Log for #openttdcoop.devzone on 11th May 2014:
Times are UTC Toggle Colours
03:27:41  *** stevenh has joined #openttdcoop.devzone
03:27:51  <stevenh> hello all, anyone awake? :)
03:28:26  <stevenh> If anyone reads this at any point, could they please allow my account access to the ssh repo for openttdcoop?
03:37:02  <stevenh> I have added my key to the files area on jpset
03:37:07  *** stevenh has quit IRC
05:08:12  *** LSky` has joined #openttdcoop.devzone
05:58:30  *** Alberth has joined #openttdcoop.devzone
06:30:12  <Alberth> yesterday I cooked   different editor syntax file generation for nml
06:30:17  <Alberth> how does it look?
06:58:20  <planetmaker> Alberth, the shebang should explicitly state python3 :)
06:58:41  <planetmaker> or python2, I guess
06:59:03  <Alberth> yeah, python2, nml is still python2
06:59:28  <Alberth> I tried 3 yesterday too, but it failed :)
07:00:44  <planetmaker> and I tried this patch on top of my py3 head and it failed :P
07:01:32  <Alberth> use 2to3 first :p
07:02:42  <planetmaker> update to default head is another option :P
07:05:06  <Alberth> that would be too easy :)
07:07:42  <DevZone> Project opengfx-mars build #103-push: SUCCESS in 2 min 30 sec:
07:07:50  <DevZone> Yippee, build fixed!
07:07:51  <DevZone> Project Townnames - German build #69-push: FIXED in 7.8 sec:
07:07:53  <DevZone> Yippee, build fixed!
07:07:53  <DevZone> Project NML - NewGRF Meta Language build #323-push: FIXED in 3 min 32 sec:
07:11:27  <Alberth> euhm, was it broken?
07:12:31  <planetmaker> 'unstable'. Due to a test job failing for stupid reasons unrelated to nml
07:12:36  <planetmaker>
07:13:34  <planetmaker> hm, I would have thought that numbers once used to be highlighted, too with my previous syntax file
07:13:47  <planetmaker> but maybe I mis-remember
07:14:42  <Alberth> didn't nml point out the position of a failed input?
07:15:24  <planetmaker> you mean like syntax error in file X, line Y? Yes
07:17:10  <Alberth> I meant printing the line, and then below  a   ^ at the position of the error
07:17:59  <Alberth> but I can only find a column search function, and I get that column number
07:18:11  <Alberth> so maybe my memory is wrong :)
07:19:55  <planetmaker> I don't think it ever quoted the line it considers faulty
07:20:43  <planetmaker> temmplate tmpl_toolbar_wide(x, y) { [x, y, 80, 40, 0, 0, NOCROP] }
07:20:54  <planetmaker> nmlc ERROR: "src/sprite_templates.pnml", line 8: Syntax error, unexpected token "tmpl_toolbar_wide"
07:20:54  <planetmaker> Included from: "ogfx-biggui.pnml", line 2
07:21:01  <planetmaker> ^ interesting error message :)
07:21:39  <planetmaker> why doesn't it complain about 'temmplate' instead?
07:22:18  <Alberth> it's a name of something, like a variable or a constant or so
07:22:44  <Alberth> it has no idea you consider it a typo in the template keyword :)
07:22:56  <planetmaker> :)
07:23:20  <Alberth> you can probably instantiate a template at that point in the specification
07:23:32  <Alberth> ie   name ( ... )
07:27:01  <planetmaker> the patch should set the execute permission to gen_editor
07:27:33  <planetmaker> (or did that simply get lost on my patch -p1 < file ?
