do the same on the bench
[14:24] <toad_> probably be faster
[14:24] <toad_> 192
[14:25] <toad_> that's 192/128
[14:25] <toad_> k=192, n=128
[14:25] <nextgens> in fact in the bench we change n all the time iirc
[14:25] <nextgens> we start with a high value and then decrease it
[14:25] <toad_> well, that's the _maximum_ we use
[14:25] <toad_> often we use less than that
[14:26] * nextgens is looking for where we set n
[14:29] <nextgens> hmm
[14:29] <nextgens> are we using the 16 or the 8 bit version ?
[14:29] <nextgens> might be auto-detected in the loader
[14:29] <toad_> 8-bit version
[14:29] * wastekan (n=blah@) has joined #freenet
[14:29] <toad_> if k<256 and n<256, the 8-bit version is used
[14:30] <toad_> otherwise the 16-bit version is used
[14:30] <nextgens> in fact we never use the 16 bit, right ?
[14:31] <toad_> yep, that is correct
[14:31] <toad_> because it is 4x slower than the 8-bit
[14:31] <nextgens> why keeping it then ? :)
[14:31] <nextgens> ok, so on average the JNI calls are twice slower than the C version
[14:31] <toad_> well, we need to decide whether we want to maintain the onion code and keep as much as possible in case we need to use it later...
[14:31] <nextgens> the tests I do is :
[14:31] <toad_> or whether we want to just strip out everything we don't use
[14:32] * nextgens doesn't explain why it's twice slower
[14:32] <nextgens> toad_> there is an other option : don't bundle things we don't use but maintain them
[14:33] * oCc (n=nodeOne@) has left #freenet
[14:34] * nextgens wonders wether he should commit what he did or not
[14:34] <nextgens> it works ...
[14:34] <toad_> nextgens: so this is using 192/128?
[14:34] <toad_> with 4k?
[14:34] <nextgens> but I'm not proud of it
[14:35] <nextgens> yep
[14:35] <toad_> hmmm
[14:35] <toad_> well, does 8k make it go faster?
[14:35] <toad_> or 16k?
[14:35] <nextgens> hmm, no, it uses 192/64
[14:35] * nextgens changes
[14:36] <toad_> what's l?
[14:37] <nextgens> the number of passes it seems
[14:39] * CIA-14 (i=cia@) Quit (Client Quit)
[14:41] <toad_> you sure you're comparing like with like?
[14:41] <toad_> i mean, does the java code have the same parameters as the c code?
[14:42] <nextgens> almost sure
[14:43] <toad_> then it's very surprising that they get radically different results
[14:43] * agsarite (i=agsarite@) has joined #freenet
[14:44] <nextgens> indeed
[14:44] <nextgens> hence I pointed it out
[14:45] * nextgens ought to run somme benchs on his 32 bit computer
[14:45] * CIA-14 (i=cia@) has joined #freenet
[14:47] <toad_> well, if it's simply a matter of JNI overhead, then increasing the size of the data passed in/out should wipe it out
[14:47] <toad_> if that doesn't then there's something else going on
[14:53] <nextgens> anyway, now I'm ready to release a new freenet-ext.jar
[14:54] <nextgens> toad_> I would need your input, using a real node
[14:54] <nextgens> please grab freenet-ext in ~nextgens/contrib/freenet_ext/freenet-ext.jar
[14:54] <nextgens> and tell me if it works
[14:55] <nextgens> btw, I've also managed to recompile all the wrapper libraries
[14:55] * nextgens wonders if we should stick to "official", released ones or if we should recompile our own
[14:56] <toad_> hmmm
[14:56] * toad_ is in the middle of coding some mods ...
[14:56] <nextgens> ok, don't mess up with it then :)
[14:56] * nextgens doesn't want to disturb
[14:56] <toad_> ok
[14:56] <toad_> you can test it yourself though
[14:56] <nextgens> true
[14:59] <toad_> i can make a diff and try the new jar on my node if you want
[14:59] <toad_> that might be the easiest way to do it
[15:00] * nextgens uploads it on emu
[15:01] <nextgens> http://emu.freenetproject.org/~nextgens/ext.jar
[15:02] * JFlesch_ (n=jflesch@) has joined #freenet
[15:03] * Jase (i=jase@) Quit (Remote closed the connection)
[15:04] * Zothar_Work (n=chatzill@) has joined #freenet
[15:07] <toad_> got the new jar
[15:08] <toad_> starting up...
[15:09] * nextgens hasn't updated the version information yet
[15:09] <toad_> http://code.bulix.org/0rk9b6-23624
[15:10] * toad_ still getting the native-FEC-doesn't-load error
[15:10] <nextgens> what's the sha1sum of the file you have ?
[15:10] <toad_> 0e826990a98549efd91630a2843a0f95085b7849
[15:10] <nextgens> hmm
[15:11] <toad_> is that the right file?
[15:11] <nextgens> that ought to be the good one
[15:11] <nextgens> INFO | jvm 1 | 2006/11/16 15:09:04 | at com.onionnetworks.util.NativeDeployer.findLibraries(NativeDeployer.java:143)
[15:11] * nextgens investigates
[15:12] * JFlesch (n=jflesch@) Quit (Read error: 110 (Connection timed out))
[15:12] <nextgens> if (p.getProperty("com.onionnetworks.native."+key+".osarch").
[15:12] <nextgens> trim().equals(OS_ARCH)) {
[15:13] <nextgens> that's the NPE
[15:13] <nextgens> ok, it's trivial
[15:13] <toad_> ok, tell me when
[15:13] * toad_ blech, 2 new child porn sites on indicia
[15:14] * toad_ shouldn't call attention to it
[15:15] * toad_ should defect to the RIAA and help eliminate it ;)
[15:16] <nextgens> hmm
[15:16] * nextgens doesn't understand
[15:17] <toad_> nextgens: anything i can help with?
