IRC logs of #shogun for Sunday, 2011-06-19

« Saturday, 2011-06-18 Index Monday, 2011-06-20 »
--- Log opened Sun Jun 19 00:00:27 2011
--- Day changed Sun Jun 19 2011
@sonney2kheh00:00
blackburn1sonney2k: pull-requested00:01
CIA-32shogun: Sergey Lisitsyn master * r80f768e / (6 files in 2 dirs): Fixed C{ROC,PRC}Evaluation classes - http://bit.ly/l4GC2700:01
CIA-32shogun: Soeren Sonnenburg master * rdfe5bdf / (6 files in 2 dirs): Merge branch 'master' of git://github.com/SergeyLisitsyn/shogun - http://bit.ly/jUAylV00:01
blackburn1answered baozeng00:03
blackburn1sonney2k: guys at scikits-learn doing manifold learning just like me00:09
blackburn1competition! :D00:10
@sonney2kblackburn1, just do the 30+ methods from the link we found somewhere00:11
@sonney2kthen the definitive answer is given :)00:11
-!- f-x [~user@117.192.200.151] has quit [Quit: ERC Version 5.3 (IRC client for Emacs)]00:12
blackburn1sure but now I'm focused on efficiency00:13
@sonney2kblackburn1, you get your arpack00:14
@sonney2kthat should be very easy to do...00:14
blackburn1sonney2k: i'm thinking about more efficient routines00:14
blackburn1http://en.wikipedia.org/wiki/LOBPCG00:16
blackburn1for example00:16
@sonney2kas you wish - simple and working things first - then speed and mem efficient stuff00:18
blackburn1I'm doing it in pace of my plan :)00:18
blackburn1hahaha00:19
blackburn1I lost link with 30+ methods00:19
blackburn1found in logs00:20
@sonney2kbtw scipy has this method implemented00:21
@sonney2kso you can prolly even borrow the code00:21
blackburn1sonney2k: which? LOBPCG?00:22
@sonney2kyes00:22
blackburn1yes, we could00:23
blackburn1wonder why dijkstra is used in LLE00:25
blackburn1in all implementations I have seen00:25
blackburn1for n vertices it is O(n^3) just as floyd-warshall00:26
blackburn1hahha this toolbox is not ideal00:28
blackburn1    % NOTE: Classical scaling is identical to performing PCA, except the00:28
blackburn1    % input data is different. Specifying pairwise similarity data is not00:28
blackburn1    % yet supported by the toolbox.00:28
blackburn1mappedX = compute_mapping(X, 'PCA', no_dims);00:28
@sonney2k1:0 for russia00:29
blackburn1btw LLE is pretty fast now00:32
blackburn1cause it is optimized a little and cholesky solver is used00:33
blackburn1sonney2k: one 1rd year student is interested in shoguning a little :)00:37
@bettybooblackburn1, ;D00:37
blackburn11rd? I wrote 1rd?00:37
blackburn1omg I should sleep00:37
@sonney2kird?00:38
@sonney2k3rd?00:38
@sonney2kor 1st?00:38
@sonney2kblackburn1, do what you can :)00:38
blackburn1ahaha00:38
blackburn11st yes00:38
@sonney2kI will have to sleep now too00:38
blackburn1good night :)00:39
@sonney2kcu00:39
-!- blackburn1 [~blackburn@31.28.40.202] has quit [Quit: Leaving.]00:48
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun10:07
-!- blackburn [~blackburn@31.28.40.202] has quit [Quit: Leaving.]10:30
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun10:30
-!- blackburn [~blackburn@31.28.40.202] has quit [Quit: Leaving.]10:37
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun10:37
-!- blackburn [~blackburn@31.28.40.202] has quit [Quit: Leaving.]11:01
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun11:01
blackburnsonney2k: oh I have started refactoring preprocessor for SGMatrix usage.. sometimes code here sucks12:01
blackburnпиздец..12:18
blackburnI have to refactor simple features too..12:18
@sonney2kblackburn, yeah many functions now need refactoring - basically all that get a float64_t* x, int32_t len argument and ** x ...12:29
blackburnsonney2k: I'll make some temporary hacks now12:29
blackburnbecause now I'm focused on Preprocessor refactoring :)12:29
blackburnsonney2k: would you fix it a bit later?12:29
blackburnnow I'm going to fix all the preprocessors.. this will take a time12:31
@sonney2kthanks12:31
blackburnsonney2k: also I'm going to introduce new class - DimReductionPreprocessor12:32
