IRC logs of #shogun for Thursday, 2014-01-09

« Wednesday, 2014-01-08 Index Friday, 2014-01-10 »
--- Log opened Thu Jan 09 00:00:43 2014
-!- FSCV [~FSCV@50.7.50.60] has quit [Quit: Leaving]00:56
-!- iglesiasg [~iglesiasg@524AE0A7.cm-4-3d.dynamic.ziggo.nl] has quit [Quit: Leaving]01:45
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]02:16
-!- thoralf [~thoralf@91-65-138-245-dynip.superkabel.de] has quit [Quit: Konversation terminated!]03:52
shogun-buildbotbuild #675 of nightly_default is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/nightly_default/builds/67504:09
-!- sonne|osx [~sonne@89.204.130.106] has joined #shogun08:11
-!- sonne|osx [~sonne@89.204.130.106] has quit [Quit: sonne|osx]08:35
-!- gsomix [~gsomix@185.6.237.115] has quit [Ping timeout: 251 seconds]09:15
-!- gsomix [~gsomix@85.26.164.130] has joined #shogun10:11
-!- gsomix [~gsomix@85.26.164.130] has quit [Read error: Connection reset by peer]10:29
-!- lisitsyn [~lisitsin@mxs.kg.ru] has joined #shogun10:54
besser82sonne|work:  hey!10:56
besser82sonne|work:  just seen Orion has done a PR  ;)10:56
besser82sonne|work:  He's one from da Fedora-Crew, too  ;)10:58
sonne|workbesser82: heh :)10:59
sonne|workbesser82: give us more :D10:59
sonne|workbesser82: and seen the python3 buildbot?10:59
besser82sonne|work:  seen it!11:08
besser82sonne|work:  Doing my best to recruit fresh devs for da shogun  :D11:08
sonne|workbesser82: I was confused - didn't you 2to3 your changes?11:08
besser82sonne|work:  2to3 is somewhat basic stuff11:08
sonne|workbesser82: I mean does doxy2swig work then or not?11:09
besser82sonne|work:  no, no with 2to3, but I got a ver from adrin which works ~98%11:09
besser82sonne|work:  just need to add some decode / encode magixx to it  ;)11:09
sonne|workbesser82: ahh cool11:10
besser82sonne|work:  and that's what I just want to get started  :D11:10
besser82sonne|work:  and fixing version.cmake, off course11:10
sonne|workbesser82: do it :)11:10
besser82sonne|work:  then I'll start to dig deeper into those failing examples11:11
sonne|workbesser82: first the compile then the tests then the examples and then the world ;)11:11
sonne|workbesser82: so what do you do instead of my hack:11:12
sonne|workimport os.path,time;print time.strftime('%Y-%m-%d %H:%M', time.gmtime(os.path.getmtime('${ROOT_DIR}/NEWS' in version.cmake?11:12
besser82sonne|work:  yes11:13
besser82sonne|work:  print is not a statement in Py3, it's a function11:13
besser82sonne|work:  so slamming brackets around the portion after `print` should do tha trick11:14
besser82sonne|work:  I hope at least11:14
sonne|workbesser82: well I meant more the if this can be done with cmake but sure that would sufficiently work11:14
besser82sonne|work:  such operations cannot be done with cmake, strftime, i mean...11:15
besser82sonne|work:  that's why cmake has such option to invoke external-command and pass stdout from that into a var  ;)11:16
@wikingthere's a swig PR somewhere that does doxy2swig but unfortunately hasn't been merged (and seems never will be)11:39
sonne|workbesser82: ok just tried - changing brackets is sufficient11:45
besser82wiking:  so it wouldn't help us, right?11:46
besser82sonne|work:  kk, will do && push11:46
sonne|workbesser82: yes please do ... I cannot from here11:46
-!- shogun-notifier- [~irker@7nn.de] has joined #shogun11:51
shogun-notifier-shogun: Bj?rn Esser :develop * 148ddb2 / cmake/version.cmake: https://github.com/shogun-toolbox/shogun/commit/148ddb20cad3cf749c9ea7ed8914f48d0387be7211:51
shogun-notifier-shogun: make version.cmake Python3-aware11:51