[15:17] <nextgens> yes, maybe
[15:17] <nextgens> it's a problem with the parser
[15:17] <nextgens> com/onionnetworks/util/NativeDeployer.java
[15:17] <nextgens> it parses a file called native.properties
[15:19] <nextgens> and in your case we get an NPE while trying to compare Properties
[15:19] <nextgens> it's stringified version with an other string we generated in fact
[15:21] <nextgens> it's in fec-common : /home/nextgens/contrib/fec/common
[15:26] <toad_> nextgens: is a key.osarch actually missing?
[15:27] <nextgens> I don't think so
[15:27] <toad_> where can i see the properties file?
[15:27] <nextgens> /home/nextgens/contrib/fec/common/lib/native.properties
[15:27] <nextgens> or do "jar xvf /home/nextgens/freenet-ext.jar lib/native.properties"
[15:27] <nextgens> so that it will extract it
[15:28] <toad_> there's one missing
[15:28] <toad_> it says fec8-win32 not fec8-win32-x86
[15:28] <toad_> it should be the latter
[15:28] <nextgens> doh
[15:28] <toad_> same with fec16
[15:28] <toad_> also change the osarch lines to -x86 if that's the arch you get (which i think it is now)
[15:29] <nextgens> hold on, I'll regenerate a new file
[15:36] * nextgens recompiles
[15:36] <toad_> nextgens: new jar?
[15:36] <nextgens> I've just commited
[15:36] <nextgens> CIA-14> c'mon!
[15:36] <toad_> can i have a new testing jar please?
[15:36] <nextgens> sure
[15:37] <nextgens> mrogers has commited :)
[15:38] <toad_> hrrm?
[15:38] * nextgens has just received 1659 N Nov 16 mrogers@freenet ( 378) [freenet-cvs] r10943 - in trunk/apps/load-balancing-sims/phase
[15:39] <nextgens> emu seems to be overloaded :/
[15:40] <nextgens> DOH, I forgot the to add junit in my classpath :<
[15:40] * toad_ has 10945
[15:40] <toad_> a few minutes ago
[15:41] <toad_> nextgens: we may have some work to do on Librarian soon
[15:41] <nextgens> [exec] /tmp/user/1009/ccPT8B1q.o(.text+0x239): In function `wrapperStartPendingSignalled':
[15:41] * toad_ thinks that would be a good thing
[15:41] <nextgens> [exec] : undefined reference to `ceil'
[15:41] <nextgens> [exec] /tmp/user/1009/ccPT8B1q.o(.text+0x353): In function `wrapperStopPendingSignalled':
[15:41] <nextgens> [exec] : undefined reference to `ceil'
[15:41] <nextgens> [exec] /tmp/user/1009/ccPT8B1q.o(.text+0x6b44): In function `wrapperProtocolRead':
[15:41] <nextgens> [exec] : undefined reference to `ceil'
[15:41] <nextgens> [exec] collect2: ld returned 1 exit status
[15:41] <nextgens> [exec] make: *** [wrapper] Error 1
[15:41] <nextgens> sorry
[15:42] <toad_> nextgens: Indicia will hopefully be producing an index soon, which means we'll need a way to link to it easily from a freesite, and also need a better format (probably some variant on the thaw/ninja format)
[15:43] * toad_ thinks being able to style its output would be quite useful too
[15:44] <toad_> brb
[15:44] * nextgens is still trying to get emu to build that damn jar file
[15:45] * MikeW (n=em@) has joined #freenet
[15:48] <nextgens> http://emu.freenetproject.org/~nextgens/ext.jar
[15:48] <nextgens> toad_> it's available there
[15:48] <toad_> yay
[15:49] <toad_> 491f86391e04332451764b3c97bac680212e8f4d ?
[15:49] <nextgens> 491f86391e04332451764b3c97bac680212e8f4d public_html/ext.jar
[15:49] <nextgens> yep
[15:49] * nextgens awaits flames
[15:50] <toad_> crashes and burns
[15:50] <toad_> SEGV
[15:50] <toad_> INFO | jvm 1 | 2006/11/16 15:50:17 | # Problematic frame:
[15:50] <toad_> INFO | jvm 1 | 2006/11/16 15:50:17 | # C [libfec8.so+0x16eb] fec_decode+0x22
[15:50] <nextgens> :'(
[15:51] <nextgens> at least now it loads the library :D
[15:51] <toad_> would you like the dump?
[15:51] <nextgens> yes please
[15:51] <nextgens> the hs_*log file
[15:51] <toad_> in your ~
[15:52] <nextgens> ok, thanks
[15:53] * Gasi (n=chatzill@) has joined #freenet
[15:53] <toad_> nextgens: http://pastebin.ca/248539 - any idea?
[15:53] <toad_> nextgens: not a transient mail problem either
[15:54] <toad_> happened at least twice on two different occasions
[15:54] <toad_> every time sandos tries to send me a mail
[15:54] <nextgens> smtp.bredband2.net has a broken dns
[15:54] <toad_> so it seems...
[15:54] <nextgens> host -t MX amphibian.dyndns.org
[15:54] <nextgens> amphibian.dyndns.org has no MX record
[15:54] <nextgens> that's probably why
[15:54] <toad_> aha
[15:55] <toad_> i should add amphibian.dyndns.org as its own primary MX?
[15:55] <nextgens> some MTAs might require it
[15:55] <nextgens> but afaik you can't using dyndns
[15:55] <nextgens> it's a premium option or something like that
[15:55] <nextgens> isn't it ?
[15:56] <toad_> "Mail Exchanger (optional):"
[15:56] <toad_> i just put amphibian.dyndns.org in that box?
[15:56] <nextgens> yes
[15:56] <toad_> it has a checkbox for Backup MX also
[15:56] <nextgens> then tick backup mx only
[15:57] <toad_> we could set up emu as a backup mx ...
[15:57] <toad_> nextgens: hmm?
[15:57] <nextgens> and if you set a MX by name, ensure you add a dot at the end
[15:57] <nextgens> sure I can set emu to be a backup for your domain
[15:57] <toad_> nextgens: why do i need to set backup mx if it's not the backup, it's the primary?
[15:57] <nextgens> do you want me to ?