07:27:44  <Alberth> I suspect the latter
07:28:23  <DevZone> Project opengfx-mars build #104-push: SUCCESS in 2 min 11 sec:
07:28:31  <DevZone> Project Townnames - German build #70-push: SUCCESS in 8.2 sec:
07:28:32  <DevZone> Project NML - NewGRF Meta Language build #324-push: SUCCESS in 3 min 7 sec:
07:29:03  <Alberth> new file mode 100755   <-- in the patch file, so it should work
08:13:13  * Alberth ponders adding macro expansion functionality to nml
08:57:46  <DevZone> Project opengfx-mars build #105-push: SUCCESS in 2 min 12 sec:
08:57:53  <DevZone> Project Townnames - German build #71-push: SUCCESS in 7.7 sec:
08:57:54  <DevZone> Project NML - NewGRF Meta Language build #325-push: SUCCESS in 3 min 6 sec:
11:10:48  *** KenjiE20 has quit IRC
11:15:11  *** KenjiE20 has joined #openttdcoop.devzone
11:17:16  *** gelignite has joined #openttdcoop.devzone
11:31:05  *** KenjiE20 has quit IRC
11:31:10  *** Supercheese has quit IRC
11:31:46  *** Supercheese has joined #openttdcoop.devzone
11:32:51  *** KenjiE20 has joined #openttdcoop.devzone
12:07:19  <DevZone> Project World Airliner Set build #124-push: FAILURE in 26 sec:
12:12:21  <DevZone> Yippee, build fixed!
12:12:22  <DevZone> Project World Airliner Set build #125-push: FIXED in 1 min 50 sec:
12:21:16  *** gelignite has quit IRC
12:27:12  *** yorick has joined #openttdcoop.devzone
15:27:57  <Alberth> action0 block allocation rewrite
15:46:10  <planetmaker> :-O big series
15:57:06  <Alberth> lots of small steps :)
15:59:14  <planetmaker> just reading through them
16:00:41  <planetmaker> might fix the houseID issue from andrew350?
16:00:53  <planetmaker> just along the way, eh?
16:01:21  <Alberth> i would guess not, but not sure
16:02:00  <Alberth> the "oh we have the block already -> ok" looks fishy in that respect
16:15:05  <planetmaker> it reads nice to me
16:15:29  <Alberth> and I guess nobody ever got to the upper limit, or they would had a crash
16:15:52  <Alberth> readable was one of the goals when making so many patches :)
16:16:20  <Alberth> I already had the implementation at that point ;)
16:21:48  <planetmaker> 010: sound effects has a max of -1 meaning infinite?
16:22:04  <DevZone> Project Japanese Buildings build #197-nightlies: SUCCESS in 20 sec:
16:23:41  <planetmaker> the use of 0xFFFF, 255, 8, 63, ... seems a bit inconsistent. Though I understand well where it comes from - it's how I'd naturally name the numbers, too
16:28:20  <Alberth> -1 means "none available", free space starts at first, upto and including last
16:29:47  <DevZone> Project ISR Industrial Station Renewal build #160-push: SUCCESS in 45 sec:
16:30:20  <Alberth> the last values looked natural to me as well
16:31:13  <Alberth> I considered using None for it, but it seemed overkill
16:31:35  <planetmaker> I'm pretty sure you tested compilation with a few NewGRFs. So, yes, please, for the series
16:31:53  <Alberth> main goal was to kill the "idx = idx + 1"  line in the search function
16:32:40  <Alberth> I didn't test that actually, I have the (perhaps wrong?) idea that the regression test is sufficient
16:33:06  <Alberth> but the code isn't really changed, just moved around a bit
16:34:08  <Alberth> and added a real class instead of this dict and a double array with magic numbers :)
16:34:13  <DevZone> Project Finnish Rail Infrastructure - Rails build #311-nightlies: SUCCESS in 8 min 30 sec:
16:37:26  <Alberth> pushed
16:37:44  <planetmaker> well, not sure we have a regression test which tests more than trivial ID allocations
16:37:53  <planetmaker> but we will learn now :D
16:38:29  <Alberth> probably not, it doesn't test any error either, I think
16:38:40  <Alberth> and we have a LOT of those :p
16:39:33  <planetmaker> dunno, mercurial has some unit tests which also tests for output or so. They also right now have a GSOC project which is about improving those
16:39:41  <planetmaker> maybe we could even learn from that?
