--- Log opened Mon Jun 17 00:00:48 2013 | ||
@sonney2k | HeikoS, I found this - which should help us a lot :D | 00:31 |
---|---|---|
@sonney2k | http://www.berlin.de/imperia/md/content/senatsverwaltungen/justiz/formularserver/vereinsregister/merkblatt_zur_eintragung_eines_vereins.pdf?start&ts=1256891879&file=merkblatt_zur_eintragung_eines_vereins.pdf | 00:31 |
@HeikoS | sonney2k: nice :) | 00:31 |
@HeikoS | just need a protocol for berlin then :) | 00:32 |
@HeikoS | and somebody has to bring this stuff to the notar | 00:34 |
@HeikoS | I will go to bed now, good night everyone, see you tomorrow | 00:34 |
@HeikoS | tomorrow GSoC starts yeaaaah ;) | 00:34 |
-!- HeikoS [~heiko@176.248.212.166] has quit [Quit: Leaving.] | 00:35 | |
-!- FSCV [~FSCV@189.139.252.135] has quit [Ping timeout: 248 seconds] | 02:01 | |
-!- lisitsyn1 [~lisitsyn@37.61.181.59] has joined #shogun | 04:10 | |
-!- gsomix [~Miranda@188.168.3.10] has quit [Ping timeout: 245 seconds] | 04:12 | |
-!- lisitsyn [~lisitsyn@188.168.3.10] has quit [Ping timeout: 268 seconds] | 04:12 | |
-!- pickle27 [~Kevin@S0106002191dec7e8.cg.shawcable.net] has joined #shogun | 04:20 | |
-!- nube [~rho@116.90.239.13] has joined #shogun | 06:16 | |
-!- lambday [67157d37@gateway/web/cgi-irc/kiwiirc.com/ip.103.21.125.55] has joined #shogun | 07:12 | |
lambday | sonney2k: moin moin :) | 07:13 |
lambday | sonney2k: with the pretty function thing, all we gotta write in the error msg is the msg itself, right? | 07:14 |
-!- lambday [67157d37@gateway/web/cgi-irc/kiwiirc.com/ip.103.21.125.55] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 07:21 | |
-!- lambday [67157d37@gateway/web/cgi-irc/kiwiirc.com/ip.103.21.125.55] has joined #shogun | 07:25 | |
@sonney2k | lambday, hey there! | 08:19 |
@sonney2k | lambday, yes exactly! | 08:19 |
lambday | sonney2k: okay | 08:20 |
@sonney2k | lambday, it should be a config in shogun/io/SGIO* to select how detailed you want to know where the error occurs | 08:20 |
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun | 08:21 | |
@sonney2k | moin votjakovr! | 08:23 |
lambday | sonney2k: okay..so once HeikoS's patch is merged, it will be enabled by default? or should we configure shogun with something? | 08:24 |
votjakovr | sonney2k: good morning! | 08:24 |
-!- gsomix [~Miranda@r206-10.smr.ru] has joined #shogun | 08:27 | |
gsomix | good morning | 08:27 |
@sonney2k | lambday, I just hacked sth up that should do it | 08:29 |
@sonney2k | compiling... | 08:29 |
lambday | sonney2k: :) :) okay | 08:29 |
lambday | good morning everyone :) | 08:30 |
@sonney2k | lambday, if I commit this - can you test this please? | 08:31 |
lambday | sonney2k: sure | 08:31 |
lambday | :) | 08:31 |
@sonney2k | patch is really small... | 08:31 |
* sonney2k compiles | 08:31 | |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 08:35 | |
shogun-notifier- | shogun: Soeren Sonnenburg :develop * 8181acb / src/shogun/io/SGIO.cpp,src/shogun/io/SGIO.h: https://github.com/shogun-toolbox/shogun/commit/8181acb60da0d7182497d30594223a9297944541 | 08:35 |
shogun-notifier- | shogun: enable location info (as in object / function name printing before SGIO | 08:35 |
shogun-notifier- | shogun: prints) with config option to disable / show line and file | 08:35 |
@sonney2k | lambday, oki committed | 08:35 |
lambday | pulling | 08:35 |
@sonney2k | okidoki | 08:35 |
@sonney2k | gtg brb | 08:35 |
lambday | sonney2k: ciao :) | 08:35 |
lambday | I'll test it | 08:35 |
-!- pickle27 [~Kevin@S0106002191dec7e8.cg.shawcable.net] has quit [Quit: Leaving] | 08:38 | |
-!- lisitsyn1 [~lisitsyn@37.61.181.59] has quit [Quit: Leaving.] | 08:44 | |
-!- sonne|work [~sonnenbu@sams-office-nat.tomtomgroup.com] has joined #shogun | 08:51 | |
sonne|work | lambday: okidoki back! | 08:53 |
-!- lambday_ [67157f37@gateway/web/freenode/ip.103.21.127.55] has joined #shogun | 08:58 | |
lambday_ | sonne|work: okie dokie tested | 08:58 |
lambday_ | sorry I got dc | 08:58 |
lambday_ | :( | 08:58 |
lambday_ | sonne|work: one problem though | 08:58 |
sonne|work | yes? | 08:58 |
-!- lambday [67157d37@gateway/web/cgi-irc/kiwiirc.com/ip.103.21.125.55] has quit [Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] | 08:58 | |
lambday_ | sonne|work: it always shows that it originates from void shogun::SGIO::message | 08:58 |
sonne|work | hah yes | 08:59 |
sonne|work | it has to be in the #define then | 08:59 |
* lambday_ is confused | 08:59 | |
lambday_ | sonne|work: how far your workplace is from where you live? :P | 09:00 |
sonne|work | I don't recall exactly but lets say ~20KMs | 09:01 |
lambday_ | sonne|work: you sure drive fast :P | 09:02 |
sonne|work | lambday_: I go by train and whenever I have to leave it and walk I say brb :D | 09:03 |
sonne|work | but yes we are all michael schumacher here ;) | 09:03 |
lambday_ | sonne|work: hahahahaha :D | 09:04 |
shogun-buildbot | build #942 of cyg1 - libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/942 blamelist: Soeren Sonnenburg <sonne@debian.org> | 09:04 |
sonne|work | lambday_: ok I know now how to do it properly. I don't have time to test this though | 09:05 |
lambday_ | sonne|work: I'll test it | 09:05 |
lambday_ | just a small test() with a REQUIRE will do, right? | 09:06 |
lambday_ | tested with a few classes that uses REQUIRE as well | 09:06 |