besser82sonne|work:  just pushed ^^^  directly to develop-branch, since that was an `approved` change  ;)11:52
sonne|workshogun-buildbot: force build --branch develop 'deb4 - python3'11:52
shogun-buildbotbuild #3 forced11:52
shogun-buildbotI'll give a shout when the build finishes11:52
sonne|workbesser82: sure thing11:52
sonne|worklets see11:52
sonne|workshogun-buildbot: go go!11:52
shogun-buildbotWhat you say!11:52
shogun-buildbotbuild #3 of deb4 - python3 is complete: Failure [failed python static]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/311:52
sonne|workbesser82: see? he obays11:52
besser82sonne|work:  yes  :D11:52
sonne|workbesser82: alright now my misconfiguration11:53
sonne|workwiking: could you help fixing it?11:53
@wikingsonne|work: fix what?11:53
sonne|workwiking: deb4 - python311:54
@wikingsonne|work: and somebody will write for me my thesis i guess :)11:54
sonne|workwiking: only if you want to become a german politician11:54
@wikingsonne|work: :P11:55
@wikingsonne|work: make python_static11:55
@wikinginstead of _python_static11:55
sonne|workwiking: I cannot login from here...11:55
@wikingic11:55
besser82sonne|work:  is you sure those crufty legacy static interf will build!?!?11:56
besser82wiking:  me cannot log either  ;)  Not in auth_keys  ;)11:56
sonne|workbesser82: no but if not then we will see and fix!11:56
-!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Quit: buildmaster reconfigured: bot disconnecting]11:56
-!- shogun-buildbot [~shogun-bu@7nn.de] has joined #shogun11:56
@wikingshogun-buildbot: force build --branch develop 'deb4 - python3'11:56
shogun-buildbotbuild #5 forced11:56
shogun-buildbotI'll give a shout when the build finishes11:56
sonne|workbesser82: I thought you have a sufficiently huge botnet ;)11:56
besser82sonne|work:  shouldn't we somewhen start to drop them static shiatz  ;)11:57
shogun-buildbotbuild #5 of deb4 - python3 is complete: Failure [failed python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/511:57
@wikingsonne|work: there u go11:57
besser82sonne|work:  :-P  No bot-net here  ;)  Just busy workerz on ~20 Servers  :D11:57
sonne|workwiking: that was lightning fast11:58
sonne|workbesser82: I would love to but there are some users on *static still11:58
sonne|workbesser82: so now we need the doxy2siwg fixes11:58
besser82sonne|work:  on it...  Just a few minutes...11:58
besser82sonne|work:  that shit static thing chokes on Python3  :-P11:59
sonne|workbesser82: not on the buildbot at least :D12:00
besser82sonne|work:  then let's announce a drop of static for next major-release12:00
sonne|workbesser82: that would kill our matlab / r support :/12:01
sonne|workbesser82: but for me fine12:01
@wikingsonne|work: r_modular we have12:01
sonne|workr_modular works so so yes12:01
besser82sonne|work:  i just wanted to say that12:01
@wikingsonne|work: and seems to be working apart from travis... which needs a little bit more love12:01
besser82sonne|work:  and matlab could be done from octave_modular, I guess?!?12:01
@wikingbesser82: could it?12:02
sonne|workbesser82: no12:02
@wikingi think we have to stick with matlab static12:02
sonne|workthere were tons of attempts of providing swig wrappers12:02
sonne|workbut no ones does it12:02
@wikingbut currently it's not being built in the right way... dunno how to generate that mex file12:02
sonne|workwiking: well if we stick to one interface we have to keep the others12:02
sonne|workotherwise people can hardly migrate12:02
@wikingsonne|work: yeah we have a problem with matlab since cmake switch as i never managed to figure out where the mex file is coming from12:03