[15:57] <nextgens> 'cause qmail is broken ? :p
[15:57] <toad_> nextgens: might be nice, when you get around to it tell me
[15:58] <nextgens> it's a matter of seconds
[15:58] * nextgens does it
[15:58] <toad_> so i should set mail exchanger: amphibian.dyndns.org. and check the backup mx button
[15:59] <nextgens> MAIL FROM: <test@test.com>
[15:59] <nextgens> 250 Ok
[15:59] <nextgens> RCPT TO: <toad@amphibian.dyndns.org>
[15:59] <nextgens> 250 Ok
[15:59] <nextgens> ok, done
[15:59] <nextgens> toad_> you can set emu as your backup, it will accept your mails now
[16:00] <toad_> so i add amphibian.dyndns.org as MX, then emu.freenetproject.org as backup
[16:00] <nextgens> yep
[16:00] <toad_> nextgens: with or without a trailing dot?
[16:00] <nextgens> but check the syntax : it might be with a trailing dot
[16:00] <nextgens> dunno
[16:00] <nextgens> usually there is one
[16:00] <nextgens> but depending on the interface they may do it on their own
[16:01] <toad_> ok, this is odd
[16:01] <toad_> it only shows one MX
[16:01] <toad_> and it's either backup MX or it isn't
[16:02] <toad_> hmmm ok it's sorted
[16:02] <toad_> amphibian.dyndns.org A 82.32.17.1
[16:02] <toad_> amphibian.dyndns.org MX 10 emu.freenetproject.org
[16:02] <toad_> amphibian.dyndns.org MX 5 amphibian.dyndns.org
[16:03] <nextgens> fine
[16:03] <toad_> nextgens: any progress on the segfault?
[16:03] <nextgens> I'm investigating
[16:03] <nextgens> but so far, no
[16:03] <nextgens> I've got no idea of what's going on
[16:04] * nextgens will build a jar from servalan
[16:05] * wastekan (n=blah@) has left #freenet
[16:18] * toad_ thinks somebody should explain anonymous remailers to AnMaster
[16:20] <nextgens> toad_> and you know the best ? it works on gij !
[16:20] * sanity (n=ian@) has joined #freenet
[16:20] * ChanServ sets mode +o sanity
[16:20] <toad_> nextgens: works as in really works?
[16:20] <nextgens> toad_> try to reach fproxy on port 7777
[16:21] * nextgens managed to get indicia
[16:21] <toad_> nextgens: make it available from the network please?
[16:21] <nextgens> whereas with sun's jvm it segfaulted
[16:21] <toad_> cool
[16:21] <nextgens> hmm, sure
[16:21] <toad_> but we can't exactly require people to run freenet under gcj/gij
[16:21] <toad_> 192.168.1.0/24
[16:21] <toad_> or at least 192.168.1.4
[16:22] <nextgens> toad_> done
[16:22] <nextgens> I got an exception from bdb stuffs
[16:22] <nextgens> but well, something not fatal imo
[16:23] <nextgens> java.lang.NullPointerException
[16:23] <nextgens> at com.sleepycat.je.txn.TxnManager.registerTxn(TxnManager.java:147)
[16:23] <nextgens> at com.sleepycat.je.txn.Txn.init(Txn.java:220)
[16:23] <nextgens> at com.sleepycat.je.txn.Txn.<init>(Txn.java:156)
[16:23] <nextgens> at com.sleepycat.je.txn.TxnManager.txnBegin(TxnManager.java:126)
[16:25] <toad_> nextgens: so the node works with gij, but not with sun, with the new ext.jar.. that's wierd
[16:25] <toad_> working with gij is good
[16:25] <toad_> nextgens: that could be bad
[16:26] <nextgens> I'm not 100% sure that everything is working
[16:26] <nextgens> I got another exception
[16:26] <nextgens> Exception in thread "freenet.clients.http.SimpleToadletServer$SocketHandler@4fb8f482" java.lang.NullPointerException
[16:26] <nextgens> at freenet.support.io.TempFileBucket.deleteFile(TempFileBucket.java:285)
[16:26] <nextgens> at freenet.support.io.FileBucket.free(FileBucket.java:337)
[16:26] <nextgens> at freenet.support.io.FileBucket.free(FileBucket.java:330)
[16:26] <nextgens> but well, it seems to work not too bad
[16:26] <nextgens> you have shut it down ?
[16:27] <toad_> i can't get it to connect to mine
[16:27] <toad_> no
[16:27] <toad_> it seems to have crashed
[16:27] * nextgens restarts it
[16:27] <toad_> all i did was try to load indicia... :|
[16:28] <nextgens> toad_> that's normal it doesn't connect to yours
[16:28] <nextgens> you need to set the AllowLocal or something
[16:28] * richard223 (i=joe@) has joined #freenet
[16:28] <toad_> i'm surprised it doesn't just connect over 82.blah
[16:28] <toad_> but yes, i did that
[16:29] <richard223> hmmm I installed windows' IPv6 stack and freenet won't connect to any peers
[16:29] <toad_> well i re-added, with local IPs, and was about to set allowlocal when it crashed
[16:30] <richard223> is there anything else I need to configure?
[16:30] <toad_> richard223: strange
[16:30] <nextgens> yes, a real OS ;p
[16:30] <toad_> richard223: you're sure IPv6 is working?
[16:30] <toad_> richard223: and IPv4 sites?
[16:31] <richard223> yes, "I can see the dancing kame on www.kame.net"
[16:31] <richard223> ipv4 also works
[16:31] <richard223> but freenet doesn't connect to either ipv4 or ipv6
[16:31] <nextgens> Goodbye. from freenet.node.Node@1eb7d323 (PacketSender deadlock)
[16:32] * sanity (n=ian@) Quit ()
[16:32] <toad_> Beta_M: the plugin system more or less works now
[16:32] <toad_> Beta_M: but there are two of them
[16:32] <Beta_M> which one should i use
[16:32] <Beta_M> and where do i find the specs
[16:33] <toad_> Librarian uses the freenet.pluginmanager api
[16:33] * agsarite (i=agsarite@) Quit (Remote closed the connection)
[16:33] <toad_> so does jSTUN
[16:33] <toad_> so probably you should use that API
[16:33] <richard223> hmmm... freenet detects my IPv6 adresses
[16:33] <toad_> i dunno if we have any docs for it
[16:33] <toad_> anything on the wiki?
