Log for #openttd on 18th October 2012:
Times are UTC Toggle Colours
00:05:33  *** KritiK [] has quit [Quit: Leaving]
00:45:32  *** tokai|noir [] has joined #openttd
00:45:35  *** mode/#openttd [+v tokai|noir] by ChanServ
00:51:34  *** tokai|mdlx [] has quit [Ping timeout: 480 seconds]
00:57:33  *** dada_ [] has quit [Ping timeout: 480 seconds]
01:00:07  *** APTX_ [] has quit [Read error: Connection reset by peer]
01:00:10  *** APTX [] has joined #openttd
01:03:13  *** dada78641 [] has joined #openttd
01:06:58  *** cyph3r [] has quit [Quit: Miranda IM! Smaller, Faster, Easier.]
01:13:44  *** roadt [~roadt@] has joined #openttd
01:21:47  *** Knogle [] has quit [Read error: Connection reset by peer]
01:22:40  *** Knogle [] has joined #openttd
01:47:16  *** Supercheese [~Password4@] has joined #openttd
02:20:05  *** Elukka [] has quit []
02:39:34  *** Djohaal [] has quit [Read error: Connection reset by peer]
02:49:08  *** supermop [] has joined #openttd
02:49:27  *** DanMacK [~AndChat61@] has joined #openttd
02:49:58  *** DanMacK [~AndChat61@] has quit []
02:56:16  *** glx [glx@2a01:e35:2f59:c7c0:4423:44f6:e56c:37bf] has quit [Quit: bye]
04:55:25  *** telanus [] has joined #openttd
04:56:01  *** Eddi|zuHause [] has quit []
04:56:17  *** Eddi|zuHause [] has joined #openttd
05:19:30  <Supercheese> Did I see a new release candidate is in the works?
05:53:25  *** Prof_Frink [~proffrink@] has quit [Quit: irssi received SIGKILL - going byebyes]
05:54:39  <hmmwhatsthisdo> Who's in charge of the Android port of oTTD?
05:59:57  <peter1138> The author of it, I suppose.
06:02:54  *** sla_ro|master [~slaco@] has joined #openttd
06:09:09  *** KouDy [~KouDy@] has joined #openttd
06:16:26  <hmmwhatsthisdo> peter1138: wasn't sure if it was a team
06:17:01  <peter1138> It might be.
06:27:47  *** supermop [] has quit [Read error: No route to host]
06:29:13  *** blathijs [] has joined #openttd
06:31:38  <Terkhen> good morning
06:31:46  <Supercheese> salve, excitate
06:35:34  *** iklu [] has joined #openttd
07:05:27  *** valhallasw [] has joined #openttd
07:09:23  *** KouDy [~KouDy@] has quit [Quit: Leaving.]
07:10:56  *** KouDy [~KouDy@] has joined #openttd
07:13:31  *** Zuu [] has joined #openttd
07:27:38  *** cyph3r [] has joined #openttd
07:36:25  <NGC3982> Morning.
07:45:11  *** HerzogDeXtEr [~Flex@] has quit [Read error: Connection reset by peer]
07:54:27  <planetmaker> hmmwhatsthisdo, in other words: non of the official devs did the android port. in tt-forums you'll find a thread of its author though
07:54:37  <hmmwhatsthisdo> mmk/
07:58:12  <peter1138> urgh, 268 junction combinations
08:03:45  *** Devroush [] has joined #openttd
08:11:01  *** iklu [] has quit [Quit: Bye]
08:19:02  *** Zuu_ [] has joined #openttd
08:20:11  *** Zuu [] has quit [Ping timeout: 480 seconds]
08:37:20  *** cyph3r [] has quit [Quit: Miranda IM! Smaller, Faster, Easier.]
08:37:42  *** Zuu_ [] has quit [Ping timeout: 480 seconds]
08:54:43  <dihedral> oi
09:04:03  <Yexo> good morning
09:04:36  <Yexo> hmmwhatsthisdo: I have been working on a new android port based on SDL2
09:05:07  <Yexo> that will still require quite some work in fixing the interaction between OpenTTD and SDL2, no idea how much time that'll take
09:11:31  *** hmmwhatsthisdo [] has quit [Ping timeout: 480 seconds]
09:20:09  <planetmaker> peter1138, I don't think that 268 junction combinations is a problem per se. But your point with level crossings, bridges , stations and tunnel portals is very important
09:23:11  <peter1138> planetmaker, only that 268 doesn't fit into a byte, if you felt like caching it
09:23:33  <planetmaker> true
09:23:50  <peter1138> personally i'd ban 180 degree bends ;)
09:23:56  <peter1138> but i don't think that'd go down all that well
09:24:11  <planetmaker> 180° bends... what's a 180° bend?
09:24:24  <peter1138> exactly!
