Log for #openttdcoop.devzone on 31st October 2014:
Times are UTC Toggle Colours
07:10:51  *** LSky` has joined #openttdcoop.devzone
11:22:25  *** yorick has quit IRC
11:22:36  *** yorick has joined #openttdcoop.devzone
11:38:51  *** oskari89 has joined #openttdcoop.devzone
12:19:38  *** oskari892 has joined #openttdcoop.devzone
12:26:26  *** oskari89 has quit IRC
12:37:44  *** Supercheese has quit IRC
13:19:46  *** oskari89 has joined #openttdcoop.devzone
13:26:27  *** oskari892 has quit IRC
13:37:21  *** LSky` has quit IRC
13:37:36  *** LSky` has joined #openttdcoop.devzone
15:38:48  *** ODM has joined #openttdcoop.devzone
15:50:25  *** frosch123 has joined #openttdcoop.devzone
16:47:26  *** Alberth has joined #openttdcoop.devzone
16:51:53  <DevZone> Project World Airliner Set build #534-push: SUCCESS in 3 min 47 sec:
16:58:10  *** LSky` has quit IRC
16:58:16  *** LSky` has joined #openttdcoop.devzone
17:01:23  <frosch123> <- i wonder whether to apply some form of alignment to make it look nicer
17:01:36  <frosch123> not sure how or whether at all
17:06:31  <Alberth> I am not even sure these things mean anything to a regular nml user
17:06:32  <planetmaker> not sure whether that is needed
17:07:10  <Alberth> the whole point of nml is not to dig around in action X (0 <= X <= 0x14)
17:07:35  <frosch123> well, better than nml suddenly says: sorry, you cannot add more
17:08:08  <frosch123> they are limits, once you reach them, you are in trouble
17:08:54  <planetmaker> I do agree it makes sense to give these info to the power users. There are several who reached one of these and then wondered about whether it's a bug
17:08:56  <Alberth> fair enough
17:09:00  <planetmaker> we got a bug opened each time :)
17:09:17  <frosch123> well, i would like to see them on jenkins
17:09:18  <planetmaker> and it can indeed help during programming to keep them somewhat known
17:09:40  <planetmaker> frosch123, that is easy to get, even when it's non-default output
17:09:41  <frosch123> there are some things which can be optimised, would be nice if jenkins gave an overview over random projects
17:09:55  <planetmaker> that can also be done
17:10:44  <frosch123> how would you achieve it via jenkins?
17:10:55  <Alberth> analyse log output?
17:10:58  <frosch123> you would not want to touch all makefiles
17:11:13  <DevZone> Project OpenGFX+ Landscape build #71-push: SUCCESS in 23 min:
17:11:16  <frosch123> Alberth: i mean passing a "-v" option to nmlc
17:11:23  <frosch123> if the info would not be default
17:11:42  <Alberth> export NML = "nml -v"   ?
17:11:59  <planetmaker> <-- I have that file. Calling nml another time would be expensive, but ^ could do the trick
17:12:10  <planetmaker> calling with non-default parameter
17:12:22  <planetmaker> though it's easier if it's default. Agreed
17:12:53  <Alberth> test for presence of NML_VERBOSE  environment var  :)
17:13:17  <planetmaker> <-- but here the call could be adjusted for all projects with one modification
17:13:28  <planetmaker> line 66
17:13:57  <planetmaker> and yes, we could also set an environment var
17:20:37  *** LSky` has quit IRC
17:21:48  <frosch123> env var sounds wrong to me :p
17:21:51  <frosch123> who does that?
17:22:08  <frosch123> omg...
17:22:18  <frosch123> what do you expect to be the most critical limit of yeti?
17:22:54  <V453000> wat
17:23:09  <frosch123> <- maybe we should make it default, and highlight 90% in red or something :p
17:23:35  <Alberth> :D
17:24:03  <frosch123> V453000: my yeti checkout is not the newest (2014-09-25), but you are doing something wrong with industry tiles
17:24:07  <frosch123> you only have 10 left
17:24:28  <V453000> idk tell Sylf
17:24:34  <V453000> best open an issue on devzone :D
17:24:40  <V453000> SYlf already found that issue
17:24:44  <V453000> was adding an industry
17:24:47  <V453000> and ran out of them
17:24:52  <frosch123> ok :p
17:24:58  <frosch123> so, how long will it take to pull?