shogun-notifier- | shogun: Soeren Sonnenburg :develop * e2142df / src/shogun/io/SGIO.cpp,src/shogun/io/SGIO.h: https://github.com/shogun-toolbox/shogun/commit/e2142df4a17e39d4ec94023be6198c59d370a963 | 09:50 |
shogun-notifier- | shogun: add pretty function macro in defines | 09:50 |
sonne|work | lambday_: no idea how to really test it in a test | 09:50 |
sonne|work | looking at it should be enough | 09:50 |
lambday_ | sonne|work: I'm checking | 09:52 |
lambday_ | sonne|work: works like a charm! :) | 09:56 |
shogun-buildbot | build #943 of cyg1 - libshogun is complete: Failure [failed configure] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/943 blamelist: Soeren Sonnenburg <sonne@debian.org> | 10:03 |
sonne|work | lambday_: what is the output of such message now? | 10:05 |
shogun-notifier- | shogun: lambday :develop * d8dc5b5 / tests/unit/base/SGObject_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/d8dc5b55611729a83213b460e4984ca959a7ffd5 | 10:05 |
shogun-notifier- | shogun: minor fix for eigen3 check in SGObject_unittest | 10:05 |
shogun-notifier- | shogun: Soeren Sonnenburg :develop * cc2f703 / tests/unit/base/SGObject_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/cc2f703ebeb36d059b632e1881b757d75fcc88ab | 10:05 |
shogun-notifier- | shogun: Merge pull request #1171 from lambday/develop | 10:06 |
shogun-notifier- | shogun: | 10:06 |
shogun-notifier- | shogun: minor fix for eigen3 check in SGObject_unittest | 10:06 |
-!- nube1 [~rho@116.90.239.3] has joined #shogun | 10:10 | |
-!- nube [~rho@116.90.239.13] has quit [Ping timeout: 256 seconds] | 10:13 | |
-!- van51 [~van51@athedsl-408350.home.otenet.gr] has joined #shogun | 10:14 | |
shogun-buildbot | build #944 of cyg1 - libshogun is complete: Failure [failed configure] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/944 blamelist: lambday <heavensdevil6909@gmail.com>, Soeren Sonnenburg <sonne@debian.org> | 10:29 |
-!- nube1 [~rho@116.90.239.3] has quit [Remote host closed the connection] | 10:41 | |
-!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun | 11:40 | |
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout] | 13:06 | |
@wiking | lisitsyn: ping | 13:08 |
lisitsyn | wiking: pong | 13:09 |
@wiking | lisitsyn: cmake? :) | 13:11 |
lisitsyn | wiking: cmdline is ready will continue | 13:11 |
@wiking | coool! | 13:12 |
lisitsyn | wiking: not today though going to have exam tomorrow :D | 13:12 |
@wiking | lisitsyn: other question: any ideas how to do a check in general for CFeatures to find out whether all the vectors in it are equal or not? | 13:12 |
@wiking | i see that with DotFeatures it's possible to do this... but with CFeatures in general i see no way with the current api | 13:13 |
lisitsyn | wiking: like vectors of CFeatures f1 == vectors of CFeatures f2? | 13:13 |
lisitsyn | wiking: yeah CFeatures are quite general | 13:14 |
lisitsyn | why do you need it? | 13:14 |
@wiking | lisitsyn: like: all f_i in F, where F is CFeatures | 13:14 |
@wiking | i mean for all i,j f_i = f_j in F | 13:14 |
lisitsyn | so all vectors are the same? | 13:15 |
@wiking | yeps | 13:15 |
@wiking | dotfeatures supports an iteration i see that | 13:15 |
lisitsyn | strange you need that! | 13:15 |
lisitsyn | why/ | 13:15 |
lisitsyn | ? | 13:15 |
@wiking | lisitsyn: decision tree | 13:15 |
lisitsyn | I mean it is quite singular case or? | 13:16 |
@wiking | yeah well i dont know any other algo that would require this api :) | 13:17 |
lisitsyn | wiking: however can you build a decision tree on features only? | 13:17 |
lisitsyn | w/o at least dotfeatures | 13:17 |
@wiking | lisitsyn: indeed | 13:17 |
@wiking | lisitsyn: it's just branching based on the features | 13:17 |
@wiking | you need to calculate an information gain value based on given amount of features | 13:18 |
@wiking | and that's it | 13:18 |
@wiking | you take the highest valued feature and branch there the tree... | 13:18 |
@wiking | mmm we don't have a factory class for CFeatures as in case for Clabels rith? | 13:19 |
@wiking | *right? | 13:19 |
-!- lambday_ [67157f37@gateway/web/freenode/ip.103.21.127.55] has quit [Ping timeout: 250 seconds] | 13:20 | |
lisitsyn | wiking: yes | 13:21 |
lisitsyn | we don't have it iirc | 13:21 |
@wiking | mmmm this is going to be funny with some dynamic_cast stuff ;) | 13:24 |
lisitsyn | wiking: which means we have wrong design somewhere :) | 13:30 |
-!- gsomix [~Miranda@r206-10.smr.ru] has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org] | 13:30 | |
lisitsyn | wiking: I developed a thing that does format("{} fdfsd {} dfs {}",2.0,'c', 1); => "2.0 fdfsd c dfs 1" | 13:31 |
lisitsyn | may be we should replace our old-style %d %s etc with it | 13:32 |
-!- lambday [67157e36@gateway/web/freenode/ip.103.21.126.54] has joined #shogun | 13:39 | |
-!- hushell [~hushell@c-24-21-141-32.hsd1.or.comcast.net] has quit [Ping timeout: 264 seconds] | 14:24 | |
-!- hushell [~hushell@c-24-21-141-32.hsd1.or.comcast.net] has joined #shogun | 14:24 | |
-!- van51 [~van51@athedsl-408350.home.otenet.gr] has left #shogun ["QUIT :Leaving."] | 14:28 | |
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has left #shogun ["ERC Version 5.3 (IRC client for Emacs)"] | 15:05 | |
-!- nube [~rho@49.244.32.220] has joined #shogun | 15:14 | |