09:25:16  *** cyph3r [] has joined #openttd
09:26:07  <peter1138> hmm
09:26:14  <peter1138> have i messed sound zoom level attenuation?
09:26:24  <peter1138> toyland seems very loud even when zoomed out o_O
09:28:18  *** sla_ro|master [~slaco@] has quit [Quit: DANGER is OFFLINE DANGER]
09:40:08  *** roadt [~roadt@] has quit [Ping timeout: 480 seconds]
09:42:07  <Rubidium> for what it's worth, aren't the images affected by the neighbouring tiles? Then what happens when those bits of tracks are sloped? Does that create yet another explosion of states? Also... shouldn't those changes be made gradual in the same way for tracks?
09:42:46  <peter1138> yes
09:47:37  <peter1138> just don't bring smooth slopes into it ;)
09:47:51  <peter1138> oh maybe you did
09:54:40  *** dada78641 [] has quit [Quit: goodbyte]
09:59:47  *** DDR [] has quit [Quit: for the love of god this is not safe for work]
10:05:07  *** roadt [~roadt@] has joined #openttd
10:05:37  *** dada_ [] has joined #openttd
10:10:52  *** cyph3r [] has quit [Quit: Miranda IM! Smaller, Faster, Easier.]
10:33:26  *** tparker [~tparker@2600:3c03::f03c:91ff:feae:42ad] has quit [Remote host closed the connection]
10:39:13  *** tparker [~tparker@2600:3c03::1d:4242] has joined #openttd
10:47:26  *** bolli [] has joined #openttd
11:13:32  *** roadt [~roadt@] has quit [Ping timeout: 480 seconds]
11:16:05  *** cyph3r [] has joined #openttd
11:17:51  *** roadt [~roadt@] has joined #openttd
11:32:23  *** Supercheese [~Password4@] has quit [Read error: Connection reset by peer]
11:32:53  *** Supercheese [~Password4@] has joined #openttd
11:40:04  <Rubidium> peter1138: ofcourse I did ;)
11:41:25  <TyrHeimdal> Are there plans for autoreplacement of trains to new rail-types?
11:42:07  <NGC3982> You mean: Replacing a diesel Engine with a Maglev engine?
11:42:14  <TyrHeimdal> mhm
11:42:40  <NGC3982> That would also demand being able to change depots while engines are parked.
11:42:50  <TyrHeimdal> It's a real pain to convert, but ofv you know that :)
11:42:54  <TyrHeimdal> yes, true
11:43:20  <TyrHeimdal> you can convert a depot with trains in them, right?
11:43:28  <TyrHeimdal> *engines
11:45:50  <NGC3982> Nope, you cannot.
11:46:08  <NGC3982> That is why this is problematic.
11:46:22  <NGC3982> And as far as i feel about it - It's good as it is.
11:46:31  <NGC3982> They game can't do everything for you. ;-)
11:47:01  <TyrHeimdal> hehe, I agree, but this is one thing I realy could do without ^^
11:47:18  <TyrHeimdal> ofc, an option would be to start at a year with maglev
11:47:19  <NGC3982> I guess you could use the Universal Railtype NewGRF.
11:47:21  <TyrHeimdal> and play from there...
11:47:48  <NGC3982> If you can handle Maglev trains running on nineteen century rails. ;-)
11:51:34  <NGC3982> TyrHeimdal: In a personal perspective (not reflecting the actual reasons for not enabling cross-railtype engine conversions), i do feel like the need for new depots and new trains kind'a reflects a realistic view on building new trains on new systems.
11:52:09  <TyrHeimdal> NGC3982: I agree with you there, but it would be nice as an option
11:52:28  <NGC3982> In a way, it feels more real to build a new facility when creating futuristic magnetic trains, instead of nineteen century cast iron steam junk.
11:52:34  <TyrHeimdal> I'm playing the game to build and learn to create the networks, and make them more and more complex
11:52:51  <NGC3982> And the fact that you can convert Railway > Electric railway suggests that someone else also had that in mind.
11:53:01  <NGC3982> TyrHeimdal: True. An option would be neat, i guess.
11:53:17  <TyrHeimdal> I don't get anything out of changing out hundreds of trains and set up all the new orders
11:53:19  <NGC3982> TyrHeimdal: Though, i highly recommend Universal Railtype GRF for that, until then. ;-)
11:53:32  <TyrHeimdal> Thanks, I'll look into that :)
11:53:51  <NGC3982> It should be in the online content, afaik.
11:54:29  <TyrHeimdal> I'll check when I get back home from work
11:54:40  * NGC3982 waits for someone to chew him out after that "nineteen century cast iron steam junk" part.
12:01:48  *** andythenorth [] has joined #openttd
12:08:52  *** cyph3r [] has quit [Quit: Miranda IM! Smaller, Faster, Easier.]
