--- Log opened Mon Apr 30 00:00:37 2012 | ||
-!- PhilTillet [~Philippe@npasserelle10.minet.net] has quit [Ping timeout: 276 seconds] | 00:04 | |
-!- PhilTillet [~Philippe@2001:660:3203:402:3545:6e43:ea02:c6b7] has joined #shogun | 00:13 | |
-!- blackburn [~qdrgsm@85.114.185.217] has joined #shogun | 01:17 | |
-!- wiking_ [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 02:04 | |
-!- wiking_ [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 02:04 | |
-!- wiking_ [~wiking@huwico/staff/wiking] has joined #shogun | 02:04 | |
-!- blackburn [~qdrgsm@85.114.185.217] has quit [Quit: Leaving.] | 02:06 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 256 seconds] | 02:08 | |
-!- wiking_ is now known as wiking | 02:08 | |
-!- emrecelikten1 [~emrecelik@92.44.126.141] has joined #shogun | 02:13 | |
-!- emrecelikten [~emrecelik@92.44.124.86] has quit [Ping timeout: 245 seconds] | 02:14 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Remote host closed the connection] | 02:20 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 02:20 | |
-!- av3ngr [av3ngr@nat/redhat/x-ripnmxahspswstqh] has joined #shogun | 02:34 | |
-!- Cheng [80fa500f@gateway/web/freenode/ip.128.250.80.15] has joined #shogun | 03:56 | |
-!- PhilTillet [~Philippe@2001:660:3203:402:3545:6e43:ea02:c6b7] has quit [Ping timeout: 260 seconds] | 04:05 | |
-!- Cheng [80fa500f@gateway/web/freenode/ip.128.250.80.15] has quit [Ping timeout: 245 seconds] | 04:40 | |
-!- wiking_ [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 05:47 | |
-!- wiking_ [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 05:47 | |
-!- wiking_ [~wiking@huwico/staff/wiking] has joined #shogun | 05:47 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Ping timeout: 250 seconds] | 05:51 | |
-!- wiking_ is now known as wiking | 05:51 | |
-!- gsomix [~gsomix@95.67.182.154] has joined #shogun | 06:04 | |
-!- gsomix [~gsomix@95.67.182.154] has quit [Client Quit] | 06:04 | |
-!- gsomix [~gsomix@95.67.182.154] has joined #shogun | 06:05 | |
-!- gsomix [~gsomix@95.67.182.154] has quit [Ping timeout: 252 seconds] | 07:53 | |
-!- n4nd0 [02893bbe@gateway/web/freenode/ip.2.137.59.190] has joined #shogun | 09:01 | |
-!- n4nd0 [02893bbe@gateway/web/freenode/ip.2.137.59.190] has quit [Quit: Page closed] | 09:18 | |
-!- blackburn [~qdrgsm@85.114.185.217] has joined #shogun | 09:49 | |
blackburn | that's nice to have holidays on monday | 10:01 |
---|---|---|
-!- av3ngr [av3ngr@nat/redhat/x-ripnmxahspswstqh] has quit [Quit: That's all folks!] | 10:07 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking] | 10:11 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 10:17 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 10:17 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 10:17 | |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has joined #shogun | 10:27 | |
blackburn | n4nd0: oops I forgot to merge your PR | 10:38 |
CIA-64 | shogun: iglesias master * r6e84b8e / (3 files in 2 dirs): * documentation JLCoverTree warnings fix - http://git.io/ZVgpUw | 10:38 |
CIA-64 | shogun: Sergey Lisitsyn master * ra11462d / (3 files in 2 dirs): Merge branch 'ct-warningfix' of git://github.com/iglesias/shogun - http://git.io/EvNH-A | 10:38 |
shogun-buildbot | build #799 of libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/libshogun/builds/799 blamelist: fernando.iglesiasg@gmail.com | 10:44 |
blackburn | oops? | 10:46 |
blackburn | how can that be :D | 10:46 |
