Log for #openttdcoop.devzone on 12th April 2014:
Times are UTC Toggle Colours
00:55:40  *** frosch123 has quit IRC
02:02:18  *** tycoondemon2 has joined #openttdcoop.devzone
02:07:29  *** tycoondemon has quit IRC
06:48:22  *** andythenorth has joined #openttdcoop.devzone
07:00:52  *** Alberth has joined #openttdcoop.devzone
07:49:07  *** andythenorth has quit IRC
08:02:58  *** andythenorth has joined #openttdcoop.devzone
08:21:17  *** andythenorth has quit IRC
08:22:51  <Alberth> the opengfx-mars thingie seems to be very stuck, ideas on how far we are, and/or how to make it move again?
08:39:41  <planetmaker> not so much in particular on either. I need some motivation to actually start working on it tbh
08:40:13  <planetmaker> it doesn't seem to me like a interesting newgrf code, just like grinding more newgrf code :(
08:41:01  <Alberth> nobody is stopping you from adding nifty newgrf features :)
08:41:36  <planetmaker> :)
08:41:52  <Alberth> perhaps we should ask zeph?
08:42:45  <Alberth> it would be great if it could be published, imho
08:43:00  <planetmaker> I planned to start looking at stations at some point. But didn't yet get around to
08:43:22  <planetmaker> Stations on grounds it could teach me something about that part of the specs I don't yet know much
08:43:42  <Alberth> sounds useful to know about
08:43:54  <planetmaker> yeah. Want to know about also for possible nml extension
08:44:09  <Alberth> my thoughts exactly :)
08:44:44  <planetmaker> with isr and chips there's good open-source reference, so... yeah.
08:45:08  <planetmaker> reference to get started beyond specs- reading only, that is
08:46:40  <planetmaker> However, I agree with you. It will be useful to prep-up mars project for a release
08:47:20  <planetmaker> I've the impression, that many large chunks are done (you did an overview recently of which graphics were not yet coded), but that there's still lots of (smaller) holes missing
08:47:41  <planetmaker> which will take a good amount of time to fill. Probably longer than the time invested so far. Pesky details
08:48:19  <Alberth> it would perhaps be a good project for a beginning coder
08:48:28  <Alberth> but we don't many of those, I think
08:51:44  <planetmaker> yes, I hoped actually that *someone* would also step up and offer help
08:52:25  <Alberth> most people seem to have forgotten about it, except for the extended mars project
08:52:37  <Alberth> but that's about 32bpp mostly, iirc
08:52:39  <planetmaker> which is strange really
08:52:58  <planetmaker> Well, tbh, I wonder that Zeph doesn't provide 32bpp and 4x sprites for Mars
08:53:14  <planetmaker> zBase shows he can. And it would be nice (IMHO)
08:54:17  <Alberth> zBase is not really nice, it looks very unfinished to me
08:54:43  <planetmaker> I do agree there
08:54:58  <planetmaker> the ground tiles need rework, all of them
08:55:09  <Alberth> the only thing going for it is that it is complete, and consistent
08:55:11  <planetmaker> the textures are not nice enough
08:55:32  <planetmaker> the houses actually start to look much nicer when on a 'decent' ground
08:56:02  <planetmaker> and worst of zBase is the unfinished GUI
08:56:05  <Alberth> I can imagine :)
08:56:06  <planetmaker> buttons missing etc
08:56:20  <planetmaker> and that it defaults to 2x GUI
08:56:40  <Alberth> yep, it shouldn't do that
08:58:40  <planetmaker> I spend days on just preparing shore and river tiles for pota-ghat. Such work can hardly be automated and it would break somewhat the quick and automatically generated sprites workflow for zBase
08:58:56  <planetmaker> though probably one could go for extensive modelling of shores alternatively in blender, too
08:59:23  <planetmaker> but the time difference between zeph's shores and mine is surely a factor of 10 if not 100 (provided you know the tools used)
08:59:48  <planetmaker> and similar additional work can be spend on (some) textures
08:59:56  <planetmaker> it would get better. Time spend would get much more
09:00:42  <Alberth> maybe pikkabird gets in this direction
09:01:20  <planetmaker> rumors say so, yes
09:01:23  <planetmaker> would be nice
09:01:40  <planetmaker> the vehicles in pineapple definitely look nicely-textured
09:01:58  <planetmaker> but he ignores the length contraction, so it looks funky at some angles :)
09:03:22  <Alberth> it's obviously a bug in openttd rather than the graphics :D
09:07:29  <planetmaker> to some extend, yes
09:10:10  <Alberth> s/d/t/
09:12:43  <planetmaker> yup :) Common mistake I make
09:13:15  <rubidium> such fun ;)
09:13:56  <rubidium> though... somehow one can state that it isn't a bug in OpenTTD, but rather a bug in the "specification" of OpenTTD
09:15:11  <planetmaker> yeah, probably the better place to localize, but doesn't change it :)
09:21:12  <rubidium> since the specification is more or less in concrete, we can't change that and as such we can't fix the "bug" in OpenTTD. If we fix the bug in OpenTTD, it won't be a real TTD descendant
09:21:27  <rubidium> at least one needs to trash a lot of the NewGRF stuff
09:22:03  <Alberth> openttd handles it better than an old simutrans version. A horizontal wagon sprite there slowly changed position while crossing a tile. It did that for each tile, as a result all gaps continually changed length :)
09:24:25  <planetmaker> I just closed a feature request from 2007 as implemented ;)
09:24:43  <Alberth> tbh I care very little about TTD descending. Trashing much newgrf stuff would give a nice new clean sheet, but I doubt we will ever recover from such a change
09:24:53  <planetmaker> anyone knows the issue# for the length bug?