12:29:10  *** Elukka [] has joined #openttd
12:33:01  *** argoneus [] has joined #openttd
13:13:50  <planetmaker> TyrHeimdal, there are no plans for that. Other than what is possible. You can autoreplace without trouble a steam engine with an electified
13:14:10  <planetmaker> Use compatible track types and it's no issue. Incompatible track types is a feature. Not a bug or missing feature
13:14:25  <planetmaker> It's the oh-so-often asked-for realism (there, I said it, the bad bad r-word)
13:14:59  <planetmaker> (and yes, non-electrified and electrified tracks ARE different track types)
13:21:36  *** Knogle [] has quit [Read error: Connection reset by peer]
13:22:43  *** Knogle [] has joined #openttd
13:29:49  *** KouDy1 [~KouDy@] has joined #openttd
13:33:58  *** KouDy [~KouDy@] has quit [Ping timeout: 480 seconds]
13:34:31  *** argoneus [] has quit [Ping timeout: 480 seconds]
13:56:19  *** andythenorth is now known as Guest2228
13:56:20  *** andythenorth [] has joined #openttd
13:57:38  *** Guest2228 [] has quit [Ping timeout: 480 seconds]
14:02:09  *** andythenorth [] has quit [Quit: andythenorth]
14:08:56  *** andythenorth [] has joined #openttd
14:45:59  *** argoneus [] has joined #openttd
15:03:18  *** andythenorth [] has quit [Quit: andythenorth]
15:08:40  *** andythenorth [] has joined #openttd
15:08:44  <andythenorth> it is raining
15:11:50  *** LordAro [] has joined #openttd
15:12:07  <LordAro> evening ladies
15:12:27  <planetmaker> a statement which seldom is wrong, andy ;-) (if ever)
15:12:34  <planetmaker> hi LordAro and andy
15:14:02  <andythenorth> bonjour
15:14:52  *** AppleGalaxy53 [] has joined #openttd
15:14:55  <V453000> it is raining beer
15:15:29  *** andythenorth is now known as Guest2242
15:15:29  *** andythenorth_ [] has joined #openttd
15:15:29  *** andythenorth_ is now known as andythenorth
15:17:34  <andythenorth> ugh
15:17:43  * andythenorth has to use the word 'artist' in connection with a newgrf :P
15:18:19  <V453000> it is doable in some cases :p
15:19:05  <andythenorth> I dislike it :P
15:19:07  <andythenorth> but me
15:19:09  <andythenorth> h
15:20:02  <planetmaker> why do you *have* to use it (and don't say 'author' or 'contributor')?
15:21:07  <andythenorth> specific to sprites
15:21:15  <planetmaker> painter :-P
15:21:15  <andythenorth> I tried using the word 'painter', it looked silly :)
15:21:59  <Pinkbeast> Designer?
15:22:06  *** Guest2242 [] has quit [Ping timeout: 480 seconds]
15:22:20  <V453000> pixel drawing fits better than painting though I think
15:22:38  *** mahmoud [] has joined #openttd
15:23:27  <V453000> for newGRF in general I would just use author as pm suggested
15:24:08  * andythenorth changes it
15:24:31  * planetmaker pulls a drawer from the desk's drawer :-P
15:24:46  <andythenorth> author :P
15:25:29  <V453000> still, pixel art is a valid category too :P
15:26:54  <V453000> at least that specifies how do you make the newgrf ... you could also be voxel ARTIST, or modeller :p
15:28:35  <peter1138> when will people faking tilt/shift realise that the focal plane needs to be constant?
15:32:04  <bolli> How long is a piece of string surely?
15:32:11  <Pinkbeast> "Drawn by"?
15:42:32  *** bolli [] has left #openttd []
15:50:47  *** HerzogDeXtEr [~Flex@] has joined #openttd
15:51:21  <NGC3982> Evening, democrats and deamons.
15:58:07  *** HerzogDeXtEr [~Flex@] has quit [Quit: Leaving.]
15:58:12  *** HerzogDeXtEr [~Flex@] has joined #openttd
16:06:07  *** Hyronymus [] has joined #openttd
16:08:14  *** sla_ro|master [~slaco@] has joined #openttd
16:12:24  *** pugi [] has joined #openttd
16:23:37  *** valhallasw [] has quit [Ping timeout: 480 seconds]
16:24:33  *** telanus1 [] has joined #openttd
16:30:31  *** telanus [] has quit [Ping timeout: 480 seconds]
16:37:13  *** cyph3r [] has joined #openttd
16:39:40  *** BadBrett [] has joined #openttd
16:42:39  *** telanus1 [] has left #openttd []
16:42:43  *** DanMacK [~AndChat61@] has joined #openttd
16:45:12  <DanMacK> Hey all
16:45:30  <Supercheese> morning
16:51:58  *** roadt [~roadt@] has quit [Ping timeout: 480 seconds]
16:55:34  *** cyph3r [] has quit [Quit: Miranda IM! Smaller, Faster, Easier.]