-!- nube [~rho@49.244.32.220] has quit [Ping timeout: 245 seconds] | 15:19 | |
-!- nube [~rho@36.253.226.40] has joined #shogun | 15:22 | |
-!- nube1 [~rho@49.244.61.96] has joined #shogun | 15:30 | |
-!- nube [~rho@36.253.226.40] has quit [Ping timeout: 252 seconds] | 15:33 | |
-!- nube1 [~rho@49.244.61.96] has quit [Ping timeout: 240 seconds] | 15:46 | |
-!- nube [~rho@49.126.230.83] has joined #shogun | 16:07 | |
-!- nube1 [~rho@49.244.29.250] has joined #shogun | 16:27 | |
-!- nube [~rho@49.126.230.83] has quit [Ping timeout: 256 seconds] | 16:27 | |
-!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.] | 16:31 | |
-!- nube1 [~rho@49.244.29.250] has quit [Ping timeout: 245 seconds] | 16:33 | |
-!- van51 [~van51@athedsl-408350.home.otenet.gr] has joined #shogun | 16:52 | |
-!- nube [~rho@49.244.72.191] has joined #shogun | 17:05 | |
-!- nube [~rho@49.244.72.191] has quit [Ping timeout: 264 seconds] | 17:09 | |
-!- lisitsyn [~lisitsyn@83.234.169.173] has joined #shogun | 17:14 | |
-!- mode/#shogun [+o lisitsyn] by ChanServ | 17:14 | |
-!- nube [~rho@49.244.37.190] has joined #shogun | 17:22 | |
-!- nube [~rho@49.244.37.190] has quit [Ping timeout: 246 seconds] | 17:39 | |
-!- HeikoS [~heiko@nat-179-180.internal.eduroam.ucl.ac.uk] has joined #shogun | 17:45 | |
-!- mode/#shogun [+o HeikoS] by ChanServ | 17:45 | |
@HeikoS | sonney2k, sonne|work, lisitsyn turns out I *will* be able to make it this evening | 17:46 |
-!- gsomix [~Miranda@83.234.169.173] has joined #shogun | 17:46 | |
gsomix | hello | 17:46 |
-!- nube [~rho@49.126.230.83] has joined #shogun | 17:56 | |
lambday | HeikoS: there? | 18:17 |
@HeikoS | lambday: yes, hi | 18:17 |
@HeikoS | I just talked to Daniel a bit | 18:17 |
lambday | HeikoS: how's your work going? :) | 18:17 |
lambday | okay.. | 18:17 |
@HeikoS | done | 18:17 |
@HeikoS | submitting today! :) | 18:17 |
lambday | HeikoS: awesome! | 18:17 |
@HeikoS | its the paper for the website you have seen | 18:17 |
@HeikoS | lambday: I talked to Daniel about the colours | 18:18 |
@HeikoS | we have to take one sample from each color | 18:18 |
@HeikoS | otherwise, there might be problems | 18:18 |
lambday | HeikoS: yes | 18:18 |
lambday | HeikoS: so, not going to a problem I guess | 18:19 |
@HeikoS | but I think we can do this via having an index variable for the probing sampler which color to use | 18:19 |
lambday | HeikoS: yes | 18:19 |
@HeikoS | and then just have one instance per color | 18:19 |
@HeikoS | and pass this on... | 18:19 |
@HeikoS | so no problems | 18:19 |
@HeikoS | phew! :) | 18:19 |
lambday | yeah! phew! :D | 18:19 |
@HeikoS | looking at your diagram ... | 18:19 |
@HeikoS | what di you change? | 18:19 |
@lisitsyn | re | 18:19 |
lambday | by the way please check the pseudo code (whis is not so pseudo but looks more like c++) :( | 18:19 |
lambday | umm.. | 18:20 |
lambday | wait | 18:20 |
lambday | mainly, added the finalize method | 18:20 |
lambday | moved shifts weights etc to the rational approximation class | 18:20 |
@HeikoS | so for the aggregation methods | 18:20 |
@HeikoS | I would add two: one which adds numbers, and one which adds up scalars | 18:20 |
@HeikoS | where finalize might divide by the number of added terms | 18:21 |
@HeikoS | rest is looking good | 18:21 |
@HeikoS | we will see whether it will work | 18:21 |
@HeikoS | so checking pseudoicode now | 18:22 |
lambday | HeikoS: I didn't get.. | 18:22 |
@HeikoS | I would not store vectors and scalars, but rather accumulate them | 18:22 |
@HeikoS | could you change the language to c++ in the gist? | 18:22 |
lambday | hehe okay | 18:22 |
@sonney2k | HeikoS, happy with the logging? | 18:23 |
@HeikoS | sonney2k: I saw the patch, not used it yet though, I guess its fine as its printing the context? :) | 18:23 |
@sonney2k | HeikoS, and regarding this evening. you have to | 18:23 |
@HeikoS | Now I have to undo all my log messages . noooooo | 18:23 |
@HeikoS | sonney2k: what? | 18:23 |
lambday | HeikoS: changed | 18:24 |
@HeikoS | lambday: did not really change that much ;) | 18:24 |
@HeikoS | lambday: for pseudo code, dont use pointers | 18:24 |
@HeikoS | but dont worry now | 18:24 |
lambday | HeikoS: oh and another thing... I was thinking of sending the classes one by one... so, should I sent PR for CLinearOperator, CDenseMatrixOperator and one unittest? | 18:24 |
@HeikoS | yeah for the basic building block at least | 18:25 |
lambday | HeikoS: alright, sending | 18:25 |
-!- zxtx [~zv@cpe-75-83-151-252.socal.res.rr.com] has quit [Ping timeout: 264 seconds] | 18:27 | |
lambday | argh stupid mistake in the cpp | 18:29 |
lambday | oops | 18:29 |
lambday | pseudocode | 18:29 |
lambday | HeikoS: in https://gist.github.com/lambday/5795796#file-logdet-cpp-L105, read it as "float64_t result=s.dot(get_imag(m_aggretation)) | 18:29 |
lambday | m_aggregate** | 18:30 |
@HeikoS | lambday: note I will be able to merge/review a bit | 18:37 |
@HeikoS | just not all day around | 18:37 |
@HeikoS | but in the evening | 18:37 |
@HeikoS | ah, lambday one more thing | 18:38 |
@HeikoS | so every GSoC student was asked to think about a real life example to present in the next meeting | 18:38 |
@HeikoS | for you, this can be computing the likelihood of the ozone model that we describe in that paper | 18:39 |
@HeikoS | so if you take a look at that ozone page on the website I sent you | 18:39 |