09:25:26  <planetmaker> such a change would split the community totally, yes
09:25:26  <Alberth> it didn't get closed immediately? :o
09:26:04  <planetmaker> Alberth, I think it was never implemented with knowledge of that issue: show canal water differently at sea level
09:26:19  <planetmaker> people who use the ttd base set will still not see a difference. But they could use newgrfs
09:26:50  <planetmaker> or use opengfx which has different water for sea and canal/river :)
09:28:11  <Alberth> blue vs red comes to mind :p
09:28:52  <Alberth> although green rivers are nicer in our family friendly environment :)
09:30:17  <planetmaker> speaking of mars river colours?
09:34:45  <Alberth> not in particular
09:41:05  <planetmaker> ah, well :) Murky-brown for Mars :P
09:47:54  *** yorick has joined #openttdcoop.devzone
10:16:02  <DevZone> Project opengfx-mars build #70-test: SUCCESS in 2 min 19 sec:
10:16:10  <DevZone> Project Townnames - German build #47-test: SUCCESS in 8 sec:
10:16:11  <DevZone> Project NML - NewGRF Meta Language build #295-push: SUCCESS in 3 min 18 sec:
10:38:38  *** mg_ has quit IRC
10:54:07  *** frosch123 has joined #openttdcoop.devzone
10:57:51  *** mg_ has joined #openttdcoop.devzone
11:11:14  *** andythenorth has joined #openttdcoop.devzone
11:31:35  *** Supercheese has quit IRC
11:32:11  *** Supercheese has joined #openttdcoop.devzone
12:18:26  *** Alberth has left #openttdcoop.devzone
12:22:13  <DevZone> Project opengfx-mars build #71-test: SUCCESS in 2 min 7 sec:
12:22:21  <DevZone> Project Townnames - German build #48-test: SUCCESS in 7.4 sec:
12:22:22  <DevZone> Project NML - NewGRF Meta Language build #296-push: SUCCESS in 2 min 57 sec:
12:32:28  <DevZone> Project Iron Horse build #803-push: SUCCESS in 1 min 24 sec:
13:13:24  <DevZone> Project road-hog build #136-push: SUCCESS in 35 sec:
13:20:26  <DevZone> Project road-hog build #137-push: SUCCESS in 37 sec:
13:45:25  *** Alberth has joined #openttdcoop.devzone
14:09:16  <planetmaker> Alberth, with respect to py2->py3 conversion. I think I changed my opinion wrt waiting for hg to change, too
14:09:39  <planetmaker> While there's some progress there, not sure it's anything which will happen fast by any means
14:10:10  <planetmaker> at least I understand what I read and heard as 'slow at best'
14:10:31  <Alberth> :)
14:10:52  <Alberth> roughly translated, "we'll move when 2.7 dies" ?   :D
14:11:25  <planetmaker> there's no such statement. But they really seem to have some problem to move some stuff over to py3
14:11:56  <planetmaker> I forgot which, but they need some things supported in py3 which are not yet, though they seem to be in the pipeline for py3
14:12:14  <Alberth> sounds weird, but perhaps in the extensions somewhere
14:12:35  <Alberth> I can imagine they have a lot of code, and no doubt some of it is "weird" in some sense :)
14:12:47  <planetmaker> also text / string / bytestream /etc
14:13:24  <Alberth> a usual source of problems
14:14:17  <Alberth> I can imagine they didn't always carefully consider the different cases, and they work a lot with files
14:16:30  <planetmaker>
14:16:31  <Webster> Title: SupportedPythonVersions - Mercurial (at
14:16:56  <planetmaker> but that's only very brief and not the lastest updates to it, but roughly.
