Log for #openttdcoop.devzone on 6th June 2010:
Times are UTC Toggle Colours
00:02:05  *** KenjiE20 has quit IRC
07:35:34  *** ODM has joined #openttdcoop.devzone
07:56:25  *** desrik has joined #openttdcoop.devzone
07:57:02  *** frosch123 has joined #openttdcoop.devzone
08:19:43  <planetmaker> Yexo: concerning regression tests: how would that best be done / implemented?
08:20:19  <Yexo> a lot of small nml files and expected nfo/grf output
08:20:27  <planetmaker> in order to properly understand cargo types etc I need to play around that anyway, so I could do that as a regression test before I tackle properly rail types
08:20:31  <Yexo> then run nml on each of those test file and compare the output with the expected output
08:20:35  <planetmaker> ok, I guess I can do that :-)
08:20:53  <planetmaker> would it be make files or implemented in python?
08:21:19  <planetmaker> first test IMHO would be testing action8 though ;-)
08:21:36  <Yexo> an nml file without action8 would work just fine, so only one test for that is needed
08:21:52  <planetmaker> would it produce a valid grf?
08:22:09  <Yexo> it would produce a grf without an action8
08:22:29  <Yexo> but for the regression test that doesn't matter, as long as the other actions are ok
08:22:42  <planetmaker> but it's hard to test that grf then ;-)
08:23:02  <planetmaker> whether that regression test actually is a good test
08:23:03  <Yexo> you don't need to test the grf ,just compare the output to known-good output
08:23:15  <planetmaker> hm, right
08:23:46  <Yexo> while testing you add the action8, then when you're sure the resulting grf works ok you remove the grf-block and put it in the regression directory
08:24:10  <planetmaker> though I feel a bit uncomfy without ingame test possibility :-) I guess that ^
08:24:47  <planetmaker> I guess I might for a starter add small but full grfs... and cut them when it works
08:25:13  <Yexo> the action8 is only a single line anyway, so adding it is ok
08:26:38  <planetmaker> hm, single line? For me it's like 5 - or wouldn't it go with name and description?
08:27:15  <Yexo> ehm, single sprite
08:27:26  <planetmaker> ah, yeah
08:30:42  <planetmaker> checking via md5sum maybe?
08:30:56  <Yexo> something like this?
08:31:02  <planetmaker> and should I prepare a small Makefile which runs the tests?
08:31:51  <planetmaker> a separate folder for each test? Yeah, why not
08:31:59  <planetmaker> sounds like the best option
08:32:05  <Yexo> it doesn't have to be a makefile, a small python script that runs each test is also fine
08:32:10  <Yexo> whatever is easiest
08:32:15  <planetmaker> though I called the dir regression instead of test ;-)
08:32:33  <Yexo> yes, regression as dir name makes more sense
08:32:38  <planetmaker> Well... I'm more proficient with a makefile I guess than writing a python script ;-)
08:33:08  <planetmaker> though the latter might make more sense
08:35:11  <Yexo> a makefile is fine really
08:40:02  <frosch123> noai regression uses diff -u to create a difference you can directly pipe into patch again to update the "known-good"
08:40:26  <planetmaker> hm?
08:40:43  <planetmaker> updating the "known-good" when exactly?
08:41:08  <frosch123> when it was not as good as it should be :p
08:41:13  <Yexo> when the formatting for nfo-output changes for example
08:42:14  <frosch123> or when you append new stuff
08:42:32  <planetmaker> hm, yes. I haven't thought about how to update regressions properly so far
08:42:52  <Yexo> noai uses one very big test, I'd rather have 20 small test files for nml then 1 big test
08:43:06  <planetmaker> yeah, I agree.
