Tuesday, 2013-09-03

*** tpb has joined #timvideos00:00
*** mithro- has joined #timvideos00:22
*** hyades has quit IRC02:30
Sewaryay! cross compiling finally succeeded!03:14
*** hyades has joined #timvideos03:22
mithroSewar, awesome!03:22
Sewarbut static linking libcurl didn't work, maybe try it again later03:25
mithrowhat does it fail with?03:25
mithrostatic linking under windows is much easier then dynamically linking under windows as dlls are horrible03:26
*** CarlFK has left #timvideos03:27
Sewarlet me try again03:27
mithroSewar, don't forget to record your build steps somewhere so people can reproduce them03:28
mithrohyades, ping?03:28
SewarMakefile should take care of that right?03:29
hyadesmithro: sorry was sleeping03:29
mithrohyades, no worries, don't execpt you to to up all the time03:29
mithroSewar, well - are you compiling libcurl ?03:29
Sewarmithro: yes03:30
mithrois that part of the makefile?03:30
Sewaryes, it will compile it for the target os as part of the build03:30
mithroSewar, okay03:31
mithrohyades, your wondering about the utf8 problem?03:31
hyadesthat should be solvable, i think putting encoding format03:32
mithrohyades, one of your files has something strange in it03:33
mithrohyades, try running coveralls locally and see if you can reproduce it03:34
tpbTitle: dpaste.de: Snippet #238842 (at dpaste.de)03:34
Sewarmithro: what am I missing? .exe still looks for .dll03:34
mithrothe following line is the problem "/home/sewar/curl-7.32.0/output/bin/curl-config --libs"03:35
mithroyou need03:36
mithro(assuming your compiling statically :)03:36
*** hyades has quit IRC03:37
Sewarnow it works :)03:37
mithroSewar, \o/03:37
mithroSewar, so I'm not sure how you'd do it, but do you have tests for your code?03:40
mithroSomething like c-unit might work03:40
mithrowine is then useful for running the tests :)03:40
Sewarmithro: noted03:42
*** hyades has joined #timvideos03:43
mithroSewar / hyades: I'm going to be away pretty much all of next week03:56
mithroand have limited access from tomorrow03:58
hyadesmithro, how do I detect the error in this file, its huge!03:58
hyades\n{u\'url\': u\'\', u\'message\': u\'Build processing error.\', u\'error\': True}\n03:59
mithrohyades, do a repr of coverage value in /usr/local/lib/python2.7/dist-packages/coveralls/report.py 1103:59
*** hyades has quit IRC04:07
mithroSewar, you still about?04:09
Sewarmithro: thanks for letting us know, I'm sure CarlFK and iiie will offer great help04:10
mithroSewar, I recommend you change the way you are working04:10
mithroSewar, Don't push directly to timvideos/edid.tv04:10
mithroSewar, instead push to your own fork04:10
mithroand when your build is green / happy send a pull request to timvideos/edid.tv04:11
mithrothen if travis still thinks it's green, merge the pull request04:11
Sewarmithro: Sure, that makes sense04:11
mithrothat way you won't keep spaming me with failure messages ;)04:11
mithroSewar, and as you can do things like rebasing, squash commits, etc04:12
mithroSewar, have you an iiie agreed on what will be considered a "successful completion of GSoC"04:14
Sewarwell, I think I'm pretty close to the main goals of the project, but we should discuss this in next meeting04:16
mithroSewar, btw it would be nice to have auto-deploy from repository to the website04:17
Sewaryeah, when I shift to triple travis builds, auto-deploy should be no problem04:18
mithroSewar, great!04:19
mithrotriple travis builds?04:19
Sewarfirst at my repo, second as pull request, third at timvideos repo, which is great actually04:20
mithrookay, I understand04:21
*** hyades has joined #timvideos05:27
hyadesmithro-: mithro ping06:06
*** mithro- has quit IRC06:08
*** mithro- has joined #timvideos06:08
hyadesmithro-: ping06:09
mithro-hyades, pongs06:09
mithro-I'm in a meeting now06:09
mithro-but will be around later06:09
hyadesk :)06:09
*** mithro- has quit IRC06:36
*** mithro- has joined #timvideos06:37
mithrohyades, about now07:38
*** mithro- has quit IRC07:40
hyadesmithro: yes, what about the repr stuff?08:10
mithroso the issue is in the coverage dictionary it is trying to turn into json08:14
hyadeseven if there is a strange character. how do I avoid it? All these files are generated at runtime08:18
mithrohyades, once we figure out what is causing it08:21
mithrohyades, we can figure out how to prevent it08:21
hyadesanother thing related to the build matrix08:22
hyadeseverytime a make a row to the matrix, it starts a separate build08:23
mithrohyades, that is correct08:23
mithrohyades, actually08:23
mithrothe matrix actually acts like a matrix08:24
mithrobut matrix under env is different08:24
tpb<http://ln-s.net/+obm> (at about.travis-ci.org)08:24
hyadescan i make it something like, reach a particular state, and then go into 4 states from that state08:25
mithrohyades, no, sadly not08:26
hyadescoz I dont want to repeat the build procedure since its very heavy08:26
mithrohyades, so I recommend doing the build on your own precise machine08:26
mithrohyades, tar it up and put it on google drive08:26