14:18:27  <Alberth> sounds reasonable
14:20:58  <Alberth>  is the current set of changes to perform for py3 from current trunk
14:21:11  <Alberth> although there are a few false positives in there
14:26:46  <planetmaker> Actually, I do believe, if we want to go for py3 - and that's the way to go - then doing that now is a good time
14:27:37  <planetmaker> Not much going on except the occasional small maintenance. The existing version does work (for certain definitions of 'works'). And the last major release is not long past
14:35:59  <Alberth> what python2 version is minimally required?
14:37:33  <planetmaker> for nml?
14:37:57  <planetmaker> it once worked with 2.5. But I don't have a py 2.5 anymore to check, so I don't know
14:38:30  <planetmaker> I'll not be sad anymore if we say 2.6 or 2.7
14:40:53  <Alberth> iirc, they have better generator support
14:41:51  <planetmaker> than py3?
14:43:19  <Alberth> than 2.5 :)
14:43:39  <planetmaker> :)
14:45:17  <Alberth> bummer   subprocess.check_output was added in 2.7
14:46:41  <planetmaker> hm. Dropping 2.6 might not yet be the best idea. Not sure.
14:52:42  <Alberth> np, I'll just use the subprocess.Popen(cmd, stdout=subprocess.PIPE).communicate()[0]   version :)
14:53:24  <planetmaker> Bit unrelated, <-- notes from pycon 2014 today. Most important is the last point :P
14:53:25  <Webster> Title: Gmane Loom (at
14:54:50  <andythenorth> “Packaging sucks, but we're improving, and we're actually doing better than other dynamic languages."
14:55:02  <andythenorth> that’s the same stuck record as for last 5 or 7 years
14:56:43  <planetmaker> hm, yeah. I somewhen put 'revisit nml packaging' on my list of things to do this year
14:57:07  <Alberth> "we no longer recommend auto-generating Python 3 compatible source code using 2to3"  :)
14:57:27  <planetmaker> yup. I wonder whether it ever was a good idea :)
14:58:36  <DevZone> Project road-hog build #138-push: SUCCESS in 46 sec:
14:58:44  <DevZone> Project Iron Horse build #804-push: SUCCESS in 1 min 35 sec:
15:01:27  <Alberth> right, .format conversion is currently the big blob of changes
15:01:44  <Alberth> that seems added in 2.6
15:01:48  <DevZone> Project opengfx-mars build #72-test: SUCCESS in 2 min 11 sec:
15:01:55  <DevZone> Project Townnames - German build #49-test: SUCCESS in 7.7 sec:
15:01:56  <DevZone> Project NML - NewGRF Meta Language build #297-push: SUCCESS in 3 min 7 sec:
15:02:44  <planetmaker> also...
15:02:45  <Webster> Title: Gmane -- death to 2.7 long live 2.7 (at
15:02:50  <planetmaker> but you likely know that :)
15:03:55  <Alberth> I didn't know the end was this close :)
15:04:42  <Alberth> ok to drop 2.5 support?
15:04:53  <Alberth> no idea whether that is mentioned somewhere :p
15:05:31  <Alberth> oh, lots of places :)
15:06:30  <planetmaker> +1 to -2.5 support :)
15:10:06  <Alberth>
15:10:30  <planetmaker> aye :)
15:10:32  <Alberth> is this a Feature?  :)
15:10:40  <planetmaker> I updated wiki.
15:10:57  <planetmaker> he. If you can sell it as feature, do so :P
15:11:40  <Alberth> Feature: we will break your setup  :)
15:11:50  <DevZone> Project World Airliner Set build #64-push: FAILURE in 20 sec:
15:12:32  <Alberth> let's see what breaks :p
15:13:00  <planetmaker> we will see. I believe that the regression tests and the test builds by DevZone will catch already quite a bit
15:13:25  *** planetmaker has left #openttdcoop.devzone
15:13:31  *** planetmaker has joined #openttdcoop.devzone
15:13:33  <planetmaker> hm :)
15:15:20  <DevZone> Project opengfx-mars build #73-test: SUCCESS in 2 min 10 sec:
15:15:28  <DevZone> Project Townnames - German build #50-test: SUCCESS in 7.2 sec:
15:15:29  <DevZone> Project NML - NewGRF Meta Language build #298-push: SUCCESS in 3 min 1 sec:
15:18:43  *** ODM has joined #openttdcoop.devzone
15:19:35  <planetmaker> uh.... I just notice. The creation of the binary will need thorough rework, when switching to py3, I think
15:20:07  <planetmaker> at least it needs a wine python3 install somehow transferred to CF
15:20:10  <Alberth> That sounds likely
15:21:13  <Alberth> how do you feel about the "map(f, ...)" versus "[f(x) for x in ...]" variants?
