Config
Log for #openttdcoop.devzone on 2nd February 2014:
Times are UTC Toggle Colours
00:29:01  *** gelignite has quit IRC
07:00:41  *** Supercheese has quit IRC
09:24:30  *** Alberth has joined #openttdcoop.devzone
10:44:26  <Alberth> mars main repo make throws some warnings (one for each sub-repo):
10:44:26  <Alberth> make[1]: warning: jobserver unavailable: using -j1.  Add `+' to parent make rule.
10:45:12  <Elyon> I have no idea how to fix that, sadly :/
10:53:46  <Alberth> I don't even know what it means :)
10:56:37  <Alberth> oh, easy to fix, just  s/make/$(MAKE)/
10:57:04  <Alberth> ie give the main make a chance to pass parameters to its sub-make processes
10:58:39  <Elyon> I don't dare touch the makefiles to be honest ^-^;
11:06:38  <planetmaker> uh, where does it throw that, Alberth ?
11:07:30  <Alberth> main repo, type "make"
11:07:43  <planetmaker> ah, nvm, yes
11:07:59  <planetmaker> I guess I only tested without -jX
11:08:22  <Alberth> I have a global variable setting it to 4 iirc :)
11:08:37  <planetmaker> let's see whether a small rewrite will solve that
11:08:43  <planetmaker> -j6 is adequate :D
11:09:02  <Alberth> there are always people in a bigger hurry :p
11:10:31  <planetmaker> hm
11:10:45  <planetmaker> nah, there's 6 projects which all run only single-threaded
11:11:01  <planetmaker> -jX on normal NewGRF makefiles is rather pointless
11:11:40  <planetmaker> as nmlc is single-threaded
11:12:04  <planetmaker> -jX makes only sense for projects which generate sprites using gimp - which we don't here either
11:18:06  *** gelignite has joined #openttdcoop.devzone
11:19:23  <Alberth> you can build the 6 projects independently
11:20:48  <planetmaker> yes. You can also build a single: make PROJECTS=indutries
11:20:50  <planetmaker> for instance
11:37:32  <Alberth> http://paste.openttdcoop.org/show/3061/    anyidea why airmines don't produce anything, and close after one month?
11:38:11  <Alberth> also, shouldn't nml barf if you don't start with a "substitute" property?
11:42:04  <planetmaker> It probably should do the latter
11:44:50  <planetmaker> I think the production multiplier should at least be 4
11:45:25  <planetmaker> hm... or i likely confuse that with production level
11:47:02  <Alberth> setting it to 4 doesn't fix the problem
11:49:05  <planetmaker> yes, I confused it with the variable 'production_level'
11:49:11  *** frosch123 has joined #openttdcoop.devzone
11:49:20  <Alberth> o/
11:49:36  <Elyon> frosch123, I just want to say thank you for the advanced sprite layout :D
11:50:18  <frosch123> yay, two people using it \o/
11:50:28  <Elyon> :p
11:50:33  <planetmaker> :D
11:50:43  <planetmaker> doesn't NML allow to use it by default?
11:51:00  <Elyon> not for stations!
11:51:02  <frosch123> well, i was thinking about stations
11:51:17  <planetmaker> :)
11:51:17  <frosch123> for industries and objects there are a lot of users
11:51:27  <Elyon> also, I found my other 16 "(pseudo)random" bits
11:52:56  <planetmaker> <Alberth> http://paste.openttdcoop.org/show/3061/    anyidea why airmines don't produce anything, and close after one month? <-- I don't see why that happens :(
11:53:21  <Alberth> probably something stupid I am doing wrong :)
11:53:53  <Alberth> "don't produce anything" as in have no output cargos, btw
11:53:53  <planetmaker> probably not that stupid :)
11:54:09  <planetmaker> that'd be a black hole industry like power plants
11:54:31  <Elyon> isn't it because your type is organic, and that requires trees to produce stuff?
11:54:42  <Elyon> or am I completely wrong there? Shouldn't it be extractive?
11:54:43  <Alberth> this is a black mountain then :)
11:55:41  <Alberth> forest is also organic, I think
11:55:48  <Elyon> so the specs say
11:56:06  <Elyon> maybe try IND_LIFE_TYPE_EXTRACTIVE?