08:43:24  <frosch123> so if you have multiple files, maybe put the output in one directory, and known-good in one directory, and run diff -Nur on them, which you can then also use to add/remove new/old stuff
08:43:25  <planetmaker> That's what I aim at now
08:43:51  <Yexo> that is a very good idea
08:43:55  <planetmaker> frosch123: I thought to make it as Yexo just proposed: one dir per test
08:44:38  <planetmaker> But possibly giving the results and expectations in one dir might be an idea... though this output can somewhat easily summarized by a proper regression script
08:44:46  <frosch123> do it as you like, i was just very impressed by how noai does it :)
08:44:51  <planetmaker> :-)
08:45:37  <planetmaker> hm... though the dir thing can be done for each test: expected and output as sub dirs there...
08:46:02  <planetmaker> but the diff of a grf will be... interesting ;-)
08:46:34  <frosch123> grf? oh i thought you would compare nfo
08:46:42  <planetmaker> grf is what matters :-)
08:46:51  <planetmaker> nfo is just... an additional test
08:46:52  <Yexo> but not really human-readable
08:46:56  <planetmaker> My opinion :-)
08:46:58  <frosch123> oh, maybe use grfcodec do decode again?
08:47:03  <planetmaker> I'll test against both
08:47:23  <planetmaker> frosch123: not as broken as it is :-P
08:47:30  <frosch123> :p
08:47:42  <planetmaker> e.g. it will fail on my machine as it doesn't produce valid grfs
08:47:45  <planetmaker> why ever
08:47:45  <Yexo> imo it'd be best to compare both nfo and grf, if there is a diff you can read in the nfo diff what's the problem and fix it
08:47:53  <planetmaker> *valid pcx
08:47:54  <frosch123> but true, if you compare both, and the nfo is fine, while the grf differs, the difference is not very important
08:48:00  <Yexo> if only the nfo or the grf differs then there is a bigger problem
08:48:18  <planetmaker> yep :-)
08:48:28  <planetmaker> though it might be a formating change in the nfo output only ;-)
08:48:32  <planetmaker> like an additonal space or so
08:48:53  <Yexo> that should be easy to spot in a diff
08:48:58  <planetmaker> grf difference is as such more crucial
08:49:07  <planetmaker> yes, easy to see, true :-)
08:49:48  <Yexo> so test both: if someone works on the output code he should expect differences in nfo formatting for example and apply those to the expected output
08:50:25  <planetmaker> yep :-)
09:16:20  <Ammler> how you know, something is known-good?
09:16:32  <Ammler> morning :-)
09:17:23  <planetmaker> Ammler: previously tested :-)
09:17:35  <planetmaker> a working grf is per definition known-good
09:17:45  <planetmaker> and good morning to you, too
09:18:42  <Ammler> so if you add a feature, you need to test the test grf ingame?
09:20:06  <Ammler> which would then be easier with one big grf, instead a lot smaller grfs
09:20:44  <Ammler> (didn't read the whole backlog, so maybe you discussed that already :-)
09:21:13  <Yexo> depends on how hard it is
09:21:38  <Yexo> for small features reading the nfo checking if grfcodec gives the same output could be enough
09:21:58  <Ammler> you like to use grfcodec as reference?
09:22:16  <Yexo> to compare nfo->grf conversion? yes
09:22:40  <Ammler> I hoped, we can replace grfcodec with nml :-)
09:23:00  <Yexo> it can, I meant just while writing the regression tests
09:23:09  <Yexo> which is something a "user" of nml would never do
09:29:38  <Ammler> Yexo: the problem is that you still don't know, if grfcodec is broken
09:30:20  <Yexo> Ammler: that's why it's only used while writing the regression tests, if the nml and grfcodec output differs you have to find out which of the 2 is wrong
09:30:23  <Ammler> so you might use a fixed grfcodec version?
09:30:54  <Ammler> ah, got it, you will save the binaries in the vcs too?
09:31:01  <Yexo> huh?