16:57:08  *** roadt [~roadt@] has joined #openttd
17:08:16  *** Alberth [] has joined #openttd
17:08:19  *** mode/#openttd [+o Alberth] by ChanServ
17:08:59  <Alberth> hi hi
17:09:16  <andythenorth> lo Alberth
17:14:25  <LordAro> hai Alberth
17:16:19  *** TrueBrain [] has quit [Quit: system failure]
17:22:49  *** andythenorth [] has quit [Quit: andythenorth]
17:29:06  *** Progman [] has joined #openttd
17:35:21  *** Wolf01 [] has joined #openttd
17:35:31  <Alberth> hi di hi Wolf01
17:35:44  <Wolf01> hello o/
17:36:04  *** frosch123 [] has joined #openttd
17:43:56  *** telanus [] has joined #openttd
17:50:50  *** keoz [] has joined #openttd
17:57:48  *** DanMacK [~AndChat61@] has quit [Ping timeout: 480 seconds]
17:58:21  *** valhallasw [] has joined #openttd
17:58:58  *** andythenorth [] has joined #openttd
18:01:36  *** TrueBrain_ [] has joined #openttd
18:01:50  *** TrueBrain_ is now known as TrueBrain
18:08:44  *** Zuu [] has joined #openttd
18:13:30  *** AppleGalaxy53 [] has quit [Remote host closed the connection]
18:23:22  <Alberth> right, what would it take to teach NML about parent/child sprites?
18:33:19  <frosch123> in spritelayouts, or in basesets?
18:33:24  *** guru3 [] has quit [Ping timeout: 480 seconds]
18:34:13  <Alberth> base sets
18:34:48  <Alberth> at least that seems like a better approach than manually figuring it out :)
18:34:52  <andythenorth> nml school :)
18:34:52  <frosch123> you would have to code stuff from ottd source (i.e. src/table) into nml
18:34:59  <frosch123> there is no algorithmic method to it
18:35:35  <frosch123> you have to make some table of parent and child sprites, and what offsets ottd applies to them
18:35:39  *** telanus [] has left #openttd []
18:35:50  <andythenorth> Alberth: template it from a .cfg file? :P
18:35:52  <andythenorth> or is that lame?
18:36:01  <Alberth> I was thinking to add a new statement for expressing that
18:36:07  *** DOUK [] has joined #openttd
18:36:46  <frosch123> so more like adding image postprocessing to nml sprites?
18:36:53  <Alberth> ie something like "parent_child(<parent>, <child>)"
18:36:55  <frosch123> add transparency on this or that side
18:37:01  <Alberth> yes
18:37:08  *** guru3 [] has joined #openttd
18:38:08  <Alberth> it should have the sizes and the offsets, so extending the parent seems quite doable
18:38:32  <frosch123> it does not have the offsets
18:38:44  <frosch123> you would have to specify some range for that via some syntax
18:38:52  <Alberth> what offsets?
18:39:14  <frosch123> huh?
18:39:35  <frosch123> what does your "it should have" refer to then?
18:39:51  <Alberth> every sprites has an offset, so I wonder what offset you refer to
18:39:57  <Alberth> *sprite
18:40:01  <frosch123> those between parent and childpsrite
18:40:21  <frosch123> they are not in the grf
18:40:43  <Alberth> I have set the position of all sprites
18:40:52  <frosch123> for basesets they are hardcoded in ottd, unknown to the outside without looking up in the source
18:41:11  *** mahmoud [] has quit [Ping timeout: 480 seconds]
18:41:16  * Alberth looks what's actually in the table
18:41:22  <frosch123> for nml spritelayouts there can be formulas, and nml will have a hard time to detemine the range of them
18:41:56  <frosch123> check stuff like _draw_industry_spec1 and _drawtile_proc1
18:42:24  <frosch123> all those special functions in industry_cmd.cpp, esp. wrt. toyland industries
18:43:14  <Alberth> extending the parent sprite so it it covers the child sprites is not enough?