[16:33] <toad_> richard223: does it pick up your ipv4 address too?
[16:34] <richard223> I have set a DNS name
[16:34] <richard223> but I usually see my ipv4 address at the end of the physical.udp line
[16:34] <toad_> is that included in your ref? (don't say physical dot udp or you'll get autokicked :) )
[16:34] <richard223> now I don't see it
[16:35] <richard223> yes, the DNS name is included on that line
[16:35] <richard223> netstat -a shows 0.0.0.0:32178 (my udp port)
[16:36] <richard223> it also shows [::]:32178
[16:36] <richard223> so I think that means it's listening on ipv4 and ipv6
[16:36] <Zothar_Work> richard223: yeah, that means it's listening; tcpdump/ethereal/wireshark show packets coming in from peers or going out to peers?
[16:37] <richard223> ok let me load up ethereal.
[16:39] <richard223> yes, there's traffic to and from my IP to port 32178
[16:40] <toad_> hmmm
[16:40] * JFlesch_ (n=jflesch@) Quit (Read error: 54 (Connection reset by peer))
[16:40] <toad_> if a node without IPv6 is connected to an IPv6 node, does it throw when it tries to send an IPv6 packet, and so not get to send an IPv4 one? where do we catch IOException's on that path?
[16:40] <richard223> java version "1.5.0_09" :S
[16:41] <richard223> nov 16, 2006 16:41:53:031 (freenet.io.comm.UdpSocketManager, UdpSocketManager sender thread on port 32178, ERROR): Caught java.lang.IllegalStateException: Cannot send auth packet: too long: 1402 from freenet.node.FNPPacketMangler@c623af
[16:42] <toad_> WTF?
[16:42] <toad_> that's bad
[16:42] <toad_> nextgens: before you say it, the old limit was 1400
[16:43] <nextgens> ;)
[16:43] <toad_> richard223: you don't have a really long node name or anything?
[16:43] <richard223> nov 16, 2006 16:43:46:421 (freenet.node.PeerNode, UdpSocketManager sender thread on port 32178, NORMAL): Completed handshake with freenet.node.PeerNode@c73e5ed4@80.5.100.137:55439@9b417dc4c894cc6e9346bf0fbeac046c5ba47e6e772839c93d8d269d000884ad on 80.5.100.137:55439 - current: null old: freenet.node.KeyTracker@da1515 for freenet.node.PeerNode@c73e5ed4@80.5.100.137:55439@9b417dc4c894cc6e9346bf0fbeac046c5ba47e6e772839c93d8d269d000884ad unv
[16:43] <richard223> my node name is 12 characters long
[16:43] <nextgens> hmm
[16:43] <nextgens> it's only the AUTH packet
[16:44] <nextgens> wich can't be fragmented iirc
[16:44] <nextgens> hence we have to use a sensible MTU
[16:44] <nextgens> or let the OS handle it
[16:44] <toad_> auth can be fragged
[16:44] <toad_> well
[16:44] <toad_> it can't be split up _by us_
[16:44] <richard223> ah ha! I found it
[16:45] * JFlesch (n=jflesch@) has joined #freenet
[16:45] <richard223> "ipv6 if" shows: enlace MTU 1280 (enlace MTU 65515)
[16:45] <toad_> richard223: we don't set dont-fragment
[16:45] <toad_> richard223: also IPv6 doesn't support fragmentation anyway...
[16:45] <toad_> but the above error was because our internal arbitrary MTU limit was too low
[16:45] <toad_> i've just fixed that
[16:45] <nextgens> toad_> he is probably using ipv6overipv4
[16:46] <nextgens> toad_> a tunnel broker
[16:46] <richard223> yes, I'm using a tunnel broker
[16:46] <toad_> nextgens: how can ipv6 have an MTU if doesn't support fragmentation anyway?
[16:46] <toad_> i suppose it can't?
[16:46] <nextgens> there is no mtu
[16:46] <nextgens> but windows ifaces might ;)
[16:47] <toad_> well richard223 can try the next build
[16:47] <toad_> 10948 i mean
[16:47] <toad_> (what's up with CIA?)
[16:47] <nextgens> dunno
[16:47] <nextgens> they have been argueing with FBI and NSA all the day long :D
[16:47] * nextgens (n=nextgens@) has left #freenet
[16:47] * nextgens (n=nextgens@) has joined #freenet
[16:47] * ChanServ sets mode +o nextgens
[16:47] <toad_> nextgens: any leads on the crash, anyway?
[16:48] <nextgens> no :'(
[16:48] <nextgens> I'm still on it
[16:48] <nextgens> even the test doesn't pass now :<
[16:50] * JFlesch (n=jflesch@) Quit (Remote closed the connection)
[16:50] * JFlesch (n=jflesch@) has joined #freenet
[16:53] <richard223> I did an "ipv6 uninstall", all my peers are back now
[16:53] <toad_> richard223: try with 10948
[16:54] <richard223> erm... :) how do I update?
[16:55] <toad_> get the new jar
[16:55] * JFlesch (n=jflesch@) Quit (Remote closed the connection)
[16:55] <toad_> from http://downloads.freenetproject.org/
[16:55] <richard223> freenet-latest?