11:56:12  * Elyon is just guessing
11:56:40  <Elyon> as I don't see the difference between a forest and a coal mine, maybe the specs meant to say Lumber Mill?
11:58:49  <Elyon> ah, no, probably not ... the lumber mill just has the IND_FLAG_CUT_TREES bit set
11:58:57  <Alberth> extractive makes no difference
11:59:01  <Elyon> aww :(
11:59:17  <Alberth> property 10 (produced cargo) is said to be not available
12:01:20  <Elyon> why are you setting input_multipliers?
12:04:47  <Alberth> failing random chances to make it work
12:05:17  <frosch123> Elyon: differene between extractive and organic is the colour on the map when selection vegetation view
12:05:23  <frosch123> or something silly unimportant
12:05:30  <Elyon> frosch123: hah, I see!
12:05:36  <frosch123> maybe also the default nearyby-station name
12:23:11  <planetmaker> Alberth, pull. The complaint about the jobserver should be gone
12:23:56  <DevZone> Project opengfx-mars build #7-push: FAILURE in 49 sec: https://jenkins.openttdcoop.org/job/opengfx-mars/7/
12:24:10  <planetmaker> meh
12:25:16  <planetmaker> I guess don't pull :S
12:26:10  <Alberth> last line is missing a \n
12:26:53  <Alberth> either add a + or do $(MAKE)  was sufficient    :)
12:27:42  <planetmaker> sorry, I don't get what you try to tell me :)
12:28:05  <planetmaker> you mean update-all misses that?
12:28:27  <Alberth> +$(MAKE)    <-- here you do both
12:29:25  <planetmaker> both what? adding + and $(MAKE)?
12:29:36  <planetmaker> That is the fix to the job server complaint
12:29:51  <planetmaker> I could just use make instead of $(MAKE)
12:30:26  <planetmaker> ah, you mean I shouldn't have anything else at all
12:30:39  <planetmaker> +changed
12:33:02  <planetmaker> that warning could have been silenced indeed by just adding the + to the front of the lines
12:33:08  <planetmaker> But the make wouldn't be parallel then
12:37:03  <Alberth> $(MAKE) is the recommended way to denote the make program
12:37:20  <Alberth> no need for + signs then
12:38:50  <planetmaker> the + is needed to silence the job server warning
12:39:02  <planetmaker> it indicates that it's safe to run in parallel
12:41:28  <planetmaker> the change from make -> $(MAKE) is independent of that. Yes, I probably should have done more atomic commits :)
12:45:06  <DevZone> Yippee, build fixed!
12:45:07  <DevZone> Project opengfx-mars build #8-push: FIXED in 2 min 8 sec: https://jenkins.openttdcoop.org/job/opengfx-mars/8/
12:57:39  <Alberth> it works for me without +
12:57:54  <Alberth> and with  make -> $(MAKE)
13:01:22  <planetmaker> ah, I see
13:05:56  <Elyon> there wouldn't happen to be a pseudorandom number generator in GRF, would there?
13:06:10  <Elyon> it should be simple enough to implement, but I was just wondering whether there already was one :)
13:06:30  <planetmaker> either the grf feature has random bits or it hasn't
13:06:44  <planetmaker> they can be re-randomized
13:07:13  <Elyon> I noticed :)
13:07:15  <planetmaker> via callback. But possibly that's not true for every occurance of random bits
13:07:55  <frosch123> Elyon: you can create a random number on construction from pretty much everything
13:08:08  <frosch123> you only need the real random bits if you want to rerandomise them on certain triggers
13:08:17  <Elyon> well, I already have that down. I'm trying to select between 256 layouts evenly using "almost truly" random bits
13:08:30  <Elyon> frosch123: yeah, platform number/length seems like a good bet, plus some station random bits
13:08:52  <Elyon> frosch123: wait, create a random number on construction?
13:08:53  <frosch123> you can also use map position, build date
13:09:45  <frosch123> with "create random number on construction" i mean some number which never changes after construction and can be uses as random value more or less
13:10:03  <frosch123> like map position, distance to nearest town, build date and such
13:10:28  <Elyon> frosch123: those seem like better bets than what I was doing. Hmm ... Thank you :)
13:10:40  <Elyon> I can't figure out how to get map position though, haven't found it yet in the specs
13:12:38  <planetmaker> I don't think you can get the absolute position on the map
13:12:55  <Elyon> I can get the position relative to the rest of the station it seems, though, so that's something
13:13:05  <Elyon> and then perform some pseudorandom magic on that
13:13:22  <planetmaker> add the build date and subtract the company# :P
13:13:23  <frosch123> hmm, yeah, position is missing for stations
13:13:32  <planetmaker> absolute position...?