blackburnit contains target dimension - all of my algorithms works with it12:32
blackburnit could re-dup some code12:33
@sonney2kDimensionReductionPreprocessor12:34
@sonney2kok12:34
-!- blackburn [~blackburn@31.28.40.202] has quit [Quit: Leaving.]13:07
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun13:07
CIA-32shogun: Heiko Strathmann master * r05ae96b / src/libshogun/features/Subset.h : added comment (+5 more commits...) - http://bit.ly/mypxuZ13:16
blackburnsonney2k: that will be a HUGE pull request14:18
-!- blackburn [~blackburn@31.28.40.202] has quit [Quit: Leaving.]14:20
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun14:20
blackburnsonney2k: NormDerivativeLem3 is not implemented, may be we will just KILL it?15:07
blackburnI want to kill somebody after 3 hrs of refactoring15:08
blackburnsonney2k: sorry for such a large pull request16:03
-!- blackburn [~blackburn@31.28.40.202] has quit [Ping timeout: 255 seconds]16:23
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun16:24
@sonney2kblackburn, feel free to kill it yes17:04
@sonney2kbut please don't kill me :D17:04
blackburnsonney2k: you too :) you probably would want to kill me for these commits :D17:04
blackburnmany-many lines17:05
@sonney2kblackburn, yes......17:09
@sonney2kdoes apply_to_feature_matrix still work in-place?17:10
@sonney2kor what does it do?17:10
blackburnehh17:10
blackburnwhat do you mean?17:10
@sonney2kI really wonder why it returns anything17:10
blackburnwell it returns SGMatrix - feature matrix17:10
blackburnbefore it was returning feature matrix too17:10
blackburnbut not of SGMatrix, just float64_t*17:11
@sonney2kthen I did it wrong already before...17:11
blackburnyou prefer not return anything?17:12
@sonney2kI think I wanted it to be applied in-place (at least that is what I remember17:12
@sonney2kfor now I would just merge what you did but one should think about it17:12
blackburnsonney2k: well it is doing preprocessing on features, then returns feature matrix17:15
@sonney2kblackburn, yeah it applies things in-place but then basically calls get-feature_matrix on the preproc'd features17:15
blackburnI didn't it this way17:16
blackburnI set different feature matrix in my algorithms17:16
@sonney2kblackburn, hmmhh it was intended this way - though of course deleting the old matrix / setting a new one can be the only option17:17
CIA-32shogun: Sergey Lisitsyn master * rf8cd650 / (22 files): SGMatrix and SGVector usage in Preprocessor - http://bit.ly/imXP6K17:18
CIA-32shogun: Sergey Lisitsyn master * r8a0e983 / (7 files in 2 dirs): Fixed naming {preproc}->{preprocessor} - http://bit.ly/mr8oJj17:18
CIA-32shogun: Sergey Lisitsyn master * r6b70388 / (18 files): Various fixes for examples - http://bit.ly/kiGzgQ17:18
CIA-32shogun: Sergey Lisitsyn master * r0c14232 / (2 files): Rearranged modular config for Preprocessor - http://bit.ly/l7Bhau17:18
@sonney2kas long as the test suite still runs through - it is all good17:18
blackburnthere are some errors but it wasn't related to my commits17:18
CIA-32shogun: Soeren Sonnenburg master * ra26d599 / (3 files in 2 dirs): remove dysfunctional/not-implemented preproc - http://bit.ly/kPg4d517:19
@sonney2kblackburn, I mean the preproc tests17:19
blackburnah I see17:19
@sonney2kthere are probably more errors still :)17:20
blackburnsonney2k: but how can we resize feature matrix?17:20
blackburnI mean I can work on given feature matrix but how can I free it partially after17:21
blackburne.g. 1000d feature matrix -> 2d feature matrix, 998 unused rows17:21
@sonney2kblackburn, I usually didn't - I just left the bigger memory chunk lying around and set new feature matrix sizes17:21
blackburndon't you think it is bad?17:22
@sonney2kI know not optimal but sometimes it was impossible to store the preproc'd data because it was just too big17:22
@sonney2ksay feature matrix twice17:22
