Config
Log for #openttdcoop.devzone on 7th February 2012:
Times are UTC Toggle Colours
00:46:08  <Brot6> NewGRF Meta Language - Feature Request #3664: Use more sensible default for loading speed (and ot... (oberhuemer) @ http://dev.openttdcoop.org/issues/3664#change-9514
03:21:32  *** dihedral has quit IRC
03:22:46  *** seberoth has quit IRC
03:37:18  *** seberoth has joined #openttdcoop.devzone
03:37:47  *** dihedral has joined #openttdcoop.devzone
06:43:03  *** JVassie has joined #openttdcoop.devzone
07:21:12  *** JVassie has quit IRC
10:38:25  <Ammler> Yexo: because I didn't found a ticket about, I had the idea, it is "solved", I will create another ticket then :-)
11:35:24  <Brot6> Dutch Trains 2.0 - Support #3632: list of vehicles to be drawn (Voyager1) @ http://dev.openttdcoop.org/issues/3632#change-9515
11:40:07  <Brot6> Dutch Trains 2.0 - Support #3632: list of vehicles to be drawn (Voyager1) @ http://dev.openttdcoop.org/issues/3632#change-9515
15:09:40  <Ammler> you guys have all python 2.7 to run nml on? Or do you not care about md5sum?
15:32:48  <Ammler> planetmaker: what python do you where you develop opengfx?
15:33:05  *** Jupix has quit IRC
15:33:09  <Ammler> and did you check md5sum last time?
15:33:18  *** Jupix has joined #openttdcoop.devzone
15:36:15  <planetmaker> I use iirc 2.6. And I simply uploaded the version from the DevZone
15:38:15  <Ammler> I take that as a "no" :-)
15:39:43  <Ammler> it seems to be something random, the md5sum is always different on the old distros
15:40:09  <Ammler> maybe it is realted to "/usr/lib64/python2.6/site-packages/ply/yacc.py:74: DeprecationWarning: the md5 module is deprecated; use hashlib instead"
15:40:25  <Ammler> so ply is the issue
15:43:45  <Ammler> no
15:46:33  <Ammler> marcel@inspiron:~/hg/opengfx> grfcodec -d ogfxe_extra.grf
15:46:34  <Ammler> Found grf container version 1
15:46:36  <Ammler> Segmentation fault
15:50:13  <Ammler> - 4578 sprites/ogfxe_extra.pcx 274 9352 01 31 34 -31 0
15:50:20  <Ammler> + 4578 sprites/ogfxe_extra.pcx 2 9352 01 31 34 -31 0
15:51:57  <Ammler> starts with sprite 3133
15:54:52  <Ammler> river shore set2
16:08:46  <Rubidium> what version of opengfx and grfcodec are that?
16:11:16  <Rubidium> as for me it decodes just fine
16:17:40  <Ammler> opengfx 0.4.1 and grfcodec tip
16:18:01  <Ammler> updated to 0.5.3 and it worked
16:37:24  <Brot6> NewGRF Meta Language - Bug #3666 (New): md5sum depends on python version (Ammler) @ http://dev.openttdcoop.org/issues/3666
16:38:25  <Ammler> maybe it's a ogfx bug
16:39:16  <Rubidium> there is no 0.5.3
16:39:45  <Ammler> oh, 5.1.3
16:40:17  <Rubidium> so I fixed a crash without knowing about it... nice! ;)
16:40:23  <Ammler> tip was 893:6a6461ca9515
16:40:35  <Ammler> no
16:40:41  <Ammler> the seg fault was with tip
16:40:46  <Ammler> 5.1.3 is working
16:41:03  <Ammler> you introduced a failure
16:41:34  <Rubidium> tip works fine for me
16:41:50  <Ammler> yes, and devzone too
16:41:50  <Rubidium> or rather, r893
16:43:30  <Rubidium> so... what is wrong?
16:43:52  <Ammler> no clue :-(
16:44:20  <Ammler> which commit should I go back?
16:44:49  <Rubidium> how the hell am I supposed to know that when you can't even make clear under which circumstances it crashes?
16:45:47  <Ammler> he, what is unclear?
16:46:07  <Ammler> 5.1.3 ok, tip crash
16:46:53  <Ammler> those fail compiles yesterday were all valid or devzone issue?
16:52:45  <Ammler> any clue, what those random images could mean?
16:52:48  <Ammler> planetmaker: ^
16:53:11  <Ammler> or any idea, how I could find those sprites in the source?
16:57:55  <Brot6> OpenGFX - Bug #3667 (New): no static md5sum on old distros (Ammler) @ http://dev.openttdcoop.org/issues/3667
16:58:27  <Rubidium> I got grfcodec tip, I got opengfx 0.4.1 zip: decode works fine
16:58:35  <Rubidium> valgrind doesn't even complain
16:58:43  <Rubidium> what is grfcodec tip for you?
16:58:56  <Ammler> you have 64bit?
16:58:59  <Rubidium> as it's kinda saying latest, but latest when?
16:59:02  <Rubidium> yes
16:59:39  <Ammler> [17:40] <Ammler> tip was 893:6a6461ca9515
17:01:01  <Rubidium> are you sure that grfcodec is ran?
17:01:02  <Ammler> how is that command again to find "bad" commit?
17:01:16  <Rubidium> bisect
17:01:28  <Ammler> you mean built?
17:01:54  <Ammler> well, I update and run make
17:02:16  <Ammler> once grfcodec -d -p2 ogf.. works, once not
17:02:22  <Rubidium> there might be many places where another grfcodec could be hiding
17:02:40  <Rubidium> (it happened more than once for me)
17:02:44  <Ammler> I also run grfcodec --version
17:03:13  <Ammler> I don't think, it does chose different grfcodec randomly, does it?