@HeikoS | there is the model (its just a big big Gaussian) | 18:39 |
@HeikoS | I have the data for that etc. So the first cool thing would be to be able to evaluate the likelihood of that model using our methods | 18:39 |
@HeikoS | thats proper real-life data | 18:39 |
@HeikoS | next steps can be maximum likelihood solutions, and then Bayesian inference over that | 18:40 |
@HeikoS | so you can think about how to present this to the others (just a few scenences) | 18:40 |
@HeikoS | sentences | 18:40 |
lambday | HeikoS: alright | 18:40 |
lambday | thanks a lot :) | 18:40 |
@HeikoS | lambday: NIPS was about adaptive MCMC, will share as soon as we make it public | 18:41 |
lambday | I was going to ask you about this :) | 18:41 |
lambday | okay :) cool! | 18:41 |
lambday | by the way, while sending PR, it says base branch "develop" - how can I send the PR against this feature branch? | 18:42 |
lambday | HeikoS: in my repo its feature/log_determinant already | 18:43 |
@HeikoS | you have to commit the stuff inside a feature branch locally | 18:43 |
@HeikoS | then push that feature branch to your fork | 18:43 |
@HeikoS | then send a PR from that feature branch against the same branch in shogun | 18:43 |
lambday | HeikoS: I already pushed local feature branch to my fork | 18:44 |
@HeikoS | so send the pr, when doing this you can select the branch | 18:44 |
lambday | shogun's base branch doesn't have the feature/log_determinant, it shows develop, and the head branch (my fork) shows feature.. | 18:46 |
lambday | would it create it once I send the PR? :-/ sorry I am still fighting with git :( | 18:47 |
@HeikoS | ah I see | 18:47 |
@HeikoS | hmm | 18:47 |
@HeikoS | lisitsyn: how to create feature branchs in shogun's repository? | 18:48 |
@HeikoS | can this be done via PRs? | 18:48 |
@lisitsyn | HeikoS: just create it locally and push | 18:48 |
@HeikoS | lisitsyn: do I have to do it or can a PR create one? | 18:48 |
@lisitsyn | HeikoS: I don't think it is possible to modify that by PR | 18:48 |
@HeikoS | lisitsyn: thanks :) one more question, how to reset to upstream current? | 18:49 |
@HeikoS | lambday: I will create the feature branch once I figured out how to reset my stuff :) | 18:49 |
lambday | HeikoS: okay :) | 18:49 |
@lisitsyn | HeikoS: like that iirc git reset origin/develop | 18:49 |
lambday | HeikoS: I'll be back after dinner | 18:50 |
@HeikoS | lambday: enjoy! | 18:50 |
@HeikoS | lambday: created a feature branch lambday-log-det for you | 18:55 |
-!- nube [~rho@49.126.230.83] has quit [Ping timeout: 276 seconds] | 19:03 | |
-!- nube [~rho@49.126.230.83] has joined #shogun | 19:07 | |
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun | 19:30 | |
votjakovr | HeikoS: hi! I've recieved your mail | 19:31 |
@HeikoS | votjakovr: hi there :) | 19:31 |
@HeikoS | should we talk abot things now or tomorrow? | 19:32 |
votjakovr | HeikoS: i'd like to hear your ideas about real world GP application... I think it should show us an advantages of using GPC wrt other classifiers | 19:34 |
@HeikoS | votjakovr: thats a good idea | 19:35 |
@HeikoS | so in contrast to SVM, we have proper binary probabilities | 19:35 |
@HeikoS | however, with Laplace approximation on binary problems, we will not massively beat SVMs | 19:36 |
votjakovr | HeikoS: yep i agree | 19:36 |
@HeikoS | since we cannot yet integrate out the hyperparameters (which would give better results) | 19:36 |
@HeikoS | It is still nice to have good example though | 19:36 |
@HeikoS | even if we reach same performance as SVM | 19:37 |
@HeikoS | an example where the probabilities count would be good | 19:37 |
@HeikoS | then we can make a point on that | 19:37 |
@HeikoS | votjakovr: I think I would go for reproducing an example that is used in papers | 19:38 |
@HeikoS | votjakovr: both a 2-dimensional (we can do nice plots of the predictive distribution then) and a higher dimensional | 19:38 |
@HeikoS | also, a graphical toy example (click point together) would be nice | 19:39 |
votjakovr | HeikoS: btw maybe take an example from Girolami's paper for multiclass. What do you think? | 19:40 |
@HeikoS | votjakovr: yes as a toy | 19:43 |
@HeikoS | the ring one is nice | 19:43 |
@HeikoS | but they also have proper data | 19:43 |
@HeikoS | votjakovr: I think we should go for multiple examples | 19:43 |
@HeikoS | toy/real for binary, both in low and higher dimensions | 19:43 |
@HeikoS | same for multiclass | 19:43 |
@HeikoS | for multiclass, we might beat SVMs with the probabilities since they only can use heuristics | 19:43 |
@HeikoS | and then a nice description and a screenshot for every example | 19:44 |
@HeikoS | that would be really cool | 19:44 |
@HeikoS | a little classification study using GPs | 19:44 |
@HeikoS | votjakovr: that will help people a lot | 19:44 |
@HeikoS | votjakovr: ok, since I am not there at the GSoC meeting, you will have to present these ideas (just a few sentences) | 19:44 |
votjakovr | HeikoS: Ok :) Also I'll try to find a good datasets for examples | 19:50 |
@HeikoS | votjakovr: nice :) | 19:50 |
@HeikoS | votjakovr: since we are covering quite a few cases , individual example do not have to be extremely complicated. Just some nice real/toy data should be fine. | 19:51 |
votjakovr | HeikoS: oh yeah | 19:52 |