[16:55] <nextgens> nah
[16:55] <nextgens> freenet-testing-latest.jar
[16:55] * JFlesch (n=jflesch@) has joined #freenet
[16:56] <richard223> oh yes I just saw it :D
[16:58] <richard223> well got it. stupid windows needs a reboot
[16:58] <richard223> to reinstall ipv6
[16:58] <richard223> brb
[17:00] * richard223 (i=joe@) Quit ()
[17:00] * mozillaman (n=borg@) has joined #freenet
[17:00] * JFlesch_ (n=jflesch@) has joined #freenet
[17:02] * hjubal (n=hjubal@) has joined #freenet
[17:06] * JFlesch_ (n=jflesch@) Quit (Remote closed the connection)
[17:06] * JFlesch_ (n=jflesch@) has joined #freenet
[17:11] * JFlesch (n=jflesch@) Quit (Read error: 110 (Connection timed out))
[17:12] * richard223 (i=joe@) has joined #freenet
[17:12] <richard223> OK, ipv4 nodes work now
[17:13] <richard223> gonna try an ipv6 node on my lan
[17:14] <nextgens> http://www.thisiscool.com/gcc_mingw.htm
[17:14] <richard223> damn it's hot down here
[17:14] <richard223> 38? C
[17:21] <toad_> richard223: :|
[17:22] <nextgens> toad_> I might have found it
[17:22] <toad_> nextgens: hmm?
[17:23] <nextgens> why it's segfaulting
[17:23] * nextgens testes a patch
[17:26] * Jase (i=jase@) has joined #freenet
[17:28] <richard223> wtf? the other node doesn't accept the new ref
[17:29] <richard223> the one with the ipv6 addresses
[17:30] * anonymouse (n=a@) has joined #freenet
[17:34] <toad_> nextgens: progress?
[17:37] <nextgens> I'm about to give up for tonight
[17:37] <toad_> ok
[17:37] <nextgens> nothing meaningfull
[17:38] * CIA-14 (i=cia@) Quit (Client Quit)
[17:40] <sandos_> auto-update 996 failed, and now 997 has been going on for a long time aswell....
[17:41] <toad_> 998 fixed the auto-updating
[17:41] <toad_> sandos_: i think i fixed the email problem, try resending that email again please?
[17:42] <richard223> wtf, still doesn't accept my new ref
[17:42] <sandos_> " 'cause qmail is broken ? :p" - haha, good thing I seem to have managed to make my manager _not_ reinstall our mail server from postfix to qmail then :)
[17:42] <sandos_> yep, I can try
[17:43] * CIA-14 (i=cia@) has joined #freenet
[17:43] <richard223> could it be because my phys udp line is too long?
[17:45] <richard223> 605 bytes long :S
[17:47] <richard223> oops... not connecting to anything again.
[17:47] <sandos_> toad_, how about my suggestion about more verbose auto-update?
[17:47] <sandos_> I dont know why they keep failing
[17:47] <sandos_> they do so every now and then...
[17:47] <sandos_> otoh, its been working flawlessly for a number of weeks now
[17:47] <richard223> Cannot send auth packet: too long: 1505 from freenet.node.FNPPacketMangler@dcb03bjava.lang.IllegalStateException: Cannot send auth packet: too long: 1505
[17:51] <Zothar_Work> heh, auth packets that long could be a problem... :)
[17:54] <richard223> hmmm too many network interfaces on this machine
[17:55] <richard223> damn, freenet is binding to 19 network addresses
[17:55] <richard223> most of these fe80:0:0: ...
[17:58] <Zothar_Work> richard223: if the IP is static, you could use bindTo to specify, though I don't know if it takes multiple IPs, so I dunno if you could have both an IPv4 and an IPv6 address in the node's bindTo for the FNP port
[17:58] <sandos_> what are the initial reqests that very often get cached? ARKs or something?
[17:58] <richard223> nope, it's a dynamic IP address
[17:58] <sandos_> my initial (first 60 seconds) cache hit-ratio is always very high
[18:05] * _admin (n=admin@) has joined #freenet
[18:05] <richard223> well, I just had to remove ipv6
[18:11] * JFlesch (n=jflesch@) has joined #freenet
[18:11] * _admin is now known as Frosty
[18:11] <Zothar_Work> sandos: I'd guess they are ARKs, yes
[18:12] * Frosty is now known as Frostyfcp
[18:13] * tubbie (n=tubbie@) has joined #freenet
[18:17] * richard223 (i=joe@) Quit ()
[18:17] * JFlesch__ (n=jflesch@) has joined #freenet
[18:18] * gybbf (n=simon_ka@) has joined #freenet
[18:22] * JFlesch_ (n=jflesch@) Quit (Read error: 110 (Connection timed out))
[18:22] <Frostyfcp> Is anybody trading node refs over on the refs channel?
[18:22] * JFlesch__ (n=jflesch@) Quit (Remote closed the connection)
[18:23] * Jflesch_ (n=jflesch@) has joined #freenet
[18:24] * gybbf (n=simon_ka@) Quit ()
[18:24] * iwantnode (n=iwantnod@) has joined #freenet
[18:24] * Zothar_Work (n=chatzill@) Quit (Remote closed the connection)
[18:26] * Iceman_B (n=Ice@) Quit (Read error: 104 (Connection reset by peer))
[18:27] * hjubal (n=hjubal@) Quit ("Leaving")
[18:28] * JFlesch (n=jflesch@) Quit (Read error: 110 (Connection timed out))
[18:28] * Jflesch_ (n=jflesch@) Quit (Remote closed the connection)
[18:34] * Frostyfcp (n=admin@) has left #freenet
[18:34] * CIA-14 (i=cia@) Quit (Client Quit)
[18:35] * Jflesch (n=jflesch@) has joined #freenet
[18:38] * sanity (n=ian@) has joined #freenet
[18:38] * ChanServ sets mode +o sanity
[18:38] * NullAcht15 (n=NullAcht@) has joined #freenet
[18:40] * CIA-14 (i=cia@) has joined #freenet
[18:45] * gybbf (n=simon_ka@) has joined #freenet
[18:50] * lionking (n=chatzill@) has joined #freenet
[18:53] * gybbf (n=simon_ka@) has left #freenet
[18:59] * lionking (n=chatzill@) Quit ("Chatzilla 0.9.76 [Firefox 2.0/2006101023]")
[18:59] * Zothar_Work (n=chatzill@) has joined #freenet
[19:02] * sich (n=sich@) has joined #freenet
[19:02] <sich> hell
[19:02] <sich> o
[19:04] <Zothar_Work> olleh
[19:06] * NullAcht15 (n=NullAcht@) Quit (Nick collision from services.)