09:31:11  <Ammler> the final grfs for the comparision
09:31:25  <Yexo> nml users won't even notice whether or not I use grfcodec
09:31:28  <Ammler> maybe I don't :-)
09:31:33  <Yexo> "nml user" is "grf developer with nml"
09:32:05  <Ammler> that was "clear"
09:32:26  <Yexo> nml can write nfo and grf, if you encode the nfo file with grfcodec you should get the exact same grf as nml wrote
09:32:56  <Yexo> if that is not the case, then either nml or grfcodec contains a bug
09:33:34  <Yexo> it's then up to one of the nml developers to find what program is wrong and if it's nml, fix it
09:34:12  <Yexo> the complete "encode with grfcodec" could be replaced with a "manually encode" step but that's very tedious
09:35:58  <Ammler> well, you could in first step test pure encoding by updating same nfo with different versions, that would be better
09:36:08  <Ammler> than testing with different apps, imo
09:36:43  <Yexo> yes, but that doesn't work for new features, as there is no "older version" to compare against
09:37:20  <Ammler> so grfcodec is only in use for the nml developer, not part of the test itself, right?
09:37:20  <Yexo> once a test is considered "correct" then the regression test will do exactly that, compare the output of the current version against a "known-good" output
09:37:26  <Yexo> exactly
09:37:42  <Yexo> it's only used to make sure the "known-good" output really is good
09:37:53  <Ammler> so that means, if we run the test on the server with nightly for example, we don't need grfcodec
09:37:58  *** KenjiE20 has joined #openttdcoop.devzone
09:38:03  <Yexo> correct again :)
09:38:16  <Ammler> then my worry is gone :-)
09:38:21  <Yexo> ok :)
09:40:03  <Brot6> NFO Meta Language - Revision 239: Feature: Add capability for regression tests (planetmaker) @
09:41:04  <Yexo> planetmaker: should the regression test use a relative path to make sure to use the correct nml2nfo?
09:41:30  <planetmaker> hm, you mean like .. in this case? Might be a good idea
09:41:50  <Yexo> and maybe add || echo "Regression test failed" after the && echo "ok"
09:41:58  <planetmaker> make fails
09:42:02  <planetmaker> if the diff is not empty
09:42:16  <Ammler> if not, we could check if the installation worked :-)
09:42:17  <planetmaker> so I considered that not necessary
09:42:26  <Yexo> ah, ok
09:42:28  <Rubidium> for regression testing you shouldn't miss the "fail" cases, i.e. you should check the std* output as well
09:43:36  <planetmaker> <-- that's how a fail looks to me
09:43:44  <planetmaker> should it continue on fail?
09:44:44  <Yexo> stop on fail is ok (or use make -k flag to continue if you really want that)
09:45:04  <Ammler> if it doesn't continue, you might miss another error?
09:45:18  <Ammler> -k, ok
09:45:20  <planetmaker> quite possible, though it's one diff for everything
09:45:34  <Yexo> hmm, for creating a single diff it'd be nice if it continues
09:45:38  <planetmaker> so it should probably show everythig different
09:45:52  <Ammler> how does output look, if ok?
09:45:56  <Ammler> empty?
09:46:02  <planetmaker> no
09:46:25  <planetmaker> ingo: ~/ottd/grfdev/nml/regression> make
09:46:26  <planetmaker> testing 001_action8.nml
09:46:28  <planetmaker> Differences found:
09:46:29  <planetmaker> Everything ok.
09:46:49  <planetmaker> better to silence it in case of success?
09:47:00  <planetmaker> posix compliance? :-)
09:47:05  <Ammler> well, like you did with md5test
09:47:11  <Ammler> I guess, there is an output too
09:47:16  <Yexo> planetmaker: the syntax in the language file is also wrong, everything after the first colon is part of the string, so align the colons, don't put spaces/tabs directly after it
09:47:32  <planetmaker> :-D
09:48:13  <Ammler> then I will add a %check macro to the build script
09:49:14  <planetmaker> I'm fine either way.
09:50:23  <Ammler> well, if no output, I could count the lines and alert like I do now with newgrfs
09:54:41  <planetmaker> also skipping the lines "testing XXX" ?
09:55:28  <Yexo> that would lead to problems, as then you wouldn't be able to see which test causes a syntax error for example
09:56:08  <Yexo> maybe use stdout for the messages and stderr for the diff or the other way around?