sonne|worksince I no longer have matlab I cannot fix it12:04
@wikingwell that's ok just somebody tell me how to generate the mex file :)12:05
sonne|workwiking: well I run the same commands mex would do ... so either run mex or run the commands mex -v would12:10
shogun-notifier-shogun: Bj?rn Esser :develop * 4efb5d8 / src/.doxy2swig.py: https://github.com/shogun-toolbox/shogun/commit/4efb5d8d0ee41fd738cf81c96ccad8d999aca63b12:11
shogun-notifier-shogun: pull in the modified doxy2swig provided by adrin12:11
besser82sonne|work:  this ^^ is not guranteed to work, but will bring us a nice step further12:12
besser82sonne|work:  I still need to add some decode/encode magixx to get it working12:13
besser82shogun-buildbot: force build --branch develop 'deb4 - python3'12:13
shogun-buildbotbuild #6 forced12:13
shogun-buildbotI'll give a shout when the build finishes12:13
besser82sonne|work:  buildbot now is @install target  :D12:27
shogun-buildbotbuild #6 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/612:37
besser82sonne|work:  just 3 fails  :D12:40
besser82sonne|work:  in tests, but some still need a fix  :(12:40
besser82sonne|work:  like da print-function and stuff12:40
besser8299% tests passed, 3 tests failed out of 53612:41
besser82Total Test time (real) =  30.27 sec12:41
besser82The following tests FAILED:12:41
besser82629 - python_modular-mathematics_linsolver_cg (Failed)12:41
besser82661 - python_modular-mathematics_logdet (Failed)12:41
besser82696 - python_modular-structure_dynprog_modular (Failed)12:41
besser82Errors while running CTest12:41
besser82sonne|work:  ^^^  :D12:41
besser82sonne|work:  math-tests are because of that array-size fail known from travis12:42
shogun-buildbotbuild #7 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/7  blamelist: Bj?rn Esser <bjoern.esser@gmail.com>12:43
-!- travis-ci [~travis-ci@ec2-50-16-20-105.compute-1.amazonaws.com] has joined #shogun12:48
travis-ci[travis-ci] it's Bj?rn Esser's turn to pay the next round of drinks for the massacre he caused in shogun-toolbox/shogun: http://travis-ci.org/shogun-toolbox/shogun/builds/1664595412:48
-!- travis-ci [~travis-ci@ec2-50-16-20-105.compute-1.amazonaws.com] has left #shogun []12:48
-!- lisitsyn1 [~lisitsin@mxs.kg.ru] has joined #shogun13:29
-!- lisitsyn [~lisitsin@mxs.kg.ru] has quit [Ping timeout: 264 seconds]13:31
sonne|workbesser82: excellent!13:32
besser82sonne|work:  but it looks like class_list.cpp.py needs some changes too, at least on Fedora13:39
sonne|workbesser82: could you please open tests/integration/python_modular/tester.py13:40
sonne|workand change the print in line 40?13:40
besser82sonne|work:  sure!  would that be all?!?13:40
sonne|workit should give us some insight why things fail13:40
sonne|workbesser82: no idea check for further prints or 2to3 even...13:40
sonne|workbesser82: yeah use 2to313:41
besser82sonne|work:  over the whole tree {test, examples} ?!?13:42
sonne|workerrm for tester.py13:42
besser82sonne|work:  kk, a minute13:42
sonne|workand generator.py too13:42
sonne|workand tests/integration/python_modular/classifier.py13:43
sonne|workbesser82: that's it13:45
shogun-notifier-shogun: Bj?rn Esser :develop * c03d93a / tests/integration/python_modular/tester.py: https://github.com/shogun-toolbox/shogun/commit/c03d93af303ea686a95031cf7779ba18af080d1013:49
shogun-notifier-shogun: 2to3ized python_modular-intergration-tests13:49
sonne|workshogun-buildbot: force build --branch=develop 'deb4 - python3'13:49
shogun-buildbotbuild #8 forced13:49
shogun-buildbotI'll give a shout when the build finishes13:49