[19:06] * NullAcht15_ (n=NullAcht@) has joined #freenet
[19:08] * _ph00 (n=z@) has joined #freenet
[19:18] <nextgens> toad_> where have you picked 10-15 from ?
[19:19] <nextgens> toad_> I've got half of it here and it performs fine
[19:19] <nextgens> do we really want people to have that many peers ?
[19:20] <toad_> nextgens: does it?
[19:21] <nextgens> yes it does
[19:21] <toad_> nextgens: hmmm ok
[19:21] <toad_> well change it however you like
[19:21] <toad_> more peers _is_ better in general, although non-darknet peers is bad
[19:21] <toad_> and ubernodes are bad
[19:22] <Zothar_Work> I've been telling people 10-15, usually on #freenet-refs where probably makes more sense as it increases the likelihood that they pick up an "old timer" peer instead of only other newbies currently on the channel
[19:23] <nextgens> there is another point to consider :
[19:23] <nextgens> the default b/w limit is 15K
[19:23] <nextgens> and most people won't change it ;)
[19:25] * nextgens reads http://wiki.freenetproject.org/NewPacketFormat
[19:25] <nextgens> I don't get why we want to have three different keys
[19:25] <nextgens> maybe I missunderstand it but I don't see the point of the "IV" one
[19:26] <nextgens> why not using a PRNG ?
[19:27] <nextgens> what about switching to HMAC while we are at changing low level stuffs ?
[19:28] * nextgens thinks we are reinventing the wheel there
[19:32] <nextgens> bbiab
[19:35] * sbc (n=sbc@) Quit ("Ex-Chat")
[19:36] <toad_> nextgens: what do you think a block cipher is if not a PRNG? :)
[19:36] <toad_> nextgens: and it does include switching to HMAC
[19:37] * CIA-14 (i=cia@) Quit (Client Quit)
[19:39] * CIA-14 (i=cia@) has joined #freenet
[19:49] * whiterabbit (n=Whiterab@) has joined #freenet
[20:00] * awatake (n=bob@) has joined #freenet
[20:06] * mozillaman (n=borg@) Quit ("Chatzilla 0.9.72-rdmsoft [XULRunner 1.8.0.1/2006012608]")
[20:13] * sich_ (n=sich@) has joined #freenet
[20:13] * sich (n=sich@) Quit (Connection reset by peer)
[20:15] * sich_ is now known as sich
[20:16] * whiterabbit (n=Whiterab@) Quit ("KVIrc 3.2.5 Anomalies http://www.kvirc.net/")
[20:19] * whiterabbit (n=Whiterab@) has joined #freenet
[20:33] * iwantnode (n=iwantnod@) Quit ()
[20:44] * NoJohnDoe (n=chatzill@) has joined #freenet
[20:50] * CIA-14 (i=cia@) Quit (Client Quit)
[20:50] * tubbie (n=tubbie@) Quit ()
[20:54] * MineHaunter (n=van85@) Quit ("brb")
[20:56] * MineHaunter (n=van85@) has joined #freenet
[20:57] * CIA-14 (i=cia@) has joined #freenet
[21:07] * whiterabbit (n=Whiterab@) Quit ("KVIrc 3.2.5 Anomalies http://www.kvirc.net/")
[21:09] * mozillaman (n=chatzill@) has joined #freenet
[21:09] * NoJohnDoe (n=chatzill@) Quit ("Chatzilla 0.9.75 [Firefox 1.5/2005111116]")
[21:33] * Zothar_Work (n=chatzill@) Quit (Remote closed the connection)
[21:34] <nextgens> toad_> have you seen CHK@9Y-PX1qIfoZX39pn2DOYbRy8lXFEWL7AKmzrFZIPoEw,tPqOCTgqLJ1dFPwQr1VL2vb5viGmdzHO0KniDM-q39E,AAEC--8?type=text/plain on frost ?
[21:36] * sich (n=sich@) Quit ("Quitte")
[21:37] * JustMe (i=JustMe_@) has joined #freenet
[21:38] * greycat (i=rfc1413@) Quit ("This time the bullet cold rocked ya / A yellow ribbon instead of a swastika")
[21:43] * CIA-14 (i=cia@) Quit (Client Quit)
[21:46] <toad_> nextgens: fetching...
[21:54] * nextgens is addicted to CIA
[21:54] <nextgens> when it doesn't work, I miss it
[21:56] <nextgens> toad_> I'll review it and commit it anyway ... it's a patch from uniqueperson, replacing some missuses of " by '
[21:57] <toad_> nextgens: good idea
[21:57] * pepe_1bm (n=pepe_1bm@) has joined #freenet
[21:57] * awatake (n=bob@) Quit ("Leaving")
[21:59] * Iceman_B (n=Ice@) has joined #freenet
[22:02] <nextgens> toad_> regarding DSA: it's really used code
[22:02] <nextgens> toad_> http://en.wikipedia.org/wiki/Digital_Signature_Algorithm
[22:03] <nextgens> they don't say it needs to be cryptographically secure
[22:03] <nextgens> they don't say it shouldn't either ...
[22:03] <nextgens> but in that instance : the special case where the generated random leads to producing an invalid value ...
[22:03] <nextgens> it doesn't matter much
[22:04] <nextgens> btw, I dunno where we use the signing code
[22:04] <nextgens> I sign node references with it ...
[22:04] <nextgens> will use it for StS
[22:04] <toad_> nextgens: hmmm
[22:05] <toad_> it's used for lots of signatures yes
[22:06] <toad_> nextgens: so it's only used in the wierd case?
[22:07] <nextgens> yes
[22:07] <nextgens> that specific call to SecureRandom at least
[22:08] <nextgens> when I implemented it, I asked you and Scott for feedback about it
[22:08] <nextgens> our DSA code is following the first version of the FIPS draft
[22:08] <nextgens> I tweaked it to follow the second
[22:08] <nextgens> or at least part of the second ...