blackburn | I fixed it other day | 10:47 |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has quit [Ping timeout: 246 seconds] | 10:47 | |
shogun-buildbot | build #800 of libshogun is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/libshogun/builds/800 | 10:48 |
-!- cheng [80fa500f@gateway/web/freenode/ip.128.250.80.15] has joined #shogun | 10:53 | |
blackburn | cheng: surprise to see you there ;) | 10:55 |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has joined #shogun | 11:00 | |
blackburn | n4nd0: ping | 11:03 |
cheng | Hi blackburn! Just logging on to see what the fuss is all about. ;-) | 11:09 |
n4nd0 | blackburn, hey | 11:09 |
blackburn | cheng: oh it would take a few days to check logs :D | 11:09 |
blackburn | ah logging on sorry | 11:09 |
blackburn | n4nd0: I merged your pr (forgot to do yesterday) | 11:10 |
blackburn | cheng: btw it would be nice to discuss multiclass idea | 11:10 |
n4nd0 | blackburn, yeah I saw it, thank you :) | 11:11 |
cheng | Yes, for sure. The idea would be to have a toolbox for converting binary classifiers into multiclass classifiers. | 11:12 |
blackburn | cheng: yeah but probably pluskid will be really fast in it so some more complex stuff should be here too | 11:13 |
cheng | I agree. I just haven't had too much time to discuss with pluskid yet. | 11:14 |
blackburn | cheng: do you have expertise in multiclass tree learning? | 11:14 |
cheng | I was traveling last week, and was offline (hence missing the first meeting). | 11:14 |
blackburn | cheng: (offtop) I wanted to ask you whether JMLR re-submission has some implicit deadline | 11:15 |
cheng | blackburn: unfortunately, I don't really have expertise (in pretty much anything). | 11:15 |
cheng | re-JMLR, not really, but if you wait too long, the reviewers tend to be more demanding. I'd say 2 months max if you can manage. | 11:16 |
blackburn | probably 2 months elapsed already :( | 11:16 |
blackburn | hmm no, rather 1.5 | 11:18 |
cheng | I mean 2 months from now. | 11:19 |
blackburn | ah | 11:19 |
cheng | re ecoc: John Langford's stuff on conditional trees seem to scale well. | 11:19 |
cheng | One thing I just learned about last week was estimating some sort of "confidence interval" | 11:20 |
cheng | This is called "conformal predictions", and there is a nice tutorial on JMLR. | 11:20 |
blackburn | cheng: you said you have no expertise? | 11:21 |
blackburn | :D | 11:21 |
cheng | Well, I know of work, but I don't yet know on how to implement the work. | 11:22 |
blackburn | cheng: not a problem - implementation is rather our part | 11:23 |
cheng | About conformal predictions, it seems to take any multiclass classifier and give back the "argmax" and another few predictions, which are in the "confidence interval" | 11:24 |
blackburn | cheng: that should be interesting | 11:25 |
cheng | It could also be an interesting research project in the structured prediction setting, but I think it is too much research for GSoC. | 11:25 |
blackburn | cheng: I actually thought of some kind of SO setting | 11:26 |
blackburn | cheng: do you think it would be reasonable to learn ECOCs? | 11:26 |
cheng | For plain multiclass however, it would be quite neat. Unfortunately, I don't yet know what other requirements the method needs. | 11:26 |
cheng | To learn ECOCs, Gunnar had an old paper "adaptive codes and polychotomies" or something like that. | 11:27 |
blackburn | really? I should check | 11:27 |