09:56:12  <planetmaker> one does on a fail in the diff
09:56:35  <planetmaker> Differences found:
09:56:37  <planetmaker> Binary files output/001_action8.grf and expected/001_action8.grf differ
09:56:53  <planetmaker> which is produced necessarily by 001_action8.nml
09:57:36  <Yexo> ok, so that is no problem then
09:57:44  <planetmaker> also it runs a full diff
09:57:51  <planetmaker> so everything is always displayed
09:57:59  <planetmaker> (of what differs)
09:59:10  <planetmaker> ok, so agreement is: no output on successful pass?
09:59:28  <Yexo> fine with me
09:59:36  <Ammler> or different output channels
10:00:02  <Ammler> but it would be nice to have a empty file, if everything is ok
10:00:07  <planetmaker> can do that
10:02:34  <planetmaker> did that
10:02:38  <Ammler> it might be good to have output so people see, the test was running...
10:02:46  <Brot6> NFO Meta Language - Revision 240: Change: Silence regression output on successful pass (planetmaker) @
10:03:05  <planetmaker> regression testing is only done by developers of NML
10:03:13  <planetmaker> we know it should be silent ;-)
10:03:20  <Yexo> and by the server nightly build (at least if not too much work)
10:03:28  <Ammler> people=you :-P
10:03:44  <Ammler> yes, it is only adding the make command to the build script
10:03:45  <planetmaker> Yexo: that's IMHO the point of adding it
10:03:57  <planetmaker> then we don't need to worry but will get told ;-)
10:04:24  <Yexo> even if it wouldn't be on the server it'd be a nice way to test code-changes
10:04:30  <planetmaker> yep :-)
10:04:33  <Ammler> you can add it to the commit hook alberth made
10:04:36  <Ammler> :-P
10:04:43  <planetmaker> he
10:05:01  <planetmaker> Ammler: it needs a simple call of 'make' in order to run the tests
10:05:11  <planetmaker> should be quite straight forward with the CF
10:05:21  <Yexo> make -C regression/, at least from the root directory
10:05:29  <planetmaker> ah, right
10:05:50  <Ammler> ok, which files do need to be copied?
10:06:00  <Ammler> or is it everything in stdout?
10:06:02  <Yexo> the output of make
10:06:05  <planetmaker> ^
10:06:09  <Ammler> ok :-)
10:06:26  <Yexo> part of the output might be in stderr, at least the message from make "compile failed" is
10:06:57  <Yexo> make: *** [all] Error 2 <- that line
10:07:15  <Yexo> so stdout is enough
10:07:22  <Ammler> ottdc@salieri:~/hg-repos/nml> make -C regression/
10:07:24  <Ammler> make: Entering directory `/home/ottdc/hg-repos/nml/regression'
10:07:25  <Ammler> make: *** [all] Error 1
10:07:53  <Yexo> hmm, use "cd regression; make" then
10:08:02  <Yexo> only it shouldn't give an error
10:08:14  <Ammler> still get Error
10:08:58  <Ammler> (with 240:7c4679c5b06d)
10:09:05  <planetmaker> what error?
10:09:13  <Ammler> no output, just that
10:09:27  <Ammler> might it be possible, you need to build it first?
10:09:54  <Yexo> that shouldn't be needed, it's all pure python code
10:10:03  <planetmaker> make -C regression/ works for me
10:10:11  <planetmaker> as does cd regression; make
10:10:19  <Yexo> both work for me too
10:10:19  <Ammler> I mean, building nml2nfo
10:10:43  <planetmaker> it doesn't need building really, does it? But it needs to be present.
10:11:02  <planetmaker> it expects to find nml2nfo
10:11:07  <Yexo> Ammler: and if you run: _V="" make -C regression/
10:11:44  <Ammler> for i in 001_action8; do ../nml2nfo --nfo output/$i.nfo --grf output/$i.grf $i.nml 2>/dev/null; done
10:11:46  <Ammler> make: *** [all] Error 1
10:11:51  <Ammler> ah
10:11:56  <Ammler> there it is :-)
10:12:09  <planetmaker> hm?