besser82sonne|work:  done;  only tester.py needed some mod on line40; the rest seemed to be good  ;)13:50
sonne|workbesser82: hmmhh python static tests are b0rken completely - the buildbot doesn't run any13:50
besser82sonne|work:  that static-legacy is last ones on my list ;)13:50
besser82sonne|work:  let's get the modular ones && all other stuff running13:51
besser82sonne|work:  then fix legacy13:51
besser82sonne|work:  with the doxy2swig from adrin, one needs to have a utf8-locale set...13:52
besser82sonne|work:  will fix that during today (encode/decode magixx)13:52
shogun-buildbotbuild #8 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/813:54
besser82sonne|work:  that `EXCEPTION 'str' does not support the buffer interface` is easy to fix  ;)14:00
shogun-buildbotbuild #9 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/9  blamelist: Bj?rn Esser <bjoern.esser@gmail.com>14:00
besser82sonne|work:  will take care of that after lunch14:01
sonne|workbesser82: thx14:02
besser82sonne|work:  but what really makes me wonder is the fact sth. is going wrong in da test, but it still passes  :S14:16
besser82sonne|work:  like an exception gets thrown and that is still *PASS*14:16
sonne|workbesser82: I think the old integration* tests fail14:25
sonne|workrest seems good I think14:25
besser82sonne|work:  they actually fail, yes, but ctest reports them as *PASS* anyways14:33
besser82test 9714:34
besser82        Start  97: integration-python_modular-tester-statistics_kmm14:34
besser8297: Test command: /usr/bin/cmake "-DEXECUTABLE=/usr/bin/python3" "-DFILE=/home/buildslave/deb4_-_python3/build/tests/integration/python_modular/tester.py" "-DARGS=statistics_kmm.py" "-DGDB_COMMAND=/usr/bin/gdb" "-DGDB_SCRIPT=/home/buildslave/deb4_-_python3/build/src/.gdb" "-P" "/home/buildslave/deb4_-_python3/build/cmake/checkgdb.cmake"14:34
besser8297: Test timeout computed to be: 9.99988e+0614:34
besser8292: LibLinear yes14:34
besser8292: classified: 4.720237e-10, bias: 1.043610e-14 <--- accuracy: 1.000000e-0414:34
besser8296: classified: 4.860116e-09, bias: 1.774136e-13 <--- accuracy: 1.000000e-0414:34
besser8297: statistics_kmm.py setting 1/1                                EXCEPTION 'str' does not support the buffer interface14:34
besser8297:14:34
besser8297: The following tests failed!14:34
besser8297:  s14:34
besser8297:14:34
besser8297: Detailled failures:14:34
besser8297:14:34
besser8297:  s14:34
besser8297: ====DIFF================14:34
besser8297: --- expected14:34
besser8297:14:34
besser8297: +++ got14:34
besser8297:14:34
besser8297: @@ -1 +1 @@14:35
besser8297:14:35
besser8297: -a14:35
besser8297: +t14:35
besser8297: ====EOT================14:35
besser8297:14:35
besser8297:14:35
besser8297:14:35
besser8297:14:35
besser82 94/536 Test  #92: integration-python_modular-LibLinear_023_1en05_1_True ...............................................................................   Passed    0.20 sec14:35
besser82 95/536 Test  #96: integration-python_modular-SVMLin_15_1en05_1_True ...................................................................................   Passed    0.18 sec14:35
besser82 96/536 Test  #97: integration-python_modular-tester-statistics_kmm ....................................................................................   Passed    0.24 sec14:35
besser82sonne|work:  like this ^^^  :S14:35
besser82sonne|work:  the test failed, but ctest says *PASS*  That's weird and may lead to unforeseeable errors...14:35
besser82sonne|work:  lemme check tester.py htat might be the key...14:36
sonne|workbesser82: but these are only the old integration-* tests14:36