mithrohyades, then just download and extract it08:26
mithrohyades, but you don't want to do that for everything - just the parts that don't matter08:27
hyadesi dont have it on precise yet, will build it soon08:27
mithrohyades, btw I have given you the ability to manipulate the gst-switch repo08:30
mithroin timvideos08:30
tpbTitle: timvideos/gst-switch · GitHub (at github.com)08:30
hyadesokay, I will look into this a bit later :)08:30
hyadesmithro: https://github.com/eddyxu/cpp-coveralls/issues/608:31
tpb<http://ln-s.net/-tc7> (at github.com)08:31
mithrohyades, good find08:32
hyadeswell, all the c files have ASCII format08:35
mithrohow do you know they are all ASCII format?08:39
mithroyou mean the output of file?08:39
tpbTitle: Ubuntu Pastebin (at paste.ubuntu.com)08:41
tpbTitle: Ubuntu Pastebin (at paste.ubuntu.com)08:41
hyadesencoding of all files in my directory http://paste.ubuntu.com/6058151/08:44
tpbTitle: Ubuntu Pastebin (at paste.ubuntu.com)08:44
mithrohyades, file just looks at a subset of the file to determine the encoding09:06
mithroif a single file is 99% ASCII and has 1 or 2 characters which are non-ascii it is unlikely that file will discover them09:07
hyadesshould i patch it with encode(utf, ignore) ??09:08
mithrohyades, no09:08
mithrohyades, we should find the file which has the invalid bytecode09:08
hyadesi am unable to replicate this exception in my machine09:11
mithro233   E9   é09:11
mithrohyades, if you could output coverage value using repr we could figure out where it is coming from09:16
hyadesshould I do this on travis?09:16
hyadesthere is probably no such character in my local copy09:17
mithrohyades, possibly09:18
mithrohyades, the fact that gsocketinputstream.c.gcov is a UTF-8 Unicode Text rather then ASCII is concerning09:19
mithrobut I don't want to jump to conclusions yet09:20
hyadesokay yes, this file has some problem09:21
tpbTitle: [Python] >>> f = open('gsocketinputstream.c.gcov') >>> a = f.read() >>> a ' -: - Pastebin.com (at pastebin.com)09:23
mithroa.find(u'0xe9') is not going to work09:25
mithroyou want09:25
mithroa.find('\xe9') is not going to work09:25
tpbTitle: Ubuntu Pastebin (at paste.ubuntu.com)09:32
hyadesmaybe those copyright characters09:32
mithrohyades, as I said I don't want to jump to conclusions09:33
mithrohyades, so lets get that coverage dictionary printed first09:33
hyadeson my machine or travis?09:34
mithrohyades, you said you couldn't reproduce the error09:36
mithrohyades, so what use is it going to be on your machine?09:37
hyadesjust a repr(coverage) is good?09:42
mithrohyades, up to you09:51
mithroyou might want to output more information09:52
hyadesyou can have some coffee, it will take around half an hour :)09:52
hyadesbut I think that coverage of only selective files whose output files have been generated should be taken, instead of all.09:54
hyadesfor files like gsocketinputstream.c coverage is not required09:55
hyadesmithro: https://api.travis-ci.org/jobs/10928061/log.txt?deansi=true11:35
tpb<http://ln-s.net/-tf-> (at api.travis-ci.org)11:35
hyadesthe repr(coverage)11:35
mithroThe log length has exceeded the limit of 4 Megabytes (this usually means that test suite is raising the same exception over and over).11:37
mithrohyades, no e9 in the limited output we have11:40
mithrohyades, you'll probably have to output it to a file locally then use wget/curl to spit it somewhere11:40
mithrohome time for me11:59
hyadesmithro: even if the char is found what to do?11:59
*** hyades has quit IRC12:21
*** hyades has joined #timvideos12:54
hyadesmithro: http://paste.ubuntu.com/6058926/plain/15:23
tpbTitle: OpenID transaction in progress (at paste.ubuntu.com)15:23
hyadesmithro: indeed these characters are present15:34
*** hyades has quit IRC16:34
*** hyades has joined #timvideos18:13
*** hyades has quit IRC18:18
*** hyades has joined #timvideos18:51
hyadesmithro, I was trying to build in precise. I get that TypeError on my VM also18:54
hyadesdef function(*args, **kwargs):18:54
hyades>       return info.invoke(*args, **kwargs)18:54
hyadesE       TypeError: Argument 1 does not allow None as a value18:54
hyadesthings are different here :/18:55
hyadesquantal no issues18:57
*** hyades_ has joined #timvideos19:56
*** hyades has quit IRC19:59
*** hyades_ has quit IRC20:28
*** hyades has joined #timvideos20:32
*** CarlFK has joined #timvideos21:12
hyadesCarlFK: the precise one was different from quantal one. A patch solved it. :)21:35
CarlFKwhich one was erroring ?21:37
*** parx has joined #timvideos21:38
CarlFKand you were developing with quantal?21:40
CarlFKI figured it was something like that21:40
CarlFKcan't you run the tests under quantal?21:41
hyadesPrecise doesn't allow bus names to be None21:41
hyadessadly travis has precise systems21:41
hyadesanyways, I put some values in bus name, and it worked.21:42
*** iiie0 has joined #timvideos21:53
*** CarlFK has quit IRC22:56
*** iiie0 has quit IRC23:51

Generated by irclog2html.py 2.12.1 by Marius Gedminas - find it at mg.pov.lt!