10:12:27  <Ammler> ply isn't installed
10:12:40  <Ammler> I will test it with osc then
10:12:41  <planetmaker> he
10:12:50  <planetmaker> osc?
10:12:56  <Ammler> the build chroot
10:13:00  <planetmaker> ah
10:20:35  <Ammler> hmm, moving and change in same commit is bad idea
10:20:59  <Ammler> I run a nightly test...
10:21:20  <Brot6> NFO Meta Language - Revision 241: DevZone: Add regression test to build script (Ammler) @
10:21:43  <Brot6> nml: update from r234 to r241 done (2 errors) -
10:22:03  <Ammler>
10:22:30  <Ammler> hmm, might be better for the log if I do "cd regression && make"
10:22:47  <Yexo> yep
10:25:41  <Brot6> NFO Meta Language - Revision 242: Fix (r241): change to regression directory before running make ... (Ammler) @
10:27:31  <Brot6> nml: update from r241 to r242 done -
10:29:37  <Ammler> hmm :-(
10:30:38  <Ammler> it returns a bad exit...
10:32:12  <Ammler> aha
10:32:21  <Ammler> I don't have ply installed for building
10:32:31  <Ammler> :-)
10:32:34  <planetmaker> :-P déja vu?
10:32:38  <Ammler> well
10:33:03  <Ammler> it isn't needed for building
10:33:28  <planetmaker> yes, it will
10:33:37  <planetmaker> if action1 etc is tested
10:33:43  <Yexo> not for building, only for testing
10:33:48  <Ammler> I hope you guys don't shoot me, if I add another FixFix :-P
10:34:03  <Yexo> not at all
10:34:36  <Ammler> also imaging might be needed?
10:35:10  <planetmaker> the biggest thing is the same of others calling one "Mr fix (rfix (rfix))" ;-)
10:35:20  <planetmaker> s/same/shame/ :-P
10:35:36  <Yexo> Ammler: all packages needed for nml projects are also needed to test nml itself
10:35:44  <Yexo> so also Imaging
10:36:01  <planetmaker> it will be once I add actionA test (working on it) ;-)
10:38:24  <planetmaker> hm... one png for all tests, or one for each?
10:38:40  <Brot6> nml: update from r242 to r243 done -
10:38:51  <Ammler> \o/
10:39:06  <Ammler> hmm
10:39:24  <Ammler> shall we mark the build as failed, if regression fails?
10:39:24  <Brot6> NFO Meta Language - Revision 243: Fix (again r241): all packages needed for nml projects are also... (Ammler) @
10:39:54  <planetmaker> It makes sense
10:39:58  <Ammler> Yexo: I stole your sentence :-P
10:39:59  <planetmaker> It needs attention
10:47:15  *** frosch123 has quit IRC
10:52:54  <Ammler> <-- if we preappend R; to the rpm, it would fail
10:53:09  <Ammler> then we don't need to count the errors :-)
10:53:30  <Yexo> sounds like a good idea
10:53:38  <Yexo> have fun
10:53:43  <Ammler> :-D
10:53:44  * Yexo is off for the rest of the day
10:53:50  <planetmaker> enjoy!
10:54:49  <Ammler> hmm, actually I didn't test building result, if it fails
10:55:15  <Ammler> the failed test was because of missing ply
10:59:23  *** Seberoth has joined #openttdcoop.devzone
11:02:15  <Ammler> planetmaker: to test failing, I simply change expected nfo a bit?