18:43:54  <frosch123> it is, but you do not know where the child sprite is positioned inside the parent sprite
18:44:02  <frosch123> and the animations even move that position
18:44:32  <frosch123> to make nml the job, you would at least need some method to specify the min and max values from the tables _draw_industry_spec1 and similar
18:45:47  <Alberth> oh, child sprites are not all relative to (0, 0) of the parent?  hmm :(
18:46:00  <frosch123> and even when it is not about basesets, also normal spritelayouts of industies or houses in nml can use formulas to position sprites relatively
18:46:23  <frosch123> Alberth: they are all positioned relatively to (0,0), but there is an additional dynamic offset
18:47:08  <frosch123> if the offsets would be fixed, how could it be animated :)
18:47:11  <Alberth> right, so much for this idea then :(
18:47:39  <Alberth> have different sprites with different relative offsets
18:48:02  <frosch123> all you can do is extending nml to add transparency to sprites (amount specified manually via the nml source), to save you from having to use gimp or some other tool
18:48:56  <frosch123> there is only one lift sprite :) ottd moves it along the house in 12 steps or whatever
18:48:58  <Alberth> doesn't seem very worthwhile
18:49:25  <frosch123> also there is only one sieve sprite and toy sprite, and toy factory stamp
18:52:49  <Alberth> thanks :)
18:54:41  *** HerzogDeXtEr1 [] has joined #openttd
19:00:58  *** HerzogDeXtEr [~Flex@] has quit [Ping timeout: 480 seconds]
19:05:24  *** mahmoud [] has joined #openttd
19:10:22  *** DOUK [] has quit [Ping timeout: 480 seconds]
19:20:00  *** keoz [] has quit [Quit: keoz]
19:27:45  *** ktns [] has joined #openttd
19:29:35  <ktns> When placing an airport to send and receive passengers, does it matter how many houses are in the coverage area?
19:30:56  *** ktns [] has quit [Quit: Leaving]
19:33:17  *** George [~George@] has quit [Read error: Connection reset by peer]
19:34:17  *** glx [glx@2a01:e35:2f59:c7c0:7c05:488c:4109:efec] has joined #openttd
19:34:20  *** mode/#openttd [+v glx] by ChanServ
19:39:51  <NGC3982> Evening.
19:39:57  <Alberth> evenink
19:40:16  *** George [~George@] has joined #openttd
19:41:07  *** KritiK [] has joined #openttd
19:43:06  *** keoz [] has joined #openttd
19:50:10  *** welterde [] has quit [Ping timeout: 480 seconds]
19:51:47  *** DanMacK [~AndChat61@] has joined #openttd
19:52:15  *** sla_ro|master [~slaco@] has quit [Quit: DANGER is OFFLINE DANGER]
19:54:16  *** welterde [] has joined #openttd
20:07:15  *** welterde [] has quit [Ping timeout: 480 seconds]
20:08:43  *** roadt [~roadt@] has quit [Ping timeout: 480 seconds]
20:08:58  *** DanMacK [~AndChat61@] has quit [Ping timeout: 480 seconds]
20:35:26  <planetmaker> oddink
20:36:00  <Alberth> o/
20:38:11  *** keoz [] has quit [Quit: keoz]
20:41:09  <Alberth> Yexo: how do you envision the interface between NML and a web translator?
20:41:23  *** keoz [] has joined #openttd
20:41:25  <Yexo> not, or rather the same as in openttd
20:41:39  <Alberth> that is, by files
20:41:47  <Yexo> yes
20:42:04  <Alberth> I agree on that :)
20:42:38  <Alberth> should the string file have strings that need no translation?
20:42:42  <Yexo> it needs to be possible to compile source offline, so it cannot depend on some online translator
20:43:04  <Alberth> yeah, also, you want to store them in a repo
20:43:08  <Yexo> exactly
20:43:44  *** mahmoud [] has quit [Ping timeout: 480 seconds]
20:43:46  <Yexo> as for strings without translation: I don't care much either way. I don't think the percentage of strings that will definitely never need a translation is high
20:44:15  <Yexo> but in the end I think this is up to how/if a web translator supports that, nml can accommodate afterwards
20:44:27  <Alberth> my reason for adding is that somewhere you need to add that information, and in a translator does not seem the right place
20:44:44  <Yexo> that's a question: is that really necessary?
20:45:02  <planetmaker> Alberth, I *can* imagine that the web translator gets write access to the repos directly on the devzone
20:45:03  <Alberth> those strings do change
20:45:15  <planetmaker> so that it can commit them and needs no intermediate storage
20:45:21  <Alberth> planetmaker: yeah, but I like bitbucket so much
20:45:25  <planetmaker> except maybe temporary for 24h or so
20:45:32  <planetmaker> how does bitbucket do it?
20:45:43  <Alberth> a random repo hosting site
20:45:48  <planetmaker> :-)
20:46:01  <Alberth> or openttd has its own repo
20:46:02  <planetmaker> Alberth, from the webtranslator's side it only needs login+pw to commit
20:46:15  <planetmaker> and URL of course
20:46:24  <Yexo> also: personally I don't care one bit about nml projects not on the devzone
20:46:26  <planetmaker> so, would work there, too
20:46:34  *** l403 [] has joined #openttd
20:46:35  <Yexo> with regards to a possible webtranslator at least
20:46:56  <Yexo> you want that service: host on the devzone and adhere to it's license requirements
20:47:08  <Yexo> don't want to? Host your own webtranslator
20:47:13  <Yexo> but that's just me
20:47:38  <planetmaker> possibly that would simplify some things
20:47:50  <planetmaker> though some interface of WT and projects would be needed anyway
20:48:13  <Alberth> seems like a sane idea, but depositing changes directly in the project repo sounds a bit too direct to me atm
20:48:22  <planetmaker> Alberth, once 24h?