@HeikoS | votjakovr: ok, nice, so how are the other things going? | 19:53 |
-!- zxtx [~zv@rrcs-74-62-200-195.west.biz.rr.com] has joined #shogun | 19:54 | |
lambday | HeikoS: back | 19:58 |
lambday | HeikoS: thanks :) I'm sending the PR | 19:58 |
lambday | HeikoS: something is wrong :( it shows 16 commits, I had only one :( | 20:01 |
votjakovr | HeikoS: today i was drawing diagrams, but i have not finished yet. Also i was reading about some related GPC stuff | 20:01 |
@HeikoS | lambday: what did you try to do? | 20:01 |
@HeikoS | votjakovr: do you think you will finish the GP-refactoring (diagrams and code) until tomorrow? | 20:02 |
@HeikoS | also lets do the logit binary case first | 20:02 |
lambday | HeikoS: I tried to sent the PR against lambday-log-det from feature/log_determinant (a branch in my fork) | 20:02 |
@HeikoS | this is in the rasmussen GP book under chapter classification, its the first case (Laplace approximation) | 20:02 |
@HeikoS | but laplace approximation is already there | 20:02 |
@HeikoS | lambday: let me check the feature branch | 20:03 |
@HeikoS | I think I forgot to rebase against develop head | 20:03 |
votjakovr | HeikoS: i agree, that laplace+logit likelihood will be the first | 20:03 |
@HeikoS | votjakovr: but really first, the classification/regression distinction needs to be made, so focus on that for now | 20:04 |
@HeikoS | I just want to make sure that you are ready to work on this stuff while I am not around. would be annoying for you if you get stuck with questions and nobody can answer them | 20:04 |
@HeikoS | lambday: yeah I have local git issues :( | 20:05 |
@HeikoS | give me a minute | 20:05 |
lambday | HeikoS: no problem | 20:05 |
votjakovr | HeikoS: Ah, ok, i'll try to finish classification/regression distinction tomorrow. | 20:07 |
@HeikoS | votjakovr: cool | 20:07 |
@HeikoS | and then afterwards, we can start worrying about the GP internals, so lets meet tomorrow and discuss if you are fine with that? | 20:07 |
votjakovr | HeikoS: ok | 20:08 |
@HeikoS | lambday: did you rebase while doing pull? | 20:08 |
@HeikoS | looks like not ... | 20:08 |
@HeikoS | argh | 20:08 |
@HeikoS | I dont know why this is, maybe just send the PR against develop | 20:08 |
@HeikoS | is also fine | 20:08 |
@HeikoS | I am annoyed by the git stuff ;) | 20:09 |
lambday | HeikoS: oops! | 20:10 |
lambday | HeikoS: I rebased against develop | 20:10 |
lambday | sorry, changing! | 20:10 |
lambday | :( | 20:10 |
@HeikoS | lambday: no that should be fine | 20:10 |
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun | 20:11 | |
shogun-notifier- | shogun: Heiko Strathmann :lambday-log-det * 512d94f / src/shogun/lib/DataType.cpp,src/shogun/lib/DataType.h: https://github.com/shogun-toolbox/shogun/commit/512d94f6a6abb1a1fe43eb22d4dbe65e2347c855 | 20:11 |
shogun-notifier- | shogun: refactored sizeof_ptype, sizeof_stype to make them available from static context | 20:11 |
shogun-notifier- | shogun: Heiko Strathmann :lambday-log-det * a5435a5 / / (3 files): https://github.com/shogun-toolbox/shogun/commit/a5435a543f2265fc4409ac487f33fc72d7e7b663 | 20:11 |
shogun-notifier- | shogun: first parts of copy_ptype with unit tests | 20:11 |
shogun-notifier- | shogun: Heiko Strathmann :lambday-log-det * 883ac63 / / (3 files): https://github.com/shogun-toolbox/shogun/commit/883ac6316c1401812621067b8890205715c6b684 | 20:11 |
shogun-notifier- | shogun: -refactored compare_stype a bit | 20:11 |
shogun-notifier- | shogun: -added copy_stype method with first bits | 20:11 |
shogun-notifier- | shogun: -added first unit test for copy_stype | 20:11 |
shogun-notifier- | shogun: Heiko Strathmann :lambday-log-det * caef78d / / (8 files): https://github.com/shogun-toolbox/shogun/commit/caef78dbe9682e5b6aa77ad8feb25dd70e8be497 | 20:11 |
shogun-notifier- | shogun: initial (working) attempt for a general purpose clone method of CSGObject. Comes with first basic unit tests. More to come | 20:11 |
shogun-notifier- | shogun: Heiko Strathmann :lambday-log-det * 855b1be / / (8 files): https://github.com/shogun-toolbox/shogun/commit/855b1be43183afdde4b7e2edea0e42d867be04ca | 20:11 |
shogun-notifier- | shogun: Merge pull request #1167 from karlnapf/develop | 20:11 |
@HeikoS | here we go | 20:11 |
shogun-notifier- | shogun: enable location info (as in object / function name printing before SGIO | 20:11 |
shogun-notifier- | shogun: prints) with config option to disable / show line and file | 20:11 |
shogun-notifier- | shogun: Soeren Sonnenburg :lambday-log-det * e2142df / src/shogun/io/SGIO.cpp,src/shogun/io/SGIO.h: https://github.com/shogun-toolbox/shogun/commit/e2142df4a17e39d4ec94023be6198c59d370a963 | 20:11 |
shogun-notifier- | shogun: add pretty function macro in defines | 20:11 |
shogun-notifier- | shogun: Soeren Sonnenburg :lambday-log-det * cc2f703 / tests/unit/base/SGObject_unittest.cc: https://github.com/shogun-toolbox/shogun/commit/cc2f703ebeb36d059b632e1881b757d75fcc88ab | 20:11 |
shogun-notifier- | shogun: Merge pull request #1171 from lambday/develop | 20:11 |
shogun-notifier- | shogun: | 20:11 |
shogun-notifier- | shogun: minor fix for eigen3 check in SGObject_unittest | 20:11 |
lambday | so many commits :'( | 20:11 |
@HeikoS | ah man | 20:11 |
@HeikoS | git kills me | 20:11 |
@HeikoS | there should be your branch now | 20:12 |