11:22:13  *** PeterT has quit IRC
11:22:31  *** PeterT has joined #openttdcoop.devzone
11:23:29  <Brot6> FIRS Industry Replacement Set - Revision 888: Change: work in progress on Aluminium Plant (andythenorth) @
11:23:29  <Brot6> OpenTTD-GUI - Feature #996 (New): support selecting the blitter in the options (planetmaker) @
11:24:01  *** PeterT has quit IRC
11:24:41  *** PeterT has joined #openttdcoop.devzone
11:25:48  <Rubidium> good luck with determining what blitters to show to the user :)
11:26:35  <planetmaker> yeah :-P
11:26:50  <planetmaker> I recall your statements last time we briefly discussed it ;-)
11:31:38  <Brot6> FIRS Industry Replacement Set - Feature #997 (New): Textile Mill needs some greeble / fenced tiles (andythenorth) @
13:08:14  *** frosch123 has joined #openttdcoop.devzone
13:55:49  <Brot6> NFO Meta Language - Revision 235:ea593d2e8247: Codechange: disable yacc debug output (yexo) @
13:55:49  <Brot6> NFO Meta Language - Revision 236:5733ff967c46: Codechange (#991): remove the cyclic dependency be... (yexo) @
13:55:49  <Brot6> NFO Meta Language - Revision 237:6b7b01f69eeb: Codechange (#991): remove the cyclic dependency be... (yexo) @
13:55:50  <Brot6> NFO Meta Language - Revision 238:0636eef6a4bc: Fix (#981): don't crash when loading an image file... (yexo) @
13:55:54  <Brot6> NFO Meta Language - Revision 239:eb7644e1bca0: Feature: Add capability for regression tests (planetmaker) @
13:55:58  <Brot6> NFO Meta Language - Revision 240:7c4679c5b06d: Change: Silence regression output on successful pass (planetmaker) @
13:56:02  <Brot6> NFO Meta Language - Revision 241:85040939ba86: DevZone: Add regression test to build script (Ammler) @
13:56:06  <Brot6> NFO Meta Language - Revision 242:49e8a6564767: Fix (r241): change to regression directory before ... (Ammler) @
13:56:10  <Brot6> NFO Meta Language - Revision 243:997db99e5a4f: Fix (again r241): all packages needed for nml proj... (Ammler) @
13:56:14  <Brot6> FIRS Industry Replacement Set - Revision 888:7d67f03ad502: Change: work in progress on Aluminium ... (andythenorth) @
13:56:18  <Ammler> sorry for spam :-P
14:14:26  <PeterT> lol
14:15:01  <Ammler> redmine does now show both, rev number and hash
14:15:18  <Ammler> but activity is broken :-(
14:16:23  <Ammler> strange
14:55:39  <PeterT> Ammler: huh, how many networks are you on?
14:55:58  <PeterT> I've seen you on Freenode, OFTC, EFnet, and Quakenet
14:56:12  <frosch123> PeterT: huh, how many networks are you on?
14:56:16  <Ammler> omg
14:56:27  <PeterT> frosch123: 6
15:01:39  <Ammler> he PeterT is running #tycoon on the old place :-)
15:04:56  <PeterT> :D
15:10:10  <Ammler> PeterT: btw. znc has ctcp flood protection
15:16:50  <desrik> any way to stop autopilot from broadcasting new games to irc?
15:17:07  <desrik> or edit the message?
15:17:44  <Ammler> what you mean exactly?
15:18:10  <Ammler> and yes, you can edit either the cfg or custom commands/callbacks
15:19:15  <desrik> when I start my server with auto pilot, I get this ....
15:21:08  <Ammler> check the callbacks
15:21:27  <Ammler> but it might be hardcoded
15:33:27  <desrik> must be hardcoded cause there isnt any callbacks for server startup or new game
15:36:04  <Ammler> join irc
15:43:27  <desrik> na that sends ap and openttd version number
16:18:50  <Brot6> firs: update from r886 to r895 done (2 errors) -
16:20:23  <Brot6> swedishrails: update from r24 to r25 done (6 errors) -
16:20:36  <Brot6> test: compile of r39 failed -
16:20:37  <Brot6> Following repos didn't need a nightlies update: 2cctrainset (r545), 32bpp-extra (r36), airportsplus (r50), bros (r12), comic-houses (r70), fish (ERROR r374), heqs (r320), newgrf_makefile (ERROR r99), nml (r243), nmts (r16), nutracks (r69), ogfxplus (r23), opengfx (r459), openmsx (r57), opensfx (r94), snowlinemod (r12), worldairlinersset (r643)
16:40:24  <desrik> ok i cant find anything in any of the tcl files or config of ap+ that would send this message....