besser82sonne|work:  so not important to py_modular?!?14:36
sonne|workbesser82: so not the newer ones we do with generator / tester.py14:36
besser82sonne|work:  shouldn't we fix that anyways?14:38
besser82sonne|work:  I mean the behaviour of ctest *PASS* / *FAIL*14:41
sonne|workbesser82: well lets say these are legacy tests that are worth keeping again because they run the exact same stuff the static interfaces run in these regression tests14:42
sonne|workbesser82: so yes14:42
besser82sonne|work:  kk, will do so too14:43
besser82sonne|work:  is there any special reason for those legacy integration tests are executed like this:14:50
besser82/usr/bin/cmake "-DEXECUTABLE=/usr/bin/python3" "-DFILE=/home/buildslave/deb4_-_python3/build/tests/integration/python_modular/tester.py" "-DARGS=statistics_kmm.py" "-DGDB_COMMAND=/usr/bin/gdb" "-DGDB_SCRIPT=/home/buildslave/deb4_-_python3/build/src/.gdb" "-P" "/home/buildslave/deb4_-_python3/build/cmake/checkgdb.cmake"14:50
besser82sonne|work:14:50
besser82sonne|work:  ^^^14:50
besser82sonne|work:  the newer ones are run like python ${test}...14:51
sonne|workerrm maybe the reason was that we run them through gdb + backtrace if the test fails14:51
besser82sonne|work:  I see14:52
besser82sonne|work:  let me find a solution ...14:52
besser82sonne|work:  easy to archive, thing  ;)14:59
besser82sonne|work:  I just need to add MESSAGE( FATAL_ERROR "FAIL" ) to checkgdb.cmake script  ;)15:00
sonne|workok...15:00
besser82sonne|work:  that would reflect an errorlevel !=0 by the cmake-exec  ;)15:00
besser82sonne|work:  which would make ctest recognize that fail15:01
shogun-notifier-shogun: Bj?rn Esser :develop * 02dead1 / cmake/checkgdb.cmake: https://github.com/shogun-toolbox/shogun/commit/02dead112c379250676b5f12b7f3f2b2f2d421d015:03
shogun-notifier-shogun: reflect a failing integration-test to ctest15:03
-!- Saurabh7 [~Saurabh7@117.200.191.191] has joined #shogun15:04
-!- lisitsyn1 [~lisitsin@mxs.kg.ru] has quit [Quit: Leaving.]15:04
shogun-buildbotbuild #10 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/10  blamelist: Bj?rn Esser <bjoern.esser@gmail.com>15:13
besser82sonne|work:  now ctest reports the fail, as should be  ;)15:23
sonne|workbesser82: hurray finally some failures ;)15:37
besser82sonne|work:  yes, and some lot false positive as well  ;)15:38
besser82sonne|work:  I need to improve that IF(RESULT) part in checkgdb.cmake  :(15:38
sonne|workbesser82: what is wrong with it?15:45
besser82sonne|work:  It always is true, even if there is just warning or some non-fail msg in RESULT15:46
sonne|workhmmhh - I guess that is because the old integration tests don't work15:47
besser82sonne|work:  So tests which pass, but produces output of any kind currently report as fail15:47
sonne|workbesser82: better focus on the new ones15:47
sonne|workbesser82: ohh15:47
sonne|workthan ignore me15:47
besser82sonne|work:  but I already have solution  :D15:47
besser82sonne|work:  I think...  let's test15:47
shogun-notifier-shogun: Bj?rn Esser :develop * 63d9882 / cmake/checkgdb.cmake: https://github.com/shogun-toolbox/shogun/commit/63d988201e06a3b6d3f6e6944f1fc18d9a0cf0f115:48
shogun-notifier-shogun: only report really failing tests as failed15:48
besser82sonne|work:  now i'm testing whether there is output matching ".*THE FOLLOWING TESTS FAILED.*"15:49
besser82sonne|work:  and only trigger the failure procedure after there's a match15:50
sonne|workbesser82: what?15:50
sonne|workbesser82: that doesn't seem correct15:51
sonne|workbesser82: we run checkgdb for each test right?15:51