@HeikoS | rebased against latest develop | 20:12 |
@HeikoS | but since I regulary fail dealing with this, lets do things in develop | 20:12 |
@HeikoS | or try another PR against lambday-log-det | 20:13 |
@HeikoS | https://github.com/shogun-toolbox/shogun/branches | 20:13 |
@HeikoS | its now under merged branches | 20:13 |
lambday | :''( | 20:14 |
@sonney2k | HeikoS, you cannot rebase on published branches | 20:14 |
@HeikoS | sonney2k: I just wanted to create a new feature branch for lambday | 20:14 |
@HeikoS | to push his stuff into | 20:14 |
@HeikoS | so I created it locally | 20:14 |
@HeikoS | from latest develop | 20:14 |
@HeikoS | and then pushed to origin | 20:15 |
@sonney2k | HeikoS, you mean git flow publish feature? | 20:15 |
@HeikoS | sonney2k: no just a new branch for lambday to develop stuff in | 20:15 |
@HeikoS | sonney2k: I though this way easy, but since it is not, I think sending PRs against develop is fine | 20:16 |
@HeikoS | we are just adding stuff anyway | 20:16 |
@sonney2k | HeikoS, well git flow feature start 'feature' | 20:16 |
@sonney2k | git flow feature publish 'feature' | 20:16 |
@HeikoS | ok lets try that | 20:16 |
@sonney2k | HeikoS, but you cannot rebase in this branch | 20:17 |
@sonney2k | you have to work until the very end in this | 20:17 |
@sonney2k | and then once done do the merge to develop! | 20:17 |
@HeikoS | my whole git is now broken :( | 20:17 |
lambday | :( | 20:17 |
@HeikoS | anyway | 20:17 |
@HeikoS | lets work in develop | 20:17 |
@sonney2k | yeah don't rebase on anything you give out! | 20:17 |
@sonney2k | you can only do that locally | 20:17 |
@HeikoS | sonney2k: well that was my plan | 20:18 |
@sonney2k | van51, how is it going? | 20:18 |
@sonney2k | progressing? | 20:18 |
@sonney2k | and hey gsomix? tests finished? | 20:18 |
@HeikoS | sonney2k: can I safely remove the lambday-log-det branch in master repository? | 20:18 |
lambday | HeikoS: I am cancelling the PR then? | 20:18 |
@HeikoS | lambday: yeah please do it in develop | 20:19 |
@HeikoS | we will get into git hell otherwise ;) | 20:19 |
@HeikoS | and I am not the guy who knows the way out of there | 20:19 |
@sonney2k | HeikoS, locally and remotely - just remove it | 20:19 |
@HeikoS | sonney2k, so the one the I see under https://github.com/shogun-toolbox/shogun/branches in merged branches, right? | 20:20 |
votjakovr | HeikoS: are we agreed to use CMachine as a base class for GP? | 20:20 |
gsomix | sonney2k: hey, nope. still preparing for exam. I'll do it little later, at night. | 20:20 |
@HeikoS | votjakovr: it is already | 20:20 |
@HeikoS | its the base class of the current GP class | 20:20 |
@HeikoS | so yes | 20:20 |
@HeikoS | votjakovr: we just need to add an extra abstraction layer | 20:20 |
votjakovr | HeikoS: yep, i understand | 20:21 |
@sonney2k | gsomix, when is your exam again? | 20:21 |
gsomix | sonney2k: 19 June. | 20:21 |
-!- pickle27 [~Kevin@S0106002191dec7e8.cg.shawcable.net] has joined #shogun | 20:21 | |
lambday | crap I just subscribed to this new git thing they were showing and now I can't find the PRs, argh!! | 20:21 |
@sonney2k | gsomix, ok then good luck and lets talk afterwards | 20:21 |
gsomix | sonney2k: ok, thanks. | 20:21 |
lambday | ah found it | 20:21 |
votjakovr | HeikoS: btw why don't we use train() method for inference for example? | 20:22 |
@HeikoS | votjakovr: we do | 20:22 |
@HeikoS | train calls some other methods internally, check CMachine | 20:22 |
@HeikoS | its not overloaded | 20:22 |
@HeikoS | but the submethods are | 20:22 |
@lisitsyn | pickle27: hey | 20:23 |
pickle27 | hey! | 20:23 |
pickle27 | GSoC has begun! | 20:23 |
@lisitsyn | pickle27: yeah | 20:24 |
van51 | sonney2k: it's going ok | 20:24 |
@lisitsyn | pickle27: remember we have to think about some application as a result of your work? | 20:24 |
van51 | sonney2k: i'm fighting a bit with the compiler atm :p | 20:24 |
votjakovr | HeikoS: but why does train_machine() method just return false? | 20:25 |
pickle27 | lisitsyn, yes of course - by application are you meaning something more involved than a really good example right? | 20:25 |
van51 | I've been meaning to ask though, the IO operations should be based on shogun classes and not iostream, right? | 20:25 |
@lisitsyn | pickle27: yes like not just an example but really fancy example | 20:25 |
van51 | because that's what I'm going to start next | 20:25 |
pickle27 | yeah! | 20:26 |
@HeikoS | votjakovr: checking | 20:26 |
pickle27 | lisitsyn, I still like the audio idea and the ekg idea Andreas mentioned but I'll have to see how to make them fancy above just a regular example | 20:26 |
@lisitsyn | pickle27: ECG is really cool | 20:26 |
lambday | HeikoS: sent the PR against develop' | 20:27 |
@lisitsyn | pickle27: well lets just make them audial | 20:27 |
pickle27 | lisitsyn, yeah I think so too | 20:27 |
pickle27 | lisitsyn, I wonder if we could separate a song into its components - that would be cool too | 20:27 |
@lisitsyn | pickle27: yeah once algos are ready we can play a bit I think | 20:28 |
lambday | HeikoS: things would have been better if we could have kept it in a separate branch.. I'll remove these debug msgs once everything is tested... :( | 20:28 |
@HeikoS | lambday: that is alright, as said, we are only adding things | 20:28 |