cheng | I had suggested that to pluskid (unfortunately in a private mail). | 11:27 |
cheng | 2002 or so, | 11:27 |
cheng | books.nips.cc/papers/files/nips15/AA04.pdf | 11:28 |
blackburn | cheng: sounds like great unimplemented idea.. | 11:29 |
cheng | Gunnar will be happy. | 11:29 |
-!- gsomix [~gsomix@95.67.153.157] has joined #shogun | 11:31 | |
gsomix | hi all | 11:31 |
cheng | It is also somehow related to cutting plane methods, which there are a few implementations of in shogun. | 11:31 |
blackburn | cheng: ideal case would be it is possible with libocas (with some patching) | 11:32 |
cheng | Hi gsomix | 11:32 |
blackburn | gsomix: hi | 11:32 |
cheng | blackburn: hopefully possible. I don't know. | 11:33 |
blackburn | cheng: however some students would work on libqp this year so it can be used too | 11:33 |
cheng | blackburn: I don't know enough about the implementation, but in principle you need a cutting plane method, and a QP solver, as you said. :-) | 11:35 |
-!- karlnapf [~heiko@host86-183-72-191.range86-183.btcentralplus.com] has joined #shogun | 11:36 | |
-!- cheng [80fa500f@gateway/web/freenode/ip.128.250.80.15] has quit [Ping timeout: 245 seconds] | 11:40 | |
wiking | fuuuuucking hell i've did the resubmission \o/ | 11:52 |
wiking | no sleep for the last 24 hours | 11:52 |
blackburn | wiking: ah! so what did you change? | 11:52 |
wiking | blackburn: ehhehe after i woke up | 11:53 |
wiking | i'll tell u | 11:53 |
blackburn | sure | 11:53 |
wiking | i think i'm gonna crash now | 11:53 |
blackburn | have a long sleep now :D | 11:53 |
-!- blackburn [~qdrgsm@85.114.185.217] has quit [Quit: Leaving.] | 11:55 | |
n4nd0 | yeah, have a good rest wiking | 11:55 |
shogun-buildbot | build #516 of ruby_modular is complete: Failure [failed test_1] Build details are at http://www.shogun-toolbox.org/buildbot/builders/ruby_modular/builds/516 blamelist: blackburn91@gmail.com | 12:14 |
-!- blackburn [~qdrgsm@85.114.185.217] has joined #shogun | 13:07 | |
-!- blackburn [~qdrgsm@85.114.185.217] has quit [Quit: Leaving.] | 13:15 | |
-!- emrecelikten1 [~emrecelik@92.44.126.141] has quit [Ping timeout: 252 seconds] | 13:46 | |
-!- cheng [73406f11@gateway/web/freenode/ip.115.64.111.17] has joined #shogun | 14:16 | |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has quit [Ping timeout: 246 seconds] | 14:27 | |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has joined #shogun | 14:27 | |
-!- karlnapf1 [~heiko@host86-176-4-122.range86-176.btcentralplus.com] has joined #shogun | 14:28 | |
-!- karlnapf [~heiko@host86-183-72-191.range86-183.btcentralplus.com] has quit [Ping timeout: 276 seconds] | 14:29 | |
-!- cheng [73406f11@gateway/web/freenode/ip.115.64.111.17] has quit [Quit: Page closed] | 14:38 | |
-!- PhilTillet [~Philippe@157.159.47.10] has joined #shogun | 15:42 | |
-!- karlnapf1 [~heiko@host86-176-4-122.range86-176.btcentralplus.com] has quit [Quit: Leaving.] | 16:31 | |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has quit [Ping timeout: 246 seconds] | 16:47 | |
-!- blackburn [~qdrgsm@85.114.185.217] has joined #shogun | 16:59 | |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has joined #shogun | 17:00 | |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has quit [Quit: Ex-Chat] | 17:13 | |
CIA-64 | shogun: Chiyuan Zhang master * ra5e054b / (2 files in 2 dirs): a better fix for the SWIG problem. - http://git.io/g5FM2Q | 17:15 |
CIA-64 | shogun: Chiyuan Zhang master * r2b13d07 / (10 files): ECOC Random Dense/Sparse Coding. - http://git.io/Ee9E3A | 17:15 |