@sonney2k(think of norm one or log(1+x) transform)17:23
blackburnno problem for preprocessor not changing dimensionality17:23
blackburnI think about my algorithms :)17:23
blackburnI can't see any way to reduce matrix size..17:23
blackburnsonney2k: do you know any?17:25
blackburnit is ok if it will broke matrix elements, I just want to reduce its size17:25
@sonney2kblackburn, then you will have to delete[] the features feature_matrix and set a new one (or jsut set a new one - I hope that should take care of the delete[])17:26
@sonney2kblackburn, it can still happen that you run out of memory17:26
blackburnbad17:27
@sonney2kthe only way I see is to use SG_MALLOC / SG_REALLOC17:27
blackburnI think about realloc17:27
blackburnsome SGMatrix::resize that doing realloc17:27
@sonney2khowever that will reduce memory on many systems but not OSX (!)17:27
blackburnwhat's wrong with osx?17:28
@sonney2kfor this to work we would need to get only SG_MALLOC'd matrices - no longer new[]'d ones17:28
blackburnis there some problems with new?17:29
@sonney2kblackburn, one cannot do REALLOC on new'd memory17:32
blackburnah17:32
blackburnI see17:32
blackburnsonney2k: anyway there are (number_of_examples)^2 matrix in both LLE and MDS17:34
blackburnso it is very memory consumpting at all17:35
blackburnwith realloc or without17:35
@sonney2kheh :)17:35
@sonney2kI guess then it does not really matter that much17:35
blackburnand there is no way to not use such large matrices17:36
@sonney2ksounds like precomputed kernel matrices17:38
@sonney2kI guess there always is a way but it is becoming computationally demanding17:38
blackburnthe only way is to use another method :D17:39
blackburnit is symmetrical17:39
blackburnI could use some fancy sym matrix class of course17:40
blackburnbut lapack doesn't work with it17:40
blackburnmake[1]: *** No rule to make target `../shogun/preprocessor/NormDerivativeLem3.h', needed by `Preprocessor_wrap.cxx'.  Stop.17:43
blackburnI've already made distclean!17:43
blackburnsonney2k: does it compiles on your machine?17:45
blackburnsonney2k: Line 13, modular/Preprocessor_includes.i17:47
blackburnsonney2k: seems I should use arpack++ not arpack..18:20
blackburnI would rather use arpack but there is no wrappers to fortran code, I don't sure I can do it :)18:21
blackburnmay be some 'Writing wrappers for Fortran 77 for dummies' book exists?18:23
-!- blackburn1 [~blackburn@31.28.40.202] has joined #shogun18:38
-!- blackburn [~blackburn@31.28.40.202] has quit [Read error: Connection reset by peer]18:39
serialhexblackburn1: fortran??19:40
blackburn1serialhex: I'd like to use arpack :) it is written in fortran 7719:42
blackburn1lapack too btw19:42
serialhexreally??  hmm  *is kind of afraid of old code*19:43
blackburn1serialhex: it is simply faster :)19:43
serialhexthat's cool, but still, who uses fortran anymore??? or fortran 77 for that matter!?!?19:43
serialhexdidn't that die out with the last of the dodo's??19:44
blackburn1hey but why should it die?19:44
blackburn1:)19:44
serialhexi dunno... i'm just wondwering19:45
serialhexso whats going on with you?19:45
blackburn1fortran compilers are optimized, safety, blabla19:45
blackburn1they are doing it for 40+, it is almost bug-free and so on19:46
blackburn1moreover some loops vectorized automatically and there are a lot of opportunities to make code parallelized19:46
blackburn1today have been doing some refactoring, it was awful19:47
serialhexcool cool... my objection was mostly in jest, at least it's not using cobol :P19:47
serialhexooh, refactoring pains...19:47
blackburn1had to introduce some new class and some naming issues19:48
blackburn1and float64_t* -> SGVector19:49
serialhexhmm, sounds like fun!19:49
blackburn1serialhex: have you seen pictures in my weekly report? :)19:50
serialhexyeah, the spooky faces & the roll :P19:50
blackburn1that's fun - refactoring not :D19:51
@bettybooha ha19:51