16:41:33  <desrik> nvm i found it
16:44:21  <desrik> if anyone was wondering its in the lang file
16:56:28  <PeterT> Ammler: so does XChat
17:08:41  <PeterT> desrik: is it ok now?
17:09:50  <desrik> yeah figured out how to change them.... now i just need to figure out how to grab the company id of the sending user so i can set up a company reset command
17:10:28  <Ammler> please post such a command here or on the bugtracker
17:10:34  <PeterT> what Ammler said
17:10:37  <PeterT> that would be great :-P
17:11:27  <desrik> i know for a fact that ap CAN grab the company id
17:11:41  <desrik> the proof is in the companies command
17:11:47  <PeterT> I bet it could too
17:12:00  <PeterT> using grep and the console command
17:12:18  <desrik> i just have to combine that and the [who] command
17:12:34  <Ammler> desrik: you mean the players command
17:12:45  <PeterT> Ammler: players == companies
17:13:01  <Ammler> ap player is openttd client :-)
17:13:04  <desrik> yeah....but you would have to use who to determine who sent the message
17:27:23  <desrik> i can use tcl in commands right?
17:27:32  <desrik> like straight tcl?
17:27:52  <desrik> lol nvm
17:27:56  <desrik> dumb question
17:28:24  <PeterT> yes, you can
17:31:10  *** DJNekkid has quit IRC
17:35:02  *** DJNekkid has joined #openttdcoop.devzone
17:57:04  <Hirundo> Brot stopped announcing ?
18:03:29  <planetmaker> brot crashed the server :S
18:03:44  <Ammler> really?
18:03:51  <Ammler> again, hmm
18:03:55  <planetmaker> or nearly. Or what was it yesterday?
18:04:06  <planetmaker> not now :-)
18:04:19  <planetmaker> no fear
18:04:37  <Ammler> puh :-)
18:04:54  <planetmaker> :-)
18:05:32  <Ammler> Hirundo: what did Brot6 miss?
18:05:55  <Hirundo> I made a few small NML commits
18:07:44  <Hirundo> hmm.. they don't show in redmine either
18:08:37  <Ammler> yes, I realized
18:08:49  <Ammler> fetch changesets hook doesn't work anymore
18:09:04  <Ammler> since my update to the patch
18:16:18  <Ammler> fetch changesets fails :-(
18:21:41  <Brot6> NFO Meta Language - Revision 235: Codechange: disable yacc debug output (yexo) @
18:21:41  <Brot6> NFO Meta Language - Revision 236: Codechange (#991): remove the cyclic dependency between nml and... (yexo) @
18:21:41  <Brot6> NFO Meta Language - Revision 237: Codechange (#991): remove the cyclic dependency between action0... (yexo) @
18:21:42  <Brot6> NFO Meta Language - Revision 238: Fix (#981): don't crash when loading an image file without a pa... (yexo) @
18:21:46  <Brot6> NFO Meta Language - Revision 239: Feature: Add capability for regression tests (planetmaker) @
18:21:51  <Brot6> NFO Meta Language - Revision 240: Change: Silence regression output on successful pass (planetmaker) @
18:21:54  <Brot6> NFO Meta Language - Revision 241: DevZone: Add regression test to build script (Ammler) @
18:21:57  <Brot6> NFO Meta Language - Revision 242: Fix (r241): change to regression directory before running make ... (Ammler) @
18:22:01  <Brot6> NFO Meta Language - Revision 243: Fix (again r241): all packages needed for nml projects are also... (Ammler) @
18:22:05  <Brot6> FIRS Industry Replacement Set - Revision 888: Change: work in progress on Aluminium Plant (andythenorth) @
18:22:39  <Brot6> NFO Meta Language - Revision 244: Fix: Always round properties with a floating point value to an ... (Hirundo) @
18:22:39  <Brot6> NFO Meta Language - Revision 245: Fix: Add newlines where it improves nfo readability. (Hirundo) @
18:22:39  <Brot6> NFO Meta Language - Revision 246: Fix: Print default graphics for graphics / livery_override bloc... (Hirundo) @
18:22:53  <Ammler> rollback to vanilla redmine
18:31:15  <planetmaker> hm
18:36:38  <Hirundo> working through the ast and all actions to add string wrappers != fun
18:37:53  * planetmaker hugs Hirundo
18:38:13  <planetmaker> you want a special NML cake on the 19th? ;-)
18:40:08  <planetmaker> Hirundo: concerning regression tests: should each regression ship with its own small png file (if it needs one) or should all use a slightly larger one and re-use stuff there?