20:48:29  <planetmaker> only needs a bit caching then
20:48:40  <Alberth> you need caching anyway
20:48:44  <Yexo> <Alberth> those strings do change <- for OpenTTD it works to have those "untranslatable" strings in all translations
20:49:12  <Rubidium> there aren't many strings that are always the same
20:49:27  <Rubidium> except the obvious ones like JUST_STRING and friends
20:49:31  <planetmaker> URL, some non-text strings
20:49:41  <Yexo> I think for nml projects that's the same, there won't be that many of them
20:49:46  <Rubidium> but the translators rarely see those strings as those can be 'pretranslated'
20:50:17  <Alberth> FIRS has a number of year strings, which change when the industry introduction changes
20:50:18  <planetmaker> true
20:50:57  <Alberth> Rubidium: but that's manual work, while adding a flag would automate it away
20:51:08  <planetmaker> however, I don't think it matters to have translators translate every single string. Even those where it doesn't really make sense, Alberth
20:51:15  <Yexo> firs can as well be changed to put those years in the code instead of the language file
20:51:20  <Yexo> that might be a better solution anyway
20:51:20  <planetmaker> especially as NML takes care to not include it in the actual grf
20:51:36  <andythenorth> gah
20:51:39  <l403> Hello. I am running gentoo building openttd from ebuild openttd randomly freezes leaving the system unstable and reboot is needed afterward. I've tried with gdb attached strace attached but no wierd errors there, openttd also doesnt spit any errors in console. The process itself becomes inkillable. HOwever someone suggested running valgrind to see if there are any errors there but with valgrind attached it just wont freeze as normal which is sometimes
20:51:39  <l403>  within few seconds still in the menu or after an hour during a game. Can anyone suggest what to do to find what the problem is?
20:51:50  <andythenorth> I have one issue with chameleon for templating: it html escapes some chars
20:51:56  <andythenorth> which is tedious
20:52:16  <Yexo> l403: if your system is unstable I'd say your system is broken, no application should be able to cause that
20:52:33  <andythenorth> Yexo: FIRS can move the years into code, no problem
20:52:45  <BadBrett> jackpot! the pros are here 8)
20:52:54  <Alberth> l403: perhaps trying to use sound?
20:53:06  <Rubidium> l403: try running OpenTTD with allegro instead of sdl to try to rule out sdl being broken
20:53:21  <l403> Yexo, I have not experienced this behavior with any other application
20:53:23  <Rubidium> s/running/building/
20:53:35  <l403> I have sound disabled but I can try different graphics
20:53:41  <Alberth> l403: compiler version?
20:53:44  <l403> dmesg also doesnt complain about anything
20:54:03  <Alberth> Yexo, planetmaker: ok, thanks for the discussion
20:54:42  <l403> Alberth, gcc 4.5.4
20:54:44  <Yexo> Alberth: for the record: even if my above opinions might sound strong, I'm willing to give a lot of room to the person actually implementing a webtranslator tool
20:54:47  <planetmaker> oh, Alberth, and when you feel like a testing environment... we can give you a VM to toy with
20:55:01  <planetmaker> full root and what you need
20:55:15  <Alberth> l403: "inkillable" usually points to a kernel (driver) problem
20:55:16  <Yexo> l403: gcc 4.5 is known to be broken, do you have another version installed?
20:55:34  <planetmaker> and... yes, let me emphasize also what yexo just said wrt "much room" :-)
20:56:30  <l403> Alberth, I was also thinking kernel problem but since I cant get any error messages anywhere I dont know where to start
20:56:34  <Alberth> Yexo: knowing how devs respond, they sound quite reasonable to me :)
20:56:44  <Zuu> Do NML projects use the same layout as GSs with a lang directory which include english.txt etc.?
20:56:56  <andythenorth> near enough
20:56:57  <Yexo> Zuu: yes (only english.lng instead of .txt)
20:56:58  <Zuu> Eg would it be easy to offer this tool also for GS projects?
20:57:00  <Alberth> planetmaker: I am not even near that point, but thanks :)
20:57:14  <Yexo> and yes
20:57:21  <planetmaker> well, it's an offer I don't plan to expire, Alberth ;-)
20:57:37  <Yexo> Zuu: GS lang directory was modelled after OpenTTD's, as is nml's structure
20:57:55  <Zuu> ok
20:58:13  <l403> Alberth, Yexo I dont have any other gcc but I can get 4.4.7 for example. does that sound good?
20:58:14  <Alberth> planetmaker: even if you did, I'd still ask when I needed something like that :)
20:58:22  <planetmaker> :-) good!