CIA-32shogun: Soeren Sonnenburg master * rbbd6767 / src/modular/Preprocessor_includes.i : really remove all occurences of NormDerivativeLem3 - http://bit.ly/loslKS19:51
blackburn1sonney2k: thanks19:52
blackburn1I didn't want to make a pull request with 1-liner :)19:52
@sonney2kblackburn1, thanks for pointing it out19:52
blackburn1thankfulness on shogun19:52
@sonney2kblackburn1, getting arpack wrappers should be easy19:52
@sonney2kwhat are the funcitons you need?19:52
blackburn1sonney2k: yes? how?19:53
blackburn1I don't know exactly19:53
blackburn1yet19:53
blackburn1there are no headers - that's a problem. or not?19:53
@sonney2kblackburn1, so? just create the header like in lib/lapack.h19:54
@sonney2kblackburn1, btw I guess your refactoring also broke the static interfaces (received complaints...)19:54
* sonney2k testing19:55
blackburn1just preproc - preprocessing issue I guess19:55
@sonney2kblackburn1, look at the fortran function signatures19:55
@sonney2kI guess so too :)19:55
@bettyboohihi19:55
blackburn1I just don't use static interfaces :)19:56
blackburn1how it works?!19:56
blackburn1we can just call fortran function?19:56
@sonney2kblackburn1, the same way as linking to any dynamic library works - we just need a .so file with the function entry point and matching signature19:59
@sonney2kblackburn1, I think I fixed the compile errrors19:59
blackburn1wow, it is just like I had enquired how child is born :D20:00
blackburn1surprise for me, never knew how it works20:00
blackburn1sonney2k: preproc?20:01
@sonney2kyup20:01
blackburn1but who uses up-to-date static shogun??20:01
@sonney2khmmhh no news from alesis :(20:01
@sonney2kblackburn1, not me :)20:01
CIA-32shogun: Soeren Sonnenburg master * rbee38a5 / (2 files): fix compilation errors in libshogunui caused by preproc refactoring - http://bit.ly/jQnzoO20:02
blackburn1I thought libshogunui is dead completely :)20:03
@sonney2kblackburn1, so which arpack functions do you need?20:08
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun20:11
-!- blackburn1 [~blackburn@31.28.40.202] has quit [Read error: Connection reset by peer]20:11
@sonney2kblackburn, so which arpack functions do you need?20:11
blackburn10 minutes :)20:14
blackburndsaupd I guess20:17
blackburnhmm. they have bad documentation20:18
@sonney2kblackburn, I guess that will help you http://www-heller.harvard.edu/people/shaw/programs/dsaupd.h20:20
@sonney2klook at the extern "C" line20:20
blackburnah!20:21
blackburnI see20:21
blackburnso I should add this to arpack.h in lib/?20:21
blackburnnice20:22
@sonney2kyes20:23
blackburnhow simple20:23
@sonney2kand some minimal code to call that function20:23
blackburnfor .configure?20:23
@sonney2kwell that dsaupd function has a trillion arguments...20:23
@sonney2kyes20:23
blackburnI see20:23
@sonney2kso non-trivial to get it working20:23
blackburnyes, many arguments, just like in some lapack routines20:23
@sonney2kwell lapack has really a lot less...20:24
@sonney2k16 arguments!20:24
* sonney2k counted20:24
blackburnwhere are drivers20:24
blackburnand three 'work' arrays20:24
blackburnand info20:24
blackburnand some params and strategies20:25
@sonney2kblackburn, did I understand this right - dsaupd is for sparse matrices?20:26
blackburnseems no20:27
blackburndseupd for eigenvalues20:28
blackburnit is some postprocessing routine, I don't know for sure20:28
blackburnah20:28
blackburnyes20:28
blackburnright20:28
serialhexwow... shogun has a total of 2,135,434 loc :P20:31
blackburnloc?20:31
serialhex1,753,324 is in C++20:31
serialhexlines of code20:31
blackburnah20:31
blackburnI see20:31
serialhexyeah20:31
blackburnwait, how?20:32
blackburnit has 130+20:32
blackburnK20:32
serialhex1,753 kloc :P20:33
@bettyboohehe!?!20:33
serialhexyes bettyboo!  insane!!!20:33
@bettybooserialhex: eh.. what you mean?20:33
blackburnehh? it cant be true :)20:33
blackburnhttp://www.ohloh.net/p/shogun/analyses/latest20:34