15:31:54  <planetmaker> no strong feelings either way really
15:36:24  <planetmaker> I've the feeling the latter, like z = [x**2 for x in range(1,6)] can be more concise than a map approach. But I might err
15:37:13  <Alberth> I favor the latter, so will do a conversion
15:37:29  <Alberth> we seem to have much fewer than anticipated :)
15:38:13  <Alberth> white_pixels = len(filter(lambda p: p == 255, mask_sprite.getdata()))    <-- this one is great :)
15:38:59  <planetmaker> yup :)
16:21:52  <DevZone> Project Japanese Buildings build #168-nightlies: SUCCESS in 17 sec:
16:34:19  <DevZone> Project Finnish Rail Infrastructure - Rails build #282-nightlies: SUCCESS in 8 min 44 sec:
16:36:34  <DevZone> Project opengfx-mars build #74-test: SUCCESS in 2 min 14 sec:
16:36:41  <DevZone> Project Townnames - German build #51-test: SUCCESS in 7.3 sec:
16:36:42  <DevZone> Project NML - NewGRF Meta Language build #299-push: SUCCESS in 5 min 51 sec:
16:41:11  <DevZone> Project road-hog build #139-nightlies: SUCCESS in 36 sec:
16:48:12  <DevZone> Project Iron Horse build #805-nightlies: SUCCESS in 1 min 37 sec:
16:49:23  <DevZone> Project Dutch Trainset build #128-push: SUCCESS in 1 min 10 sec:
16:50:52  <DevZone> Project FISH is ships build #179-push: SUCCESS in 1 min 29 sec:
16:50:54  <DevZone> Project xussrset - Trains from Russia build #262-push: SUCCESS in 3 min 28 sec:
16:52:10  <DevZone> Project Iron Horse build #806-push: SUCCESS in 1 min 17 sec:
17:08:17  *** gelignite has joined #openttdcoop.devzone
17:13:42  *** oskari89 has joined #openttdcoop.devzone
17:25:25  *** yorick has quit IRC
18:19:06  *** andythenorth has quit IRC
18:21:58  *** andythenorth has joined #openttdcoop.devzone
18:51:18  <planetmaker> andythenorth, on the contrary, it gave me a small enough problem which sounds interesting. Totally pointless probably. But interesting ;)
18:51:30  <andythenorth> ok
19:43:38  *** andythenorth has quit IRC
19:52:12  *** andythenorth has joined #openttdcoop.devzone
20:18:13  *** ODM has quit IRC
20:31:19  *** andythenorth has quit IRC
20:50:59  <DevZone> Project NML - NewGRF Meta Language build #300-push: FAILURE in 6.9 sec:
20:56:57  <DevZone> Project opengfx-mars build #75-test: SUCCESS in 2 min 10 sec:
20:57:05  <DevZone> Project Townnames - German build #52-test: SUCCESS in 7.2 sec:
20:57:06  <DevZone> Yippee, build fixed!
20:57:07  <DevZone> Project NML - NewGRF Meta Language build #301-push: FIXED in 3 min 4 sec:
20:57:46  <Alberth> \o/
21:01:30  <planetmaker> \o/
21:11:05  <Alberth> ok, I have most changes copied into nml
21:11:45  <Alberth> the print debug statements should get some love, but after that, it's basically the 2to3 conversion that needs to be done
21:13:15  <DevZone> Project opengfx-mars build #76-test: SUCCESS in 2 min 8 sec:
21:13:19  <Alberth> and that should be pretty simple due to my preparations today :)
21:13:22  <DevZone> Project Townnames - German build #53-test: SUCCESS in 7.3 sec:
21:13:23  <DevZone> Project NML - NewGRF Meta Language build #302-push: SUCCESS in 2 min 57 sec:
21:22:13  *** Alberth has left #openttdcoop.devzone
21:40:22  *** oskari89 has quit IRC

Powered by YARRSTE version: svn-trunk