sonne|workbesser82: so if the test fails it should return RESULT that triggers15:51
sonne|workotherwise nothing15:52
besser82sonne|work:  some tests report stuff like "multiclass/MulticlassStrategy.cpp line 45: MulticlassStrategy::CMulticlassStrategy(): register parameters!" to RESULT15:52
shogun-buildbotbuild #2592 of deb1 - libshogun is complete: Failure [failed test]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/2592  blamelist: Bj?rn Esser <bjoern.esser@gmail.com>15:53
besser82sonne|work:  which isn't a fail but just some garbage15:53
sonne|workbesser82: I mean this works OK under python215:53
besser82sonne|work: nope15:53
sonne|workbesser82: http://buildbot.shogun-toolbox.org/builders/deb3%20-%20modular_interfaces/builds/2129/steps/test%20python%20modular/logs/stdio15:53
besser82sonne|work: the warning is always there  ;)15:53
sonne|workit looks all ok there ^^15:54
besser82sonne|work:  same test report that on py2, too  ;)  "multiclass/MulticlassStrategy.cpp line 45: MulticlassStrategy::CMulticlassStrategy(): register parameters!"15:55
sonne|workbesser82: yeah but that is not an error15:55
besser82sonne|work:  i know, but the previous RESULT-logic made it one  ;)15:55
sonne|workno15:55
sonne|workthey all passed15:55
besser82sonne|work:  sine it only tested "IF(RESULT)"15:56
sonne|workbut now it doesn't work at all15:57
besser82sonne|work:  which was actually an effect of that b0rken RESULT logic  ;)15:57
besser82sonne|work:  because now I trigger a FATAL_ERROR in cmake to get /usr/bin/cmake to exit(1) if there was sth.  that made IF(RESULT) == true15:58
besser82sonne|work: which would be the case on `garbage` output15:58
besser82sonne|work: so i enhaced that logic with a REGEX MATCH15:59
besser82sonne|work: like above15:59
besser82shogun-buildbot: force build --branch=develop 'deb4 - python3'16:00
shogun-buildbotbuild #11 forced16:00
shogun-buildbotI'll give a shout when the build finishes16:00
shogun-buildbotbuild #11 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/1116:04
@wikinghehe so now we are the stage where we were in august16:05
besser82wiking:  :-P16:07
@wikingbesser82: integration tests always failed with this error when using python3: EXCEPTION 'str' does not support the buffer interface16:09
@wikingand the problem is that i've checked into the issue16:09
@wikingand it's really a shitty one to fix16:09
besser82wiking:  for me it looks simple to fix  ;)16:10
besser82wiking:  Lemme do a random guess StringIO is used, rye?16:10
@wikingbesser82: ok16:10
@wikingbesser82: noup it's not that unfortunately16:11
@wikingbesser82: have u had a look in integration script?16:11
besser82wiking:  the it is sth not yet :(16:11
besser82wiking:  but the fix can't be that hatd  ;)16:11
@wikingbesser82: unfortunately it is16:11
shogun-notifier-shogun: Bj?rn Esser :develop * b2989bb / cmake/checkgdb.cmake: https://github.com/shogun-toolbox/shogun/commit/b2989bb2bb32f41381355a63b439ab1905a3f4c216:18
shogun-notifier-shogun: only report a failed test if there is some certain output16:18
shogun-buildbotbuild #2593 of deb1 - libshogun is complete: Success [build successful]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb1%20-%20libshogun/builds/259316:22
besser82wiking:  If assume the exception is raised caused by `b = pickle.load(open(fname))`  ?!?16:25
@wikingyeps16:25
besser82wiking:  haha, easy shiats  :D16:25
besser82wiking:  one needs to do BufferIO and use decode  ;)16:26
besser82wiking:  will try to fix  ;)16:27
shogun-buildbotbuild #12 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/12  blamelist: Bj?rn Esser <bjoern.esser@gmail.com>16:27