16:39:58  <planetmaker> I haven't looked yet much at any details, just seen a few patches on hg-devel
16:40:30  <DevZone> Project opengfx-mars build #106-push: SUCCESS in 2 min 28 sec:
16:40:39  <DevZone> Project Townnames - German build #72-push: SUCCESS in 9.6 sec:
16:40:40  <DevZone> Project NML - NewGRF Meta Language build #326-push: SUCCESS in 3 min 30 sec:
16:40:59  *** oskari89 has joined #openttdcoop.devzone
16:41:01  *** ODM has joined #openttdcoop.devzone
16:41:24  <planetmaker> and they seem to have a rule which means that many fixes come with a new test :)
16:41:24  <DevZone> Project road-hog build #192-nightlies: SUCCESS in 40 sec:
16:41:32  <Alberth> unit tests explodes even more in number, as you're testing a few routines together, for example all methods in BlockAllocation
16:42:04  <Alberth> as well as all cases with a single routine
16:42:23  <Alberth> useful if you really depend on correctness of that code
16:42:55  <Alberth> but in general it's a lot of effort, much more than just writing the code itself
16:43:05  <planetmaker> yes. Our requirements are different and ^^
16:43:29  <Alberth> where the return on investment is quite unclear.
16:43:42  <Alberth> basically every test that never fails is wasted time
16:44:03  <planetmaker> yes. And we don't havea  requirement that the resulting grfs are always the same. Just valid
16:44:45  <Alberth> in my experience, if you have the logic right, you either make systematic errors that show up in almost any use, or it never fails
16:45:25  <Alberth> there are of course exceptions to this rule, eg the townname set allocation algorithm
16:45:50  <Alberth> but then again it's also not quite clear if you would have found that with a unit test
16:48:32  <DevZone> Project Iron Horse build #874-nightlies: SUCCESS in 1 min 37 sec:
16:49:07  * Alberth hopes jenkins is not test-builfing for every nml commit :p
16:49:43  <planetmaker> nope, every push
16:49:59  <DevZone> Project Dutch Trainset build #158-push: SUCCESS in 1 min 26 sec:
16:50:08  <planetmaker> the newest commit from every push and - if present - the newest tag therein.
16:50:28  <Alberth> oh, that's quite alright then
16:51:12  <planetmaker>
16:51:16  <DevZone> Project xussrset - Trains from Russia build #280-push: SUCCESS in 3 min 47 sec:
16:51:23  <planetmaker> ^ that's the hook which decides on the revs which to build
16:51:29  <DevZone> Project 2ccts build #110-push: SUCCESS in 1 min 30 sec:
16:51:42  <DevZone> Project ecs build #11-push: SUCCESS in 25 sec:
16:57:43  <Alberth> quite short, even
17:00:20  *** KenjiE20 has quit IRC
17:01:11  <DevZone> Project ISR Industrial Station Renewal build #161-push: SUCCESS in 46 sec:
17:17:34  *** frosch123 has joined #openttdcoop.devzone
17:40:22  *** oskari892 has joined #openttdcoop.devzone
17:47:49  *** oskari89 has quit IRC
17:51:48  *** KenjiE20 has joined #openttdcoop.devzone
18:40:28  *** oskari89 has joined #openttdcoop.devzone
18:47:15  *** oskari892 has quit IRC
19:40:27  *** oskari892 has joined #openttdcoop.devzone
19:47:20  *** oskari89 has quit IRC
20:25:54  <DevZone> Project ecs build #12-push: SUCCESS in 25 sec:
20:32:38  *** ODM has quit IRC
20:40:34  *** oskari89 has joined #openttdcoop.devzone
20:47:19  *** oskari892 has quit IRC
20:53:44  *** frosch123 has quit IRC
21:24:42  *** LSky` has quit IRC
21:24:53  *** oskari89 has quit IRC
21:31:00  *** Alberth has left #openttdcoop.devzone
23:51:26  *** yorick has quit IRC

Powered by YARRSTE version: svn-trunk