18:40:19  <planetmaker> I need a good name proposal in the latter case
18:40:48  <Hirundo> 19th ? = r20k party ?
18:41:05  <planetmaker> And I wonder... it probably needs documenting what is tested in which regression test...
18:41:12  <planetmaker> yes, 19th June = r20k party
18:41:36  <planetmaker> I understood that you'd come, right?
18:42:07  <Hirundo> .. no :( lack of time, money, and upcoming exams, I'm afraid
18:42:31  <planetmaker> oh :`-(
18:42:57  <planetmaker> no shared car trip with folks?
18:43:02  <planetmaker> just for a day?
18:43:36  <planetmaker> Like 4lberth, Rubi, TB, Y3xo... might fit 5 :-)
18:43:57  * planetmaker desperately tries to avoid highlight ;-)
18:44:45  <Hirundo> Ah, I wondered were all that 1337 sp34k came from
18:44:46  <planetmaker> there were even some thoughts among them that they return the very same day
18:45:08  <Brot6> FIRS Industry Replacement Set - Revision 889: Change: progress on Aluminium Plant (andythenorth) @
18:45:08  <Brot6> FIRS Industry Replacement Set - Revision 890: Change: remove unused town index check for Power St... (andythenorth) @
18:45:08  <Brot6> FIRS Industry Replacement Set - Revision 891: Add: template for remapping industry colors (andythenorth) @
18:45:09  <Brot6> FIRS Industry Replacement Set - Revision 892: Change: Aluminium Plant does look better without ex... (andythenorth) @
18:45:13  <Brot6> FIRS Industry Replacement Set - Revision 893: Feature: improved Aluminium Plant graphic (andythenorth) @
18:45:19  <Brot6> FIRS Industry Replacement Set - Revision 894: Change: tweak Aluminium Plant graphic (andythenorth) @
18:45:22  <Brot6> FIRS Industry Replacement Set - Revision 895: Change: tweak Aluminium Plant graphic (andythenorth) @
18:47:29  <Hirundo> sorry pm
18:47:42  <planetmaker> *sob*
18:47:56  * Hirundo hands towel
18:47:59  <planetmaker> :-)
18:48:12  <planetmaker> hm... the towel day was only recently, right?
18:51:26  <Rubidium> yep
18:51:50  <Rubidium> ~2 weeks
18:53:14  <planetmaker> 25th or so IIRC
19:21:32  *** Seberoth has quit IRC
19:21:37  *** Seberoth has joined #openttdcoop.devzone
19:33:24  <desrik> is there any other information on ap+ other than the openttdcoop wiki?
19:34:06  <Ammler> yes, devzone
19:34:12  <planetmaker> :-P
19:35:04  <desrik> ok. anyone know if there is a way to have commands not echoed back to the game or irc?
19:36:00  <Ammler> if you don't add reply, it doesn't echo
19:38:17  <desrik> ok thx
19:51:21  <desrik> how do I use "quotes" in a say message?
20:02:14  *** frosch123 has quit IRC
22:40:59  *** PeterT is now known as PeterT-ghost
22:41:21  *** PeterT-ghost is now known as PeterT
22:52:13  *** ODM has quit IRC

Powered by YARRSTE version: svn-trunk