shogun-notifier-shogun: Bj?rn Esser :develop * d57df18 / cmake/checkgdb.cmake: https://github.com/shogun-toolbox/shogun/commit/d57df18935d3c759e6578be10f609ccf942daaf416:32
shogun-notifier-shogun: fixup regex-matching16:32
besser82shogun-buildbot: force build --branch=develop 'deb4 - python3'16:33
shogun-buildbotbuild #13 forced16:33
shogun-buildbotI'll give a shout when the build finishes16:33
shogun-buildbotbuild #13 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/1316:37
shogun-buildbotbuild #14 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/14  blamelist: Bj?rn Esser <bjoern.esser@gmail.com>16:43
shogun-notifier-shogun: Bj?rn Esser :develop * 8f59231 / cmake/checkgdb.cmake: https://github.com/shogun-toolbox/shogun/commit/8f59231931ee6ffd9daba11a8724d9b09e27a0f316:44
shogun-notifier-shogun: be more verbose on failure16:44
besser82shogun-buildbot: force build --branch=develop 'deb4 - python3'16:44
shogun-buildbotbuild #15 forced16:44
shogun-buildbotI'll give a shout when the build finishes16:44
shogun-buildbotbuild #15 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/1516:48
shogun-buildbotbuild #16 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/16  blamelist: Bj?rn Esser <bjoern.esser@gmail.com>16:54
shogun-notifier-shogun: Bj?rn Esser :develop * bf1af37 / tests/integration/python_modular/tester.py: https://github.com/shogun-toolbox/shogun/commit/bf1af3796cf9e580d0eba36cad43821b03b7fb8917:04
shogun-notifier-shogun: fix loading pickles with Python317:04
-!- lambday [67157c4d@gateway/web/freenode/ip.103.21.124.77] has joined #shogun17:04
besser82shogun-buildbot: force build --branch=develop 'deb4 - python3'17:05
shogun-buildbotbuild #17 forced17:05
shogun-buildbotI'll give a shout when the build finishes17:05
shogun-buildbotbuild #17 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/1717:09
shogun-buildbotbuild #18 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/18  blamelist: Bj?rn Esser <bjoern.esser@gmail.com>17:14
shogun-notifier-shogun: Bj?rn Esser :develop * 5f8a6c0 / tests/integration/python_modular/tester.py: https://github.com/shogun-toolbox/shogun/commit/5f8a6c02ca8fb6044e04fb071b97b2f811105e6917:14
shogun-notifier-shogun: fix indention17:14
besser82shogun-buildbot: force build --branch=develop 'deb4 - python3'17:14
shogun-buildbotbuild #19 forced17:14
shogun-buildbotI'll give a shout when the build finishes17:14
shogun-buildbotbuild #19 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/1917:18
besser82sonne|work, sonney2k, wiking: that small fix actually fixed 101 intergration tests with py3  ;)17:20
shogun-buildbotbuild #20 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/20  blamelist: Bj?rn Esser <bjoern.esser@gmail.com>17:24
-!- Saurabh7 [~Saurabh7@117.200.191.191] has quit [Ping timeout: 265 seconds]17:27
shogun-notifier-shogun: Bj?rn Esser :develop * 2518fbe / tests/integration/python_modular/tester.py: https://github.com/shogun-toolbox/shogun/commit/2518fbe9846e31e601e782d864db571e33b0dc8217:44
shogun-notifier-shogun: cover the numpy-incompats between Py2 and Py317:44
shogun-notifier-shogun: Bj?rn Esser :develop * 9eff7f3 / cmake/checkgdb.cmake: https://github.com/shogun-toolbox/shogun/commit/9eff7f380bc7c54f21a90c9318fc0d954ad3527817:44
shogun-notifier-shogun: get rid of the overlaborated and useless output-processing17:44
besser82shogun-buildbot: force build --branch=develop 'deb4 - python3'17:44
shogun-buildbotbuild #21 forced17:44
shogun-buildbotI'll give a shout when the build finishes17:44