serialhexhttps://gist.github.com/1034570  thats the results i got from running http://cloc.sourceforge.net/20:34
serialhexor, more specifically: cloc --by-file-by-lang --exclude-dir=.git .20:35
blackburnsomething is wrong, it can't be true for sure20:36
serialhexmaybe it included the .git dir accidentially?20:36
blackburnmay be, or shogun-data :)20:36
blackburnbecause linux kernel have ~14M20:37
serialhexmaybe20:37
blackburnif shogun was ~2M lines of code sonney2k would be happy :D20:38
@bettyboo<:*)20:38
@sonney2kblackburn, I don't think so20:38
@sonney2kthen it would never compile20:38
@sonney2kserialhex, I guess you should do git clean -dfx before doing so20:39
serialhexoh, der that might help :P20:39
serialhexahh yeah, much more reasionable!20:40
serialhexhttps://gist.github.com/103457020:40
serialhex74,644 loc in C++20:42
serialhexwith 134,648 loc total20:43
serialhexmuch more sane!! :D20:43
blackburnsonney2k: I imagine I we had 2M loc and doing ST* -> SGVector refactoring :D20:44
serialhexhmm... though this gives me a different number than the one you sent me to blackburn... idk, different was of counting i guess :P20:44
blackburnmay be it is not actual at ohloh20:46
blackburnObjective C                       6             23              0            38520:47
blackburnhahah20:47
blackburnshogun for iPhone™20:47
-!- blackburn [~blackburn@31.28.40.202] has quit [Quit: Leaving.]21:01
-!- blackburn1 [~blackburn@31.28.40.202] has joined #shogun21:01
-!- blackburn1 [~blackburn@31.28.40.202] has quit [Quit: Leaving.]21:13
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun21:13
-!- blackburn [~blackburn@31.28.40.202] has quit [Read error: Connection reset by peer]21:15
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun21:16
-!- blackburn [~blackburn@31.28.40.202] has quit [Ping timeout: 240 seconds]21:21
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun21:22
-!- blackburn [~blackburn@31.28.40.202] has quit [Ping timeout: 255 seconds]21:28
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun21:29
@sonney2kblackburn, it did run on Gunnar's iphone btw...21:38
blackburnshogun on iphone? :D21:38
blackburnfor what?? :)21:39
@bettyboo8)21:39
blackburnsonney2k: I was surprised gunnar answered with some praise about my progress :)21:39
@sonney2kme too21:39
@sonney2kI thought he has -inf time and doesn't read mail :)21:40
@bettyboo<:*)21:40
@sonney2kmaybe he is on holidays :D21:40
-!- blackburn [~blackburn@31.28.40.202] has quit [Ping timeout: 255 seconds]21:44
-!- blackburn [~blackburn@31.28.40.202] has joined #shogun21:47
blackburndamn I'm tired with this wi-fi21:51
blackburnI guess something is wrong with driver for adapter on other machine21:51
blackburnsonney2k: do we plan to integrate some 'online' for preprocessors too?21:56
blackburnI have some ideas how to make my algos 'online' but not sure21:58
@sonney2knot a top priority22:00
blackburnsonney2k: what about apply_to_feature_matrix?22:11
blackburnshould it return nothing?22:12
@sonney2kwaht about it?22:12
blackburnyou said it is wrong it returns anything22:12
@sonney2kyeah22:30
@sonney2kyou should check if this is used though22:30
blackburnactually nothing bad with it but may be it should return pointer to given features22:35
blackburni.e. it is _modifying_ function and it should be emphasized somehow22:36
@sonney2kblackburn, up to you...22:37
blackburnwill think about it22:38
blackburnbtw naming sucks22:38
blackburnapply_to_feature_matrix but features is passed :)22:38
@sonney2kyou mean apply() only?22:39
blackburni don't know but naming is wrong - apply_to_features(features) is ok, but apply_to_feature_matrix(features) looks strange22:39
blackburnmay be apply(SGMatrix or SGVector) should be better22:40
blackburnjust like in ->machine transition :)22:40
@sonney2kbut that does not make sense for preproc22:40
@sonney2kbecause things can be strings / sparse matrix etc22:40