20:58:30  <Yexo> nml is lacking in string code support due to limitations in the grf spec, but that should be the main difference
20:58:37  <Yexo> l403: yes
20:58:55  <andythenorth> ho ho
21:00:09  <Alberth> Zuu: it should be generic enough to handle AIs , GSs, NewGrfs, and even openttd itself, imho
21:00:31  <Yexo> l403: this is the gcc bug report, it is not fixed in gcc 4.5, but gentoo might have added custom fixes
21:00:44  <Yexo> so using 4.4 might not help, but I'd be good to rule out that problem
21:01:05  <Yexo> although now I think about it, it's probably not that since you say you can sometimes play for an hour
21:01:33  <Zuu> Alberth: great, although AI support will be hard or easy depending on how you see it as AIs don't support translations.
21:02:11  <l403> Yexo, thanks. I'll take a look
21:02:19  *** Chris_Booth [] has joined #openttd
21:03:56  <frosch123> @topic set 1 1.2.2, 1.2.3-RC1
21:03:56  *** DorpsGek changed the topic of #openttd to: 1.2.2, 1.2.3-RC1 | Website: * (translator: translator, server list: servers, wiki: wiki, patches & bug-reports: bugs, revision log: vcs, release info: finger) | Don't ask to ask, just ask | 'Latest' is not a valid version, 'Most recent' neither | English only | for dev-talk | #openttd.notice for commit notices
21:06:15  <andythenorth> "nml is so bracing"
21:06:44  <Alberth> try LISP instead :p
21:06:48  <andythenorth>
21:06:59  <andythenorth> hmm
21:07:04  * andythenorth has solved more things
21:08:07  <Zuu> Try something truly scope-less for a change. :-p
21:08:40  <andythenorth> ugh
21:08:54  <andythenorth> like nfo? :)
21:08:57  <andythenorth> I guess nfo has scopes
21:11:17  <Zuu> Yea, like single-purpose languages only invented for use within an application.
21:11:42  <Rubidium> what is nfo? If it's the stuff GRFCodec just assembles, then I guess it's effectively scopeless. If it's the meaning of the bytes of the GRF as read by OpenTTD, then it has scopes
21:13:44  <andythenorth> I think of it as unscoped for action IDs
21:13:59  <andythenorth> and scoped for the items that have 'this item' or 'related item' etc
21:14:02  <andythenorth> but meh
21:14:38  *** keoz [] has quit [Quit: keoz]
21:15:10  * andythenorth needs a faster nml
21:15:52  <Alberth> try building a 32bpp baseset every now and then :)
21:16:03  <Wolf01> 'night
21:16:06  *** Wolf01 [] has quit [Quit: Once again the world is quick to bury me.]
21:16:09  <andythenorth> FIRS is about 20s, and I build it every couple of minutes :P
21:16:58  <Rubidium> full zbase is an hour
21:17:02  <andythenorth> :o
21:17:17  <Rubidium> which is annoying when you want to do some sprite offsets or so
21:17:25  <andythenorth> indeed
21:17:27  <Rubidium> ask Alberth about the toyland industries
21:17:28  <andythenorth> partial compilation? :P
21:17:45  <Yexo> does that include use of the sprite cache or is that a full rebuilt?
21:18:02  <Rubidium> it luckily only compiles the needed files
21:18:08  *** keoz [] has joined #openttd
21:18:13  <Rubidium> and the sprite cache helps lots
21:18:30  <Rubidium> bringing it down to like a minute when there are barely any changes
21:18:47  <Rubidium> mostly due to "copying" a 200 MiB file
21:19:20  <Yexo> I wonder how much speed could be gained by rewriting nml in c++
21:19:37  <andythenorth> what about pypy ?
21:21:07  <Alberth> yexo  start with the backend of handling the images
21:21:58  <Rubidium> and the code to emit the GRF?
21:22:18  <Yexo> yep, that could work
21:22:46  <Yexo> first change nml to it generates the "base" part of the grf (all non-image sprites) and a list of {filename, offsets) for real sprites
21:22:53  <Terkhen> good night
21:23:13  <Yexo> then write c++ code to translate that to a file instead of using the current python code + spritecache
21:23:13  <Alberth> Yexo: just hook pieces of C into nml
21:23:58  <Yexo> Alberth: it has to be quite a big part, hooking a too small part causes too much overhead with the python<>native conversions
21:24:08  <Yexo> I've tried that already with an lz77 implementation in c
21:24:18  * andythenorth tries to figure out how to get the FIRS makefile to use grfcodec
21:24:33  <Alberth> use cython may also be an option
21:25:00  <Yexo> it certainly was a nice project to learn python, but looking back I wouldn't start with python now
21:25:01  <andythenorth>
21:25:13  <Yexo> (and that's not just because of performance reasons)
21:25:18  <andythenorth> doing it in python got you some commits by me :)
21:25:20  <Alberth> nml has grown quite a lot I think
21:25:30  <andythenorth> and I have learnt some of the format by just reading source
21:25:55  * Yexo missing static typing
21:26:25  <NGC3982> Im sorry, but how does the Openttd site know what kind of CPU i use?