[22:09] <toad_> nextgens: you use an ordinary SecureRandom for STS too?
[22:09] <nextgens> and got stuck, awaiting for a reply from you or scott regarding whether it was worthy to implement FIPS-3...
[22:09] <nextgens> toad_> no, I seed from yarrow iirc
[22:10] <toad_> ok
[22:10] <toad_> using Yarrow is a good idea anyway
[22:10] <nextgens> and I've never done it as it implied modifying the DSAGroups
[22:10] <toad_> nextgens: what's the difference?
[22:10] <toad_> nextgens: oh, the extra checks?
[22:10] <nextgens> yes
[22:10] <toad_> nextgens: bug scott
[22:11] <toad_> nextgens: if you don't get a response, then bug me again
[22:11] <toad_> is that blocking STS?
[22:11] <nextgens> I'll do when I get around StS
[22:11] <nextgens> not really
[22:11] <nextgens> but I don't like building a castle with weakened bricks ;)
[22:12] * Zothar_Work (n=chatzill@) has joined #freenet
[22:12] <nextgens> btw I don't think that's really important (maybe I'm wrong though)
[22:12] <nextgens> I bet freenet won't ever be attacked directly on the crypto.
[22:14] <nextgens> '\"'
[22:14] <nextgens> can't we use '"' in java ?
[22:14] <toad_> dunno
[22:15] <_ph00> I was thinking about some possible legal problem:
[22:15] <_ph00> in a country where data encryption is completely banned, would freenet be automatically outlawd?
[22:15] <nextgens> toad_> I think we already have a method to get a ticker
[22:16] <nextgens> toad_> something static somewhere iirc
[22:16] <nextgens> _ph00> the answer is obvious : yes
[22:16] <toad_> nextgens: it shouldn't be static
[22:16] <_ph00> or, in a country where you must decrypt or you get a sentence, and you can't decrypt your store of course?
[22:16] <_ph00> how's that there?
[22:17] <toad_> _ph00: if they want to execute you they'll find a reason :)
[22:17] <nextgens> in case you want to have several nodes on the same jvm ?
[22:17] <toad_> _ph00: very few countries outlaw encryption completely
[22:17] <toad_> nextgens: yes
[22:17] <toad_> _ph00: because it's required for e-commerce
[22:17] <_ph00> hm
[22:17] <_ph00> fair enough
[22:17] <toad_> of course, SOME countries do effectively outlaw the internet - cuba, iran, ...
[22:18] <_ph00> what about the must-decrypt thing? still no trial about that in UK?
[22:18] <toad_> cuba, north korea it's limited to the elite, and then it's whitelisted
[22:18] <toad_> iran you have to use a cybercafe
[22:18] <toad_> which are logged
[22:18] <toad_> _ph00: i dunno
[22:18] * CIA-14 (i=cia@) has joined #freenet
[22:18] <_ph00> which leads to the question: how to reach those people?
[22:18] <toad_> _ph00: haven't heard anything..
[22:18] <_ph00> cuba iran etc
[22:18] <toad_> _ph00: good question
[22:19] <toad_> well, it's not whitelisted in cuba
[22:19] <toad_> but it is severely limited
[22:19] <_ph00> once they install freenet and get some peers running shouldn't be a problem, but the question is how to reach them in the first place, make them aware of freenet
[22:19] <nextgens> +headers.put("Content-Disposition", "attachment; filename=\"" + key.getPreferredFilename() + '\"');
[22:19] <nextgens> is it the same as :
[22:19] <toad_> _ph00: maybe
[22:20] <_ph00> we may need some agents to travel there (very risky)
[22:20] <nextgens> +headers.put("Content-Disposition", "attachment; filename=\"" + key.getPreferredFilename() + '"'); ?
[22:20] * nextgens thinks that it is
[22:20] <toad_> _ph00: that's always a problem in any hostile regime
[22:20] <toad_> _ph00: and people do
[22:20] <toad_> _ph00: our job is to provide them with infrastructure that will decrease rather than increase the chance of folk getting killed
[22:21] * toad_ not sure that's possible mind you
[22:21] * MikeW (n=em@) Quit ()
[22:21] <_ph00> with the current darknet setup, I guess that knowing how to set freenet up, getting the software, and a bunch of e-mail addresses (if they can use encryption) maybe they could get connected
[22:21] <_ph00> OK for the decrease killings, not increase them
[22:22] <toad_> :)
[22:22] <_ph00> but I was thinking about those who would run freenet despite their "bad" govt
[22:22] <toad_> the problem is freenet will always be attackable if you spend enough money on it
[22:22] <_ph00> IO think it's our "duty" to try to reach those ppl
[22:22] <_ph00> I*
[22:23] <_ph00> will it get more difficultly attackable if it get bigger?
[22:23] <toad_> well, china's a more realistic goal
[22:23] <_ph00> china is a very important goal
[22:23] <toad_> a lot more people using the internet, for a start
[22:23] <_ph00> 1.5 billion ppl...