blackburn | unstoppable pluskid | 17:15 |
-!- Priyans [~Priyans@115.248.130.148] has joined #shogun | 19:58 | |
-!- Priyans [~Priyans@115.248.130.148] has quit [Quit: Leaving] | 20:04 | |
CIA-64 | shogun: Sergey Lisitsyn master * r40c2582 / src/shogun/lib/external/shogun_liblinear.cpp : Significant performance improvement for multiclass liblinear - http://git.io/63encw | 20:17 |
blackburn | sonney2k: yay | 20:17 |
blackburn | at least 3x faster | 20:17 |
wiking | blackburn: "Congratulations on having your paper accepted to the Workshop on Medical Computer Vision at CVPR 2012" | 20:18 |
wiking | !!! | 20:18 |
blackburn | wiking: congrats :) | 20:18 |
blackburn | I guess you did great job there | 20:18 |
blackburn | wiking: btw could you please do regression test of liblinear? | 20:18 |
blackburn | just run and check if results are still the same | 20:19 |
blackburn | btw it should be much faster now | 20:21 |
-!- PhilTillet [~Philippe@157.159.47.10] has quit [Remote host closed the connection] | 20:38 | |
wiking | blackburn: eheheh i've just got back | 21:15 |
wiking | i had a little bit of FUCKYEAH moment :) | 21:15 |
blackburn | wiking: for a hour :D | 21:16 |
wiking | well yeah it happens :) | 21:16 |
wiking | ok | 21:17 |
wiking | we need a kernel | 21:17 |
wiking | laplacian | 21:17 |
wiking | which is almost like a gaussian | 21:17 |
wiking | that's what i've realized yesterday around 4am | 21:17 |
wiking | :> | 21:17 |
blackburn | wiking: ExponentialKernel? | 21:18 |
blackburn | CExponentialKernel | 21:18 |
wiking | http://crsouza.blogspot.com/2010/03/kernel-functions-for-machine-learning.html#laplacian | 21:18 |
wiking | is that it? | 21:18 |
wiking | almost | 21:18 |
wiking | see the difference on the page | 21:19 |
wiking | but i wonder what's the implementation in shogun | 21:19 |
blackburn | wiking: ah these differencies make no sense to me | 21:19 |
blackburn | according sigmas will make it equal | 21:20 |
shogun-buildbot | build #224 of nightly_all is complete: Success [build successful] Build details are at http://www.shogun-toolbox.org/buildbot/builders/nightly_all/builds/224 | 21:22 |
blackburn | it's been a long w/o success :D | 21:24 |
@sonney2k | blackburn, regarding learning ECOC - I was a young student sitting in the room when gunnar wrote that paper. he in the end said that there is very little gain if at all | 21:35 |
blackburn | sonney2k: great news :D | 21:35 |
@sonney2k | and couldn't really produce some example that was convincing | 21:35 |
@sonney2k | so I guess you can save some time :) | 21:35 |
blackburn | sonney2k: you weren't so young these days I think ;) | 21:36 |
@sonney2k | blackburn, what? 9 years ago is like inf time | 21:37 |
blackburn | sonney2k: you were older than me still | 21:37 |
@sonney2k | gsomix, did you manage to get things to compile with the new sgvector? | 21:37 |
blackburn | 5 years or so :D | 21:37 |
@sonney2k | blackburn, no 2002! | 21:39 |
blackburn | sonney2k: I mean you still were 5 years older than me | 21:39 |
gsomix | sonney2k, nope. I'm working on this now. | 21:39 |
blackburn | sonney2k: to not be surprised with gsomix patch I suggested it | 21:40 |
blackburn | (feature titles namely) | 21:40 |
@sonney2k | blackburn, I will always be older :) | 21:42 |
@sonney2k | blackburn, from gunnars paper " So far, we could only show a proof of concept and more experimental work is necessary. | 21:42 |