17:25:02  <V453000> I think he solved it by making one industry tile show different sprites
17:25:03  <V453000> idk
17:25:14  <V453000> probably all sprites
17:26:47  <frosch123> <- anyway, >= 80 is new or very different to wednesday
17:27:00  <frosch123> i mean >= 70
17:28:30  <frosch123> the biggest thing that affects exising projects is probably having to add ".nmlcache" to .hgignore
17:34:19  <Alberth> :O lots of new diffs
17:34:42  <Alberth> I need to make some food first :)
17:34:48  <frosch123> :)
17:37:02  <frosch123> not much new yeti, less than 15 minutes pull
17:37:28  <frosch123> that gives hopes that recompile is not much slower :p
17:40:36  <V453000> what are you trying to do frosch123 ? :D
17:41:28  <frosch123> first goal was to make yeti compile on 32bit systems
17:41:36  <frosch123> next goal was to print id usage
17:41:45  <frosch123> remaining goal is partial compilation
17:42:18  <V453000> very interesting :)
17:43:05  <V453000> good luck! :P
18:04:34  <frosch123> aw, more 3B
18:05:08  <frosch123> encoding crawled to 3k, jumped to 15k, now back to encoding more 3B
18:07:47  <V453000> 3B is animated :)
18:15:48  <Alberth>  paramters      old typo,  760  (, around line 20)
18:19:06  <Alberth> I wonder about having a class for the statistics
18:19:25  <frosch123> they are quite different
18:20:11  <frosch123> there is even a class for id ranges, but it is only used for temporary registers and labels
18:20:21  <frosch123> subtle differences everywhere :)
18:24:04  <Alberth> patches look fine
18:28:32  <frosch123> <- new yeti, two tiles less :p
18:32:31  <Alberth> add more industries to get less used tiles?  :)
18:33:02  <frosch123> well, changing them would break the savegame, so maybe sylf will only do that when really hitting the limit
18:33:28  <frosch123> Alberth: thanks for review :)
18:33:41  <frosch123> i fixed the typo, i guess it's time to push, and see what sets break :p
18:33:55  <Alberth> :)
18:35:24  <Alberth> I never actively switch to a different version of a newgrf in a game, and I keep all old stuff, so savegame compatibility is not a big issue for me :)
18:35:45  <V453000> savegame compatibility with industries is a bit meh tbh
18:36:30  <Alberth> frosch123: It's nice that you move nml forward
18:40:38  <DevZone> Project road-hog build #461-nightlies: SUCCESS in 36 sec:
18:53:19  <DevZone> Project opengfx-mars build #164-push: SUCCESS in 2 min 15 sec:
18:53:36  <DevZone> Project Townnames - German build #106-push: SUCCESS in 7.6 sec:
18:53:39  <DevZone> Project NML - NewGRF Meta Language build #371-push: SUCCESS in 3 min 49 sec:
19:00:09  <frosch123> hmm i recall that an nml push would trigger some other projects
19:00:15  <frosch123> but i see none on jenkins
19:16:10  <frosch123> wut? why does xussrset define cargo items?
19:18:10  <frosch123> hmm, custom cargo icons
19:19:16  <V453000> xd
19:19:19  <V453000> sounds silly
19:19:48  <frosch123> no idea what it actually does, if ecs is present, it redefines all cargos
19:19:55  <frosch123> with possibly different properties
19:20:08  <frosch123> anyway, i thought it was a bug in my nml :p
19:25:36  *** Zuu has joined #openttdcoop.devzone
19:25:38  <frosch123> oh my, all these grfs have such different stats
19:27:21  <frosch123> <- :p
19:32:02  <frosch123> ("generated/pnml/orchard_piggery.pnml", line 56939) <- generated code :p
19:42:30  <V453000> XD
20:04:03  <frosch123> wow, nuts doesn't use a single temporary action2 register
20:04:40  <frosch123> i wonder whether we can create a content matching algorithm
20:04:57  <frosch123> detect author by nml statistics
20:08:03  <V453000> wtf is action2 register :R
20:08:05  <V453000> :D
20:09:40  <frosch123> to translate for you:
20:09:55  <frosch123> "Concurrent Action2 registers" means the grf is complicated
20:10:08  <frosch123> "Concurrent spritegroups:" means the grf is terrible organised
20:10:23  <frosch123> "Concurrent Action10 labels" means the grf is configurable
20:11:36  <frosch123> ogfx+landscape and firs perform quite well under those stats
20:15:25  <V453000> hm :D
20:18:17  <frosch123> i didn't dare to write that in the output though :)
20:48:19  <michi_cc> Checked CETS yet? eddi-nml partial compilation might complicate that though.
20:49:26  <frosch123> oi, i had a question for you
20:49:47  <frosch123> michi_cc: does reusing graphics in container 2 only benefit filesize, or also spritecache size in ottd?
20:51:00  *** LSky` has joined #openttdcoop.devzone
21:04:30  <michi_cc> The sprite cache is indexed by SpriteID and contains a raw pointer to the blitter data, which means filesize only for now. Changing that would mean another indirection with a counter for each raw data.
21:07:50  <frosch123> hmm, SpriteID also need to be consecutive in spritesets
21:08:00  <frosch123> so that cannot be changed either
21:18:27  *** Alberth has left #openttdcoop.devzone
21:27:04  <frosch123> planetmaker: any idea what "documentation" is mentioned in #2785?
21:27:27  <frosch123> i have no idea what could be documented there, so i would suggest to close the issue
21:31:23  *** ODM has quit IRC
21:32:33  <Rubidium_> frosch123: I think it's about the scoping
21:33:20  <Rubidium_> though the current behaviour is pretty typical, *except* when you were used to templated with the C preprocessor
21:33:43  <Rubidium_> which might've been the pitfall triggering the initial issue
21:34:19  <Rubidium_> having said that...
21:35:37  <Rubidium_> how does it behave in ?
21:36:08  <Rubidium_> will the ground sprite be sprite 2 or 4?
22:21:27  *** LSky` has quit IRC
23:31:30  *** frosch123 has quit IRC
23:43:22  *** Zuu has quit IRC

Powered by YARRSTE version: svn-trunk