13:13:48  <frosch123> you can get it more or less for industries
13:13:57  <frosch123> clamped to 8 bits or so though
13:14:05  <frosch123> so, really only useful for random data
13:14:09  <frosch123> not for actual map position :p
13:14:14  <planetmaker> hm, how? which var?
13:14:34  <frosch123> industry 0x80
13:15:11  <planetmaker> uhm... are they listed anywhere?
13:15:24  <frosch123> yes, in source
13:15:28  <planetmaker> :P
13:15:37  <frosch123> don't list them on the wiki, their meaning is not granted
13:15:54  <frosch123> people will start using them to try detect map border or something silly, and then report that they don't work
13:16:05  <Elyon> haha
13:16:17  <frosch123> but for "random" they are fine
13:18:58  <Elyon> questions: 1) is there any limit to the number of different cargoes waiting at a station? 2) is there any limit to the number of different cargoes available in the same game?
13:21:26  <Alberth> the latter is 32
13:23:02  <Elyon> Alberth: thanks :)
13:23:29  <frosch123> i think some variables are limited to read 16k units of cargo, if there is more it will be clamped to 16k or so
13:24:07  <Elyon> frosch123: I see, good to know - but how about the number of different cargoes when there are smaller amounts?
13:24:12  <frosch123> so maybe you may not be able to show different graphics for 16k vs 32k units waiting
13:24:26  <Elyon> like, can I have 1 unit of all 32 types waiting at a station?
13:24:31  <frosch123> i doubt there is a 12 limit in ottd
13:24:43  <frosch123> 12 is the number of cargos there were in original ttd, so maybe it applies to ttdp
13:24:49  <frosch123> but in ottd there is no 12 limit anywhere
13:25:31  <Elyon> is the browsable repo supposed to return bad gateway?
13:25:39  <frosch123> ah, yes, 12 is the limit for cargo information when using grf version <= 6
13:27:03  <frosch123> (when using nml you use grf version 8)
13:27:22  <frosch123> (if anyone uses grf version <= 6, they cannot be helped)
13:27:37  <Elyon> I'm using NFO and grf version 8 :)
13:27:51  <frosch123> ah, right, stations :)
13:28:22  <Elyon> indeed! it is funtimes :)
13:28:31  <planetmaker> which repo, Elyon ?
13:28:44  <Elyon> planetmaker: which repo for what?
13:28:45  <planetmaker> (if you mean openttd, better use hg.openttd.org
13:29:11  <Elyon> hg.openttd.org gotcha
13:29:13  <Elyon> thanks :)
13:29:14  <frosch123> planetmaker: is it corrrect that is is fine to use symlinks in zbasebuild to link to opengfx assuming the zbuild subrepo layout?
13:29:37  <planetmaker> why do you need symlinks at all?
13:29:52  <frosch123> to link zbasebuild/sprites/extra/extra-2ccmap.pnml -> opengfx/sprites/extra/extra-2ccmap.pnml
13:29:55  <frosch123> instead of copying it
13:30:46  <planetmaker> you could include that file with relative path instead. Symlink will work as long as people clone the master repo and possibly as long as they don't use windows. Not sure
13:31:10  <frosch123> hmm, ok, include instead of symlink
13:31:58  <frosch123> hmm, currently there are no cross references at all
13:32:10  <planetmaker> except graphics?
13:32:28  <frosch123> nothing uses the "opengfx" subrepo
13:32:33  <planetmaker> hm
13:32:41  <planetmaker> then I wonder why it's there at all
13:32:52  <frosch123> or it is very well hidden
13:33:14  <frosch123> ah wait, there are infact symlinks
13:33:20  <frosch123> extra -> ../opengfx/extra
13:33:23  <frosch123> graphics also use symlinks
13:33:30  <planetmaker> well, then use it :)
13:34:54  <frosch123> let's see whether hg is better with symlinks than svn :p
13:40:21  <Elyon> hrm, I need 128 bits of triggerable random per tile :(
13:40:43  <Elyon> only need to find 124 :p
13:42:19  <planetmaker> Elyon, where the heck you need to choose from 2**128 choices?!