@sonney2k | It is in particular interesting to find practical examples, where a non-trivial choice of the | 21:42 |
@sonney2k | code (via optimization) helps simplifying the embedding and finally leads to additional | 21:42 |
@sonney2k | improvements. Such problems often appear in Computer Vision, where there are strong relationships between classes. Preliminary results indicate that one can achieve considerable | 21:42 |
@sonney2k | improvements when adapting codes and embeddings [3]." | 21:42 |
@sonney2k | haha :)) | 21:42 |
blackburn | sonney2k: yes twice | 21:42 |
blackburn | sonney2k: are you familiar with this paper? | 21:43 |
blackburn | what do they do in particular? | 21:43 |
blackburn | learn codes or learn classifier using structured output framework and codes as outputs? | 21:43 |
@sonney2k | blackburn, I just had a brief look - they learn ECOC and gunnar (as usual) uses column generation which is kind of cutting plane technique | 21:43 |
@sonney2k | they learn the 'optimal' code | 21:44 |
@sonney2k | but it didn't gain anything | 21:44 |
blackburn | nahh I don't want to do it | 21:44 |
blackburn | and never did | 21:44 |
blackburn | I want to learn classifiers with ECOC outputs | 21:44 |
blackburn | in SO fashion | 21:44 |
@sonney2k | blackburn, still it is intersting if one can learn the *optimal* coding | 21:44 |
@sonney2k | maybe it can help wikings vision problems or so | 21:45 |
@sonney2k | blackburn, good job on liblinear :) | 21:45 |
@sonney2k | that is for multiclass only right? | 21:45 |
blackburn | sonney2k: yes sure | 21:45 |
blackburn | I am not interested in binary | 21:45 |
@sonney2k | or did you do it for L1 regul too? | 21:45 |
blackburn | hmmm is it needed? I have never checked the code | 21:46 |
@sonney2k | blackburn, I guess just grep for the feature iterator stuff is best | 21:47 |
blackburn | sonney2k: L1 makes no use of iterator so far | 21:47 |
@sonney2k | blackburn, what did you suggest to gsomix ? | 21:47 |
@sonney2k | gsomix, ok np | 21:47 |
blackburn | sonney2k: names/titles of features | 21:47 |
blackburn | like 1st feature is 'weight' | 21:47 |
blackburn | and 99th is 'hair color' :D | 21:48 |
blackburn | sonney2k: yes liblinear needs similar stuff | 21:49 |
blackburn | L1 as you said | 21:49 |
blackburn | L1LR | 21:49 |
blackburn | terrible | 21:50 |
blackburn | a lot of | 21:50 |
@sonney2k | blackburn, you really want to maximize overhead don't you? | 21:53 |
blackburn | sonney2k: overhead of cfeatures?? | 21:53 |
blackburn | you must be kidding | 21:54 |
@sonney2k | blackburn, I thought we are talking about sgvector? | 21:54 |
blackburn | sonney2k: cfeatures | 21:54 |
@sonney2k | you didn't say so | 21:54 |
blackburn | was implicit :D | 21:54 |
blackburn | sonney2k: no I do not want any overhead in sgvector | 21:54 |
blackburn | however I do not care about overhead of sgobjects | 21:55 |
blackburn | we have giant parameters map and other stuff | 21:55 |
blackburn | this already makes sgobjects not so lightweight but hey who cares | 21:55 |
blackburn | I wouldn't say typical shogun based program would use more than 100 objects | 21:56 |
gsomix | sonney2k, what about ref() in copy constructor and unref() in destructor? can I add it? or do we prefer manually control? | 21:56 |
@sonney2k | blackburn, sure. agreed. | 22:02 |
@sonney2k | gsomix, yes ref in copy constructor / unref in destructor | 22:02 |
gsomix | \o/ | 22:03 |