pickle27 | lisitsyn, do you have a pref for the language the final example is done in? | 20:28 |
pickle27 | I'm going to write my basic examples in c++ obviously but I was thinking python for the finale | 20:28 |
@HeikoS | lambday: sorry for being unable to deal with this git stuff . it just always causes me a headache, debug messages are fine though | 20:29 |
@lisitsyn | pickle27: I think we should go for something web | 20:29 |
pickle27 | okay so probs python or ruby then | 20:29 |
@HeikoS | votjakovr: I am sorry, I was wrong | 20:29 |
@lisitsyn | pickle27: well python is totally ok but there is a need to employ something to play audio etc | 20:30 |
@lisitsyn | pickle27: I think that's crucial to have a web demo everbody can check | 20:30 |
@HeikoS | votjakovr: so training in the GP case means just to compute cholesky stuff | 20:30 |
lambday | HeikoS: no problem :( I suck at git stuffs every time, always checking stackoverflow solns :( | 20:30 |
@HeikoS | votjakovr: currently, train is empty | 20:30 |
@HeikoS | votjakovr: and only apply is called, | 20:31 |
pickle27 | lisitsyn, yeah I like that idea too | 20:31 |
@lisitsyn | lambday: git gets easier once you get the basic idea that branches are homeomorphic endofunctors mapping submanifolds of a Hilbert space | 20:31 |
@HeikoS | votjakovr: so the question is whether we should split training and testing | 20:31 |
votjakovr | HeikoS: exactly! | 20:31 |
@HeikoS | votjakovr: I would say yes | 20:32 |
votjakovr | HeikoS: i agree | 20:32 |
@HeikoS | so train should precompute everything that only needs the training features | 20:32 |
@HeikoS | votjakovr: currently this is all done in a lazy way right? | 20:32 |
lambday | lisitsyn: kill me! kill me now :D | 20:32 |
votjakovr | HeikoS: yep | 20:32 |
@lisitsyn | lambday: that's a famous joke about git | 20:32 |
lambday | lisitsyn: *dies* | 20:33 |
@HeikoS | votjakovr: so which parts do only depend on training data? | 20:33 |
* lambday dies | 20:33 | |
votjakovr | HeikoS: we perform inference on training data | 20:34 |
@HeikoS | votjakovr: no, we also have testing data | 20:34 |
pickle27 | lisitsyn, python and ruby both have nice web frameworks, is that what you had in mind? | 20:35 |
@lisitsyn | pickle27: oh well all we need is to pass some data to javascript actually | 20:35 |
@lisitsyn | can be done in any language you'd like to use | 20:35 |
pickle27 | lisitsyn, oh okay right | 20:35 |
@HeikoS | votjakovr: so for exact inference in regression, training is to compute the cholesky of K+\sigma^2 * I | 20:36 |
votjakovr | HeikoS: for testing data we compute mean and variance using posterior | 20:36 |
@HeikoS | votjakovr: exactly, and for that we need the cholesky | 20:37 |
@HeikoS | votjakovr: so I would say then we should have an abstract method on inference method | 20:38 |
@HeikoS | which corresponds to train | 20:38 |
votjakovr | HeikoS: yep, so train() method should perform inference on training data | 20:39 |
@HeikoS | votjakovr: for the exact inference, this then updates the cholesky and the training kernel matrix | 20:39 |
@HeikoS | votjakovr: thats not really called inference, but yes, do all the stuff on training data | 20:39 |
@HeikoS | votjakovr: and then we can use the CMachine interface for train/apply | 20:39 |
@HeikoS | where apply calls submethods apply_regression etc | 20:39 |
votjakovr | HeikoS: yep, something like compute_posterior() in inference method class ? | 20:40 |
@HeikoS | and the GP-subclasses implement this | 20:40 |
@HeikoS | votjakovr: that sounds good | 20:40 |
@HeikoS | and GP::train_machine then calls this method | 20:40 |
@HeikoS | votjakovr: dont call it compute_posterior, since in classification its different | 20:41 |
@HeikoS | votjakovr: maybe call it just train :) | 20:42 |
@HeikoS | votjakovr: btw we should also refactor the Laplace approximation class, we cannot use the regression one for classification but it shares a lot of code | 20:43 |
votjakovr | HeikoS: but what if i just make an update_all() method public? | 20:45 |
@HeikoS | votjakovr: yeah maybe, we will see about this later I guess | 20:45 |
@HeikoS | votjakovr: first goal is to have GP base class which inherits from CMachine and two subclasses for regression and classification | 20:46 |
votjakovr | HeikoS: Ok! | 20:46 |
pickle27 | lisitsyn, also | 20:47 |
@lisitsyn | pickle27: yes? | 20:47 |
pickle27 | lisitsyn, I was planning on using the shogun PCA class as a sort of template for the various ADJ techniques | 20:48 |
pickle27 | lisitsyn, Im just looking at the PCA class now to see if that makes any sense to do at all lol | 20:49 |
@lisitsyn | pickle27: oh well | 20:49 |
@lisitsyn | PCA is so easy to implement | 20:49 |
@lisitsyn | :) | 20:49 |
pickle27 | lisitsyn, yeah it is lol | 20:49 |
pickle27 | lisitsyn, just seemed like the structure of the class might be a bit similar | 20:49 |
@lisitsyn | pickle27: you are welcome to patch that part of shogun for sure | 20:49 |
pickle27 | but maybe not | 20:50 |
@lisitsyn | pickle27: PCA is also implemented in tapkee of course | 20:50 |
pickle27 | yeah I would happily do some work on the PCA module | 20:50 |
pickle27 | yes I know! | 20:50 |
pickle27 | hmm PCA is still LAPACK in shogun | 20:51 |
@lisitsyn | pickle27: actually if you find it easier you may first implement something apart | 20:51 |