-!- lisitsyn [~lisitsyn@80.252.20.67] has joined #shogun17:48
shogun-buildbotbuild #21 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/2117:48
besser82sonney2k, sonne|work, wiking: looks like we need to create separate pickles for Py3-stuff  :(17:53
besser82sonney2k, sonne|work, wiking: binary-shit simply isn't compatible17:53
shogun-buildbotbuild #22 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/22  blamelist: Bj?rn Esser <bjoern.esser@gmail.com>17:53
besser82sonney2k, sonne|work, wiking: like da data-structures from modshogun17:54
besser82sonney2k, sonne|work, wiking: one can unpickle, but don't get useful data ---> SEGFAULT17:54
besser82sonney2k, sonne|work, wiking: as that SEGFAULT definetively comes from modshogun reconstructor...18:12
besser82sonney2k, sonne|work, wiking: that bin-incompat is known and stuff like numpy-arrays are affected as well...18:14
besser82sonney2k, sonne|work, wiking: but that acutally lives in the nature of pickles  :(18:14
-!- lambday [67157c4d@gateway/web/freenode/ip.103.21.124.77] has quit [Ping timeout: 272 seconds]18:40
@wikingbesser82: i thought it's an easy fix18:51
besser82wiking:  loading / upickling that shiats, yes18:51
besser82wiking:  patching python2 / python3 to keep bin-compat no  :-P18:52
@wikingyeah, i knew18:52
besser82wiking:  at least for more than the half of the integrationtests it works that way  ;)18:53
besser82wiking:  so the easy fix "repairs" 101/190 failing intergr test18:54
-!- iglesiasg [~iglesiasg@524AE0A7.cm-4-3d.dynamic.ziggo.nl] has joined #shogun19:50
-!- mode/#shogun [+o iglesiasg] by ChanServ19:51
@iglesiasgsup people19:52
besser82iglesiasg:  Hey!20:13
-!- shogun-notifier- [~irker@7nn.de] has quit [Quit: transmission timeout]20:44
besser82lisitsyn:  Wassup?!?  Any news?  How is it going?22:27
lisitsynbesser82: preparing for my exam ;)22:28
lisitsynbesser82: what about u?22:28
besser82lisitsyn:  good luck!22:28
lisitsynthanks22:28
besser82lisitsyn:  Heavily involved in getting Python3-support for da shogunzzzz22:28
lisitsynbesser82: cool, I see quite a lot of troubles here?22:45
besser82lisitsyn:  the most annoying shiatz is about da binary-incompat of pickles with numpy && modshogun structures22:46
@sonney2kbesser82, hey there23:03
@sonney2kbesser82, is the ascii aka level 0 pickle protocol compatible?23:03
-!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Quit: buildmaster reconfigured: bot disconnecting]23:16
-!- shogun-buildbot [~shogun-bu@7nn.de] has joined #shogun23:16
@sonney2kshogun-buildbot: force build --branch=develop 'deb4 - python3'23:16
shogun-buildbotbuild #23 forced23:16
shogun-buildbotI'll give a shout when the build finishes23:16
-!- lisitsyn [~lisitsyn@80.252.20.67] has quit [Quit: Leaving.]23:20
shogun-buildbotbuild #23 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/2323:21
-!- shogun-buildbot [~shogun-bu@7nn.de] has quit [Quit: buildmaster reconfigured: bot disconnecting]23:24
-!- shogun-buildbot [~shogun-bu@7nn.de] has joined #shogun23:25
-!- iglesiasg [~iglesiasg@524AE0A7.cm-4-3d.dynamic.ziggo.nl] has quit [Quit: Leaving]23:27
@sonney2kshogun-buildbot: force build --branch=develop 'deb4 - python3'23:43
shogun-buildbotbuild #24 forced23:43
shogun-buildbotI'll give a shout when the build finishes23:43
shogun-buildbotbuild #24 of deb4 - python3 is complete: Failure [failed test python modular]  Build details are at http://buildbot.shogun-toolbox.org/builders/deb4%20-%20python3/builds/2423:47
--- Log closed Fri Jan 10 00:00:45 2014
« Wednesday, 2014-01-08 Index Friday, 2014-01-10 »

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