blackburnwell I understand why it is called apply_to_feature_matrix but it could be confusing22:41
@sonney2kwe had the apply_to_example or so function to be able to apply the preproc to just one example22:41
@sonney2kand matrix to the whole set of examples22:41
@sonney2kblackburn, you are now the preproc master - so naming is up to you22:43
blackburnyeap :)22:44
@sonney2kI am thinking of removing all the legacy typemaps now22:44
@sonney2kotherwise it is too hard to see where examples fail22:44
@sonney2kdammed I will just do it22:45
@sonney2kthis reminds me that I still have to fix r_modular / octave_modular typemaps.22:45
blackburnyou mean all these float32_t** dst ..?22:47
@sonney2kyes22:53
@sonney2kdie22:54
@sonney2kfloatxxx** bla22:54
@sonney2kall these modular/*.i files that do apply -> die22:54
blackburnhehe22:58
blackburnlet's make a funeral22:58
blackburntragic moment22:58
@sonney2kblackburn, if you'd like to help go wild and over all the .h / .cpp files and replace these doulbe* or ** thingies with SGVector/SGMatrix,SGStringList, SGSparseMatrix / SparseVector SGNDArray23:13
@sonney2kI erased all typemaps now and am now porting octave / r_modular to new datatypes23:14
blackburnsonney2k: ok, will do tomorrow23:14
blackburnah, damn it, will do now23:23
@sonney2kalright we have octave typemaps23:24
blackburnsonney2k: oh hehe it is much code to change23:24
@sonney2kyes23:25
@sonney2kdo file after file please23:25
@sonney2ksuch that we can parallelize23:25
@sonney2kI will notify you when I have the r_typemaps23:25
@sonney2kthen I join you23:25
blackburnsonney2k: I will do classifier now23:25
blackburnjust after get landmarkmds prototype to work :)23:26
blackburnsonney2k: just interface issues now, not guts of algos, right?23:33
@sonney2konly interfaces for now23:34
blackburnsonney2k: CKNN::classify_for_multiple_k23:35
blackburnphew!23:36
@sonney2k?23:36
blackburndo we need it?23:36
@sonney2ksure23:36
blackburnit works strange way23:36
@sonney2kthis is cool for knn model selection23:36
blackburnnot returning labels but some array23:36
@sonney2kI mean you run knn once with large k and get the results for free for all23:37
@sonney2kk23:37
@sonney2kk < max_K23:37
blackburnyes I understand23:37
blackburnokay will adapt it23:37
blackburnhm it is simple23:38
blackburnhehe23:38
@sonney2kcompiling for libshogun,libshogunui,r,matlab,octave,python,r_modular,octave_modular,ruby_modular,csharp_modular,python_modular,java_modular23:44
blackburnhahaha23:44
blackburnbetter go sleep, at morning it will be done23:44
blackburnknn done23:46
@sonney2kblackburn, let me have a look at kernels in the meantime23:49
blackburnsonney2k: could you please do it for distributions?23:49
@sonney2kor distributions if you want23:49
blackburnand classifier/PluginEstimate23:50
blackburnbecause too much code for me, I think it is easier for you23:50
@sonney2khmmhh alesis didn't send any email - I wonder what happened....23:51
@sonney2klast time I met him on IRC he was pretty enthusiastic23:51
blackburnbad23:51
blackburnWTF is classifier/svm/Tron??23:53
@sonney2ksome liblinear solver23:53
blackburnis it used somewhere externally?23:53
@sonney2kno23:54
@sonney2k(I think)23:54
blackburnI hope so23:54
blackburnI've never seen such23:54
blackburnoh23:54
@sonney2khmmhh alesis did his last commit exactly 1 week ago23:55
blackburni won't sleep this night :D23:55
blackburnsonney2k: bad..23:56
@sonney2kblackburn, he might have had a good reason... hope he is well23:57
blackburnsonney2k: his mentor is in trip, right?23:57
blackburnsonney2k: just write him email, no any suggestion23:57
@sonney2kyes - his mentor (vojtech) just returned23:57
@sonney2kdid already23:58
blackburnoh this solvers make me mad23:59
--- Log closed Mon Jun 20 00:00:01 2011
« Saturday, 2011-06-18 Index Monday, 2011-06-20 »

Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!