@lisitsyn | pickle27: yes | 20:51 |
pickle27 | that could be another great entrance commit PCA in Eigen3 | 20:51 |
pickle27 | yeah I am planning to implement separate for a while but I still want to use a shogun class | 20:52 |
@lisitsyn | pickle27: I don't know if we should change shogun's PCA to use tapkee | 20:53 |
@lisitsyn | actually I support three possible methods | 20:53 |
@lisitsyn | dense from eigen3, randomized like in redsvd | 20:53 |
@lisitsyn | and arpack | 20:53 |
pickle27 | right well for the time being Im going to focus on ADJ anyways | 20:54 |
@lisitsyn | so may be it makes sense to do a few steps to make it compatible | 20:54 |
@lisitsyn | I'll be back a bit later | 20:54 |
pickle27 | okay! | 20:54 |
-!- lambday [67157e36@gateway/web/freenode/ip.103.21.126.54] has quit [Ping timeout: 250 seconds] | 20:59 | |
shogun-notifier- | shogun: lambday :develop * a536834 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/a536834a2c2e1c80a997cc62bb651c25bb9cd43e | 21:04 |
shogun-notifier- | shogun: class CLinearOperator and DenseMatrixOperator added | 21:04 |
shogun-notifier- | shogun: Heiko Strathmann :develop * 6642f28 / / (4 files): https://github.com/shogun-toolbox/shogun/commit/6642f286d42a6b7c175d7c3a2cd1cba4f5122a76 | 21:04 |
shogun-notifier- | shogun: Merge pull request #1173 from lambday/feature/log_determinant | 21:04 |
shogun-notifier- | shogun: | 21:04 |
shogun-notifier- | shogun: class CLinearOperator and DenseMatrixOperator added | 21:04 |
-!- nube [~rho@49.126.230.83] has quit [Ping timeout: 276 seconds] | 21:05 | |
-!- nube [~rho@49.244.53.255] has joined #shogun | 21:06 | |
shogun-buildbot | build #945 of cyg1 - libshogun is complete: Failure [failed configure] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/945 blamelist: lambday <heavensdevil6909@gmail.com> | 21:07 |
gsomix | operators, operators everywhere | 21:09 |
@HeikoS | gsomix: smooth operators! :) | 21:09 |
@HeikoS | https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&ved=0CDcQtwIwAA&url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D3L3T4tFNxOo&ei=4V6_UaXnHMGqO7K9gPAB&usg=AFQjCNFm45xhogGxX82-g9mygCmsxq7hyA&sig2=cZj210ZL6TC7GQ4Tb-AfXg&bvm=bv.47883778,d.ZWU | 21:09 |
-!- nube [~rho@49.244.53.255] has quit [Ping timeout: 248 seconds] | 21:12 | |
gsomix | HeikoS: cool! | 21:18 |
gsomix | now I know what to tell the teacher about operators on the exam :) | 21:18 |
@HeikoS | gsomix: is your exam also about operators? :) | 21:21 |
gsomix | HeikoS: methods of mathematical physics. there is some operators and eigenproblems related to quantum mechanics examples. | 21:24 |
gsomix | *there are | 21:24 |
@HeikoS | gsomix: nice! | 21:24 |
@HeikoS | good stuff, physicists always have the best ideas, statisticians always steal them ;) | 21:25 |
shogun-buildbot | build #946 of cyg1 - libshogun is complete: Failure [failed compile] Build details are at http://www.shogun-toolbox.org/buildbot/builders/cyg1%20-%20libshogun/builds/946 blamelist: Heiko Strathmann <heiko.strathmann@gmail.com> | 21:26 |
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has quit [Ping timeout: 256 seconds] | 21:31 | |
gsomix | HeikoS: huh. btw, there is no strong course about statistics on my major. only in the form of the statistical physics. it's sad. | 21:32 |
@HeikoS | gsomix: seems to happen anyways | 21:32 |
@HeikoS | We just stole another method from statistical physikcs literatur here :) | 21:32 |
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has joined #shogun | 21:33 | |
-!- nube [~rho@49.244.119.220] has joined #shogun | 21:35 | |
-!- lambday [67157c36@gateway/web/freenode/ip.103.21.124.54] has joined #shogun | 21:38 | |
-!- nube [~rho@49.244.119.220] has quit [Ping timeout: 264 seconds] | 21:42 | |
-!- van51 [~van51@athedsl-408350.home.otenet.gr] has quit [Remote host closed the connection] | 21:52 | |
-!- HeikoS [~heiko@nat-179-180.internal.eduroam.ucl.ac.uk] has quit [Quit: Leaving.] | 22:02 | |
-!- votjakovr [~votjakovr@host-46-241-3-209.bbcustomer.zsttk.net] has left #shogun ["Fallen asleep!"] | 22:08 | |
-!- gsomix [~Miranda@83.234.169.173] has quit [Ping timeout: 245 seconds] | 22:10 | |
-!- lisitsyn [~lisitsyn@83.234.169.173] has quit [Ping timeout: 248 seconds] | 22:10 | |
-!- nube [~rho@49.244.113.132] has joined #shogun | 22:17 | |
-!- lambday [67157c36@gateway/web/freenode/ip.103.21.124.54] has quit [] | 22:18 | |
-!- sonney2k [~shogun@7nn.de] has left #shogun ["Ex-Chat"] | 22:18 | |
-!- lisitsyn [~lisitsyn@83.234.169.173] has joined #shogun | 22:20 | |
lisitsyn | oh HeikoS mentioned sade | 22:23 |
-!- nube [~rho@49.244.113.132] has quit [Ping timeout: 240 seconds] | 22:24 | |
-!- gsomix [~Miranda@83.234.169.173] has joined #shogun | 22:31 | |
-!- nube [~rho@49.244.88.103] has joined #shogun | 22:58 | |
-!- iglesiasg [d58f3226@gateway/web/freenode/ip.213.143.50.38] has joined #shogun | 23:02 | |
-!- nube [~rho@49.244.88.103] has quit [Ping timeout: 246 seconds] | 23:03 | |
-!- nube [~rho@49.244.13.21] has joined #shogun | 23:20 | |
iglesiasg | thanks for the patch hushell, I will check | 23:29 |
-!- nube [~rho@49.244.13.21] has quit [Ping timeout: 252 seconds] | 23:46 | |
--- Log closed Tue Jun 18 00:00:50 2013 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!