blackburn | sonney2k: patching L1 stuff.. | 22:04 |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has joined #shogun | 22:05 | |
-!- emrecelikten [~emrecelik@92.44.126.141] has joined #shogun | 22:08 | |
* wiking is in workmoooode | 22:17 | |
* gsomix tired today | 22:18 | |
* blackburn drinks moar vodka and keep patching | 22:19 | |
gsomix | sonney2k, it seems, I need a sleep. | 22:19 |
gsomix | blackburn, I'll come in Samara tomorrow, wait for me. leave vodka :) | 22:20 |
blackburn | gsomix: I am not in samara | 22:21 |
gsomix | blackburn, ah. =___= | 22:21 |
blackburn | and you said you prefer not to drink :P | 22:21 |
blackburn | gsomix: actually you should let us know you had a bday | 22:21 |
blackburn | :D | 22:21 |
blackburn | wiking: n4nd0: sonney2k: this guy gsomix have a bday today and keep it in secret :D | 22:22 |
n4nd0 | wow | 22:22 |
n4nd0 | happy birthday gsomix! | 22:23 |
wiking | gsomix: !!!!!!!!! | 22:23 |
n4nd0 | felicidades! | 22:23 |
n4nd0 | grattis pÄ födelsedagen! | 22:23 |
wiking | gsomix: sretan rodjendan | 22:23 |
blackburn | lol | 22:24 |
wiking | i wonder if it's similar to russian :P | 22:24 |
blackburn | wiking: s dnem rojdenia | 22:24 |
gsomix | n4nd0, thanks | 22:24 |
wiking | :>> | 22:24 |
blackburn | wiking: sounds a little similar | 22:24 |
gsomix | wiking, spasibo :] | 22:24 |
wiking | gsomix: :) | 22:25 |
gsomix | you guys are cool | 22:27 |
gsomix | unfortunately I need a sleep | 22:27 |
gsomix | good night :) | 22:27 |
n4nd0 | good night! | 22:28 |
blackburn | that L1 and transposed features is something I do not like.. | 22:30 |
blackburn | sonney2k: can you motivate this transposition? | 22:33 |
blackburn | is it to use feature iterator as iterator on one feature of different vectors? | 22:35 |
-!- karlnapf [~heiko@host86-176-4-122.range86-176.btcentralplus.com] has joined #shogun | 22:37 | |
-!- gsomix [~gsomix@95.67.153.157] has quit [Ping timeout: 252 seconds] | 22:43 | |
-!- PhilTillet [~Philippe@2001:660:3203:402:1d2a:c7fe:1fff:9fdb] has joined #shogun | 22:46 | |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has quit [Ping timeout: 246 seconds] | 22:47 | |
-!- PhilTillet [~Philippe@2001:660:3203:402:1d2a:c7fe:1fff:9fdb] has quit [Ping timeout: 245 seconds] | 22:51 | |
-!- PhilTillet [~Philippe@157.159.42.154] has joined #shogun | 22:58 | |
-!- n4nd0 [~n4nd0@190.Red-2-137-59.dynamicIP.rima-tde.net] has joined #shogun | 23:01 | |
-!- wiking [~wiking@huwico/staff/wiking] has quit [Quit: wiking] | 23:23 | |
-!- blackburn [~qdrgsm@85.114.185.217] has quit [Quit: Leaving.] | 23:25 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has joined #shogun | 23:25 | |
-!- wiking [~wiking@78-23-189-112.access.telenet.be] has quit [Changing host] | 23:25 | |
-!- wiking [~wiking@huwico/staff/wiking] has joined #shogun | 23:25 | |
-!- blackburn [~qdrgsm@85.114.185.217] has joined #shogun | 23:25 | |
-!- blackburn [~qdrgsm@85.114.185.217] has quit [Quit: Leaving.] | 23:31 | |
-!- blackburn [~qdrgsm@85.114.185.217] has joined #shogun | 23:33 | |
-!- emrecelikten [~emrecelik@92.44.126.141] has quit [Read error: Operation timed out] | 23:33 | |
blackburn | oh nice I started to receive strange characters instead of russian words in icq :D | 23:33 |
-!- blackburn [~qdrgsm@85.114.185.217] has quit [Quit: Leaving.] | 23:43 | |
-!- blackburn [~qdrgsm@85.114.185.217] has joined #shogun | 23:51 | |
--- Log closed Tue May 01 00:00:37 2012 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!