21:26:41  <Alberth> yeah, that's the biggest problem of bigger non-trivial Python programs
21:26:52  <Alberth> NGC3982: OS version?
21:26:55  <andythenorth> NGC3982: it checks some user agent properties
21:27:09  <NGC3982> andythenorth, Alberth: Is that supplied by my browser?
21:27:13  <andythenorth> i.e. your browser yes
21:27:19  <NGC3982> Ah, i see.
21:27:24  <NGC3982> A tad suprising.
21:27:30  <andythenorth>
21:27:31  <Alberth> NGC3982: yes, it gives a lot of information
21:28:26  * andythenorth -> bed
21:28:27  *** andythenorth [] has left #openttd []
21:29:07  *** valhallasw [] has quit [Ping timeout: 480 seconds]
21:31:41  <Guest1496> Yexo: why not python, btw?
21:32:09  <Yexo> as I said: no static typing (which is fine for small projects, but for me quickly becomes annoying the more a project grows) and speed
21:32:16  *** keoz [] has quit [Quit: keoz]
21:32:58  * Guest1496 shrugs
21:33:02  *** Guest1496 is now known as Xaroth
21:33:12  * NGC3982 finally upgrades to 1.2.2
21:33:34  *** Xaroth is now known as Guest2295
21:33:50  <Yexo> will you wait on 1.2.3 before you upgrade to 1.2.3-RC1?
21:33:54  <Guest2295> grr, stupid nickserv
21:34:10  <NGC3982> Yexo: I ..don't know.
21:34:24  <NGC3982> Im not really following the updating structure of OpenTTD
21:35:14  <Alberth> good night
21:35:38  <Yexo> good night
21:36:02  *** Guest2295 is now known as xaroth
21:36:07  *** Alberth [] has left #openttd []
21:37:13  <xaroth> Yexo: the non-static-typed nature of Python can also be used for it's advantage, but I do agree that at times it can be annoying as fook
21:37:47  <xaroth> and speed, well, things can be optimized
21:38:00  <Yexo> sure, but only to a certain extend
21:38:07  <xaroth> for a long time EVE online ran both client and server(cluster) fully on python
21:38:42  <xaroth> so optimizations can go a long way with python
21:39:21  *** Chris_Booth [] has quit [Quit: ChatZilla 0.9.89 [Firefox 17.0/20121010150351]]
21:39:26  <Yexo> sure, but as soon as you go number-crunching / image processing you can't beat a lower-level language
21:39:32  <xaroth> yep
21:39:41  <xaroth> but that's what the native bindings are for
21:39:46  <xaroth> well, the image processing part at least
21:39:52  <NGC3982> Where can i alter the setting of if "Go To" should be unmarked or not, after setting a destination in the order window?
21:39:54  <TinoDidriksen> Stackless Python...not normal Python.
21:41:11  <frosch123> night
21:41:15  *** frosch123 [] has quit [Remote host closed the connection]
21:42:08  <xaroth> TinoDidriksen: Microthreads are hardly useful for something like NML
21:42:09  *** KouDy1 [~KouDy@] has quit [Quit: Leaving.]
21:42:30  <TinoDidriksen> No, I meant EVE uses Stackless, not bog standard Python.
21:42:54  <xaroth> yes, they use it for the microthreads
21:42:59  <xaroth> but the main point remains
21:43:50  <xaroth> optimization in python is key
21:45:32  <Yexo> nml is not just a bit too slow, it's several orders of magnitudes slower than I think is possible
21:48:14  <Yexo> NGC3982: Advanced Settings -> Interface -> Quick creation of vehicle orders
21:52:43  *** Prof_Frink [~proffrink@] has joined #openttd
21:55:02  <Yexo> good night
21:55:46  <BadBrett> why not just use a GUI to automatically set the offsets like I do?
21:57:36  *** Progman [] has quit [Remote host closed the connection]
21:58:51  *** argoneus [] has quit [Ping timeout: 480 seconds]
22:03:55  <NGC3982> Yexo: Oh, thanks.
22:35:22  *** Zuu [] has quit [Ping timeout: 480 seconds]
22:39:33  *** Devroush [] has quit []
22:46:48  *** Hyronymus [] has quit [Quit: Hyronymus]
23:18:30  *** LordAro [] has quit [Remote host closed the connection]
23:25:28  *** TheMask96 [] has quit [Ping timeout: 480 seconds]
23:30:21  *** TheMask96 [] has joined #openttd

Powered by YARRSTE version: svn-trunk