[22:23] * nextgens has reviewed half the patch and start getting bored
[22:24] <nextgens> he should make smaller patches
[22:24] <nextgens> :p
[22:24] <toad_> and china censors the internet, but they allow it to exist in the general population (private broadband) because they recognize its economic importance
[22:24] <_ph00> I guess someone in cuba, iran etc would run freenet if they had a chance
[22:24] <toad_> _ph00: i've talked to an iranian here, he said they'd managed to block tor somehow
[22:24] <toad_> that was before they abolished private broadband
[22:25] <_ph00> one problem with china is that there are lots of ppl who only speak chinese (i.e. wee need a chinese webpage, chinese wiki etc)
[22:25] <toad_> banning private broadband is a very serious step, it will cost iran billions of dollars in the long term
[22:25] <_ph00> woah
[22:25] <toad_> _ph00: yes, we need translations
[22:25] <_ph00> they managed to block tor?!? I didn't even know that was theoretically possible
[22:25] <toad_> it's on the todo list, but not before the beta
[22:25] <toad_> _ph00: yes, it's not hard
[22:25] <toad_> tor is trivially harvestable
[22:25] * nextgens has a lot of potential traductors aviable here
[22:26] <toad_> yes, we have folk lined up
[22:26] <toad_> but we need the infrastructure put in first
[22:26] <nextgens> including for chineese, japoneese, korean
[22:26] <toad_> and that's not a priority until the beta
[22:26] <nextgens> agreed
[22:26] <_ph00> to french, yes, and other EU languages wouldn't be so difficult to find translators for
[22:26] <_ph00> ah OK
[22:26] <_ph00> including korean chinese etc
[22:26] <_ph00> cool
[22:26] <toad_> well, if you find any, add their email addresses to the translation bug
[22:27] * toad_ goes to investigate dinner
[22:30] * Gasi (n=chatzill@) Quit ("Chatzilla 0.9.67+ [SeaMonkey 1.0.6/2006103003]")
[22:31] * pepe_1bm (n=pepe_1bm@) Quit ()
[22:34] <nextgens> is a regexp wizard around ?
[22:35] <nextgens> I'd like to match lines starting with a +, containing '\"' to replace them with '"'
[22:35] <nextgens> ^\+(.*)'\\"' would that work ?
[22:45] <Zothar_Work> hmm.. that's not the type of regexp I'm most familiar with, but let me munch my brain on that a sec
[22:47] <Zothar_Work> you want to replace '\"' with something and my understanding is that you'll want something like this for your matching part
[22:47] <Zothar_Work> : ^\+.*(\\")
[22:47] <Zothar_Work> bach
[22:47] <Zothar_Work> bah
[22:47] <Zothar_Work> ^\+.*(\\")
[22:48] <Zothar_Work> then how you replace what's in the () depends on the language's functions I think (I've done regexp for replacements very rarely; perhaps this is a case where string replace is faster/easier?
[22:48] <Zothar_Work> nextgens: sound about right?
[22:50] <nextgens> I've found what I need
[22:50] * nextgens is commiting
[22:50] <nextgens> thanks anyway
[22:51] <nextgens> ok they went in
[22:52] <nextgens> Message body is too big: 213584 bytes with a limit of 40 KB
[22:52] <nextgens> ^^-^^
[22:52] <nextgens> emu is gonna spend the whole night recompiling now :/
[22:52] <nextgens> and we are going to get flooded :|
[22:53] <nextgens> at least it allows me to test whether my auto-building process works and scalles :p
[22:53] <nextgens> -l
[22:54] <nextgens> and the main file ?
[22:54] <nextgens> cool
[22:58] * Muixirt (n=trixium@) has joined #freenet
[23:03] <nextgens> whaaa$
[23:03] <nextgens> I don't know if it's a false impression or not but some pages are loading far faster now
[23:03] <nextgens> things like /config/
[23:05] <nextgens> toad_> my nodeAveragePingTime: 236ms is much higher that it used to be
[23:05] <nextgens> toad_> but the input/output rate is almost the same ...
[23:05] <nextgens> toad_> I bet that's related to MTU changes
[23:06] * anonymouse (n=a@) Quit (Read error: 54 (Connection reset by peer))
[23:06] * anonymouse (n=a@) has joined #freenet
[23:07] <Zothar_Work> nextgens: perhaps if Subversion is the only thing that send email to freenet-cvs, that list's limit could be raised higher than the default 40 KB?
[23:07] <Zothar_Work> sends
[23:08] * IMCensored1 (n=KMIntern@) Quit (Read error: 110 (Connection timed out))
[23:09] <nextgens> sometimes we don't want the full diff to go through the list though
[23:09] <nextgens> when would we introduce new versions of our backdoor otherwise ? :p
[23:10] <nextgens> more seriously, sometimes it's better to send a short explanation of what have been done
[23:11] <nextgens> and to give people willing to read it the opportunity to do so, giving them a pointer to a diff
[23:11] <Zothar_Work> :)
[23:11] <nextgens> http://emu.freenetproject.org/cgi-bin/viewcvs.cgi?view=rev&rev=10961
[23:14] <Zothar_Work> most of r10962 wasn't too hard to audit for real changes
[23:15] <nextgens> they are only trivial changes
[23:16] <nextgens> but well, sending a 200KB email to everyone is probably stupid
[23:16] <nextgens> most won't read it anyway
[23:17] <nextgens> 47 /var/backups/freenet/daily/mailman/members_cvs.conf
[23:17] <nextgens> 2218 /var/backups/freenet/daily/mailman/members_announce.conf
[23:17] <nextgens> 215 /var/backups/freenet/daily/mailman/members_chat.conf
[23:17] <nextgens> 248 /var/backups/freenet/daily/mailman/members_devl.conf
[23:17] <nextgens> 390 /var/backups/freenet/daily/mailman/members_support.conf
[23:17] <nextgens> 178 /var/backups/freenet/daily/mailman/members_tech.conf
[23:17] <nextgens> we have more people on @devl than @tech :/
[23:18] <nextgens> that's bad
[23:19] * colione (n=colione@) Quit ("Leaving")
[23:20] * colione (n=colione@) has joined #freenet
[23:24] <Zothar_Work> true on most won't read it :)
[23:24] <Zothar_Work> perhaps tech got created awhile after devl?
[23:25] <nextgens> I dunno
[23:26] * nextgens is going to bed
[23:26] <nextgens> cya
[23:28] <Zothar_Work> nn
[23:31] * JustMe (i=JustMe_@) Quit ()
[23:38] * CIA-14 (i=cia@) Quit (Client Quit)
[23:50] * CIA-14 (i=cia@) has joined #freenet
[23:52] * richard223 (i=joe@) has joined #freenet
[23:58] * Muixirt (n=trixium@) Quit ("Gute Nacht !")
Irc logs of #freenet : 2008 2007 2006 2005
These logs were automatically created by FreenetLogBot on chat.freenode.net using the Java IRC LogBot.