17:03:24  <Rubidium> no
17:03:49  <Rubidium> but it's not the first time someone claims to be running X, but is actually running Y
17:03:52  <Ammler> marcel@inspiron:~/hg/grfcodec/.devzone/build> which grfcodec
17:03:54  <Ammler> /home/marcel/bin/grfcodec
17:03:55  <Ammler> marcel@inspiron:~/hg/grfcodec/.devzone/build> l /home/marcel/bin/grfcodec
17:03:57  <Ammler> lrwxrwxrwx 1 marcel users 33 Jul 31  2010 /home/marcel/bin/grfcodec -> /home/marcel/hg/grfcodec/grfcodec*
17:04:39  <Ammler> cd
17:04:41  <Rubidium> what do you mean with "once ... works, once not"?
17:04:58  <Ammler> hg up 5.1.3; make -> works
17:05:06  <Ammler> hg up; make -> doesn't
17:06:19  <Rubidium> so hg bisect -b tip; hg bisect -g 5.1.3
17:06:28  <Rubidium> and *go*
17:06:34  <Rubidium> but you might hit other problems
17:06:46  <Rubidium> which have been solved subsequently
17:07:03  <Rubidium> could you at least get a backtrace of the crash?
17:07:19  <Ammler> you know, how well I do with debugging :-P
17:07:35  <Ammler> I might have no debug tools installed anymore
17:07:40  <Rubidium> to Makefile.local add CXXFLAGS = -g -ggdb -O0
17:07:49  <Rubidium> make clean && make
17:09:47  <Rubidium> gdb --ex run --args grfcodec <parameters>
17:09:52  <Rubidium> when it crashed:
17:09:53  <Rubidium> bt full
17:10:00  <Rubidium> pastebin output
17:11:19  <Ammler> http://paste.openttdcoop.org/show/1078/
17:11:28  <Brot6> grfcodec: update from r892 to r893 done - http://bundles.openttdcoop.org/grfcodec/nightlies/r893
17:16:00  <Rubidium> aha!
17:18:00  <Ammler> The first bad revision is:
17:18:01  <Ammler> changeset:   869:4fd22bd6f8f9
17:18:22  <Ammler> and it has the segfault
17:19:46  <Ammler> marcel@inspiron:~/hg/grfcodec> hg bisect -c "make; cd ~/hg/opengfx; grfcodec -d -p2 ogfxe_extra.grf"
17:20:32  <Ammler> was quite cool :-)
17:21:37  <Brot6> bandit: update from r223 to r225 done (1 warnings) - http://bundles.openttdcoop.org/bandit/nightlies/r225
17:22:54  <Rubidium> not quite sure that revision introduced the bug though
17:23:28  <Ammler> so it's not 869?
17:24:17  <Rubidium> 865 also exhibits sings
17:24:42  *** frosch123 has joined #openttdcoop.devzone
17:24:53  <Ammler> Changeset 868:9aae2b3afe86: good
17:25:42  <Ammler> http://paste.openttdcoop.org/show/1079/
17:26:51  <Ammler> hmm, 870 didn't segfault?
17:27:30  <Ammler> now that is very silly
17:27:41  <Ammler> ..
17:30:36  <Ammler> now I set good to 870 and it returns 884 as bad
17:31:29  <Rubidium> that's more plausible
18:08:21  <Brot6> Following repos rebuilds successful without any difference to earlier nightlies builds: bros (1 warnings), transrapidtrackset, 2cctrainset (78 warnings), cets (251 warnings), worldairlinersset, heqs, basecosts, water-features, isr (2 warnings) (Diffsize: 1367), 32bpp-extra (2 warnings), newgrf_makefile (Diffsize: 12), snowlinemod, metrotrackset (Diffsize: 1), fish, ttrs (7 warnings), ogfx-trees, smts (Diffsize: 8), chips (1 warnings),
18:08:21  <Brot6> comic-houses (3 warnings) (Diffsize: 22)
18:12:26  <Rubidium> Ammler: does it work now?
18:13:17  <Brot6> GRFCodec - Revision 894:03bf8bebbba0: -Fix: null pointer dereferences when decoding to pcx (Rubidium) @ http://dev.openttdcoop.org/projects/grfcodec/repository/revisions/03bf8bebbba0
18:13:17  <Brot6> GRFCodec - Revision 895:385c23d20541: -Fix: valgrind error/warning regarding reading out of bounds (Rubidium) @ http://dev.openttdcoop.org/projects/grfcodec/repository/revisions/385c23d20541
18:13:17  <Brot6> GRFCodec - Revision 896:387f273acb0f: -Fix: minor memory leak (Rubidium) @ http://dev.openttdcoop.org/projects/grfcodec/repository/revisions/387f273acb0f
18:17:50  <Ammler> Rubidium: yes, thanks :-)
18:18:26  <Ammler> but well, grfcodec isn't really needed anymore
18:18:33  <Ammler> I need working nml :-P
18:19:21  <Ammler> I just found the issue because I was looking for the ogfx issue
18:23:00  <Ammler> usually I do not like youtube videos but this was fun: http://www.youtube.com/watch?v=mcncoGh-etI
18:23:01  <Webster> Title: Chihuahua Plays with English Mastiff - YouTube (at www.youtube.com)
19:10:15  *** andythenorth has joined #openttdcoop.devzone
20:00:46  *** JVassie has joined #openttdcoop.devzone
21:17:10  *** andythenorth has quit IRC
22:28:35  *** JVassie has quit IRC
23:12:06  *** frosch123 has quit IRC

Powered by YARRSTE version: svn-trunk