13:42:23  <planetmaker> calc 2**128
13:42:27  <planetmaker> @calc 2**128
13:42:27  <Webster> planetmaker: 340282366920938463463374607431768211456
13:42:49  <planetmaker> I believe something in the concept *could* be reworked
13:42:54  <Elyon> shush
13:43:20  <Elyon> okay, I have 16 cargo slots, each cargo slot can have 8 different cargo sprites
13:43:27  <Elyon> I'd like those to get rerandomised
13:43:37  <Elyon> :D
13:43:53  <frosch123> why do they need randomising independently?
13:44:13  <Elyon> because /awesome/
13:44:23  <Elyon> because there shouldn't be discernable patterns
13:44:29  <frosch123> don't confuse games with spreadsheet applications
13:44:42  <Elyon> hmm. Fair point
13:45:04  <planetmaker> Elyon, no pattern != many choices
13:45:07  <Elyon> I can probably just use the four triggerable bits to select between 16 random-/looking/ layouts of cargo :)
13:45:53  <Elyon> that way I also avoid a random tile having 16 of the same variant of cargo by chance :)
13:46:03  <Elyon> fair points! I will make do with the 4 bits, again
13:49:11  <planetmaker> Elyon, for instance, swedishrails uses 2 pseudo-random bits based on absolute tile position. It decides on showing different fences and transition date for level crossings
13:49:34  <planetmaker> I'm pretty sure that it's nothing people will notice. Even though there's not a single really random bit involved
13:50:00  <Elyon> planetmaker: but there's no absolute position for stations
13:50:15  <Elyon> however, I have 4 really random bits
13:50:18  <Elyon> so I should be fine
13:50:55  <planetmaker> Elyon, exactly. each tile has its own independent random bits. Thus it will be no fixed pattern
13:51:27  <Elyon> :)
13:55:48  *** ChanServ changes topic to "http access to repos currently down for maintenance. ssh access availble, ask if you need it. | Talk about things hosted and developed on http://dev.openttdcoop.org | Downloads log: http://bundles.openttdcoop.org/log.csv | Sandbox passwords are the same as the usernames"
14:00:41  <Alberth> aquiferpump sprites get mixed up, even though I think I do the same as in the airmine
14:01:44  <Elyon> oh no
14:01:47  <Elyon> subset sum :(
14:04:13  <Alberth> oh joy, in NFO :p
14:05:51  <Elyon> I'm more worried about performance than implementing it in NFO to be honest :p
16:18:49  <DevZone> Project NML - NewGRF Meta Language build #238-nightlies: SUCCESS in 1 min 3 sec: https://jenkins.openttdcoop.org/job/nml/238/
16:48:37  <DevZone> Project Dutch Trainset build #69-push: SUCCESS in 1 min 7 sec: https://jenkins.openttdcoop.org/job/dutchtrains/69/
16:50:05  <planetmaker> ^ also jenkins comes with an IRC announcement plug-in, Ammler ;)
16:54:04  <Alberth> but it speaks in the future :p
16:54:18  <planetmaker> in the future?
16:54:48  <planetmaker> oh... yeah, can be understood that way :)
16:55:36  <planetmaker> never noticed :)
17:22:02  <DevZone> Project Japanese Buildings build #99-nightlies: SUCCESS in 16 sec: https://jenkins.openttdcoop.org/job/jpbuild/99/
17:33:58  <DevZone> Project Finnish Rail Infrastructure - Rails build #213-nightlies: SUCCESS in 8 min 12 sec: https://jenkins.openttdcoop.org/job/frissrails/213/
17:48:16  <DevZone> Yippee, build fixed!
17:48:17  <DevZone> Project Iron Horse build #660-nightlies: FIXED in 1 min 27 sec: https://jenkins.openttdcoop.org/job/iron-horse/660/
17:53:06  *** frosch123 has quit IRC
17:56:09  *** frosch123 has joined #openttdcoop.devzone
18:24:16  <DevZone> Project OpenGFX build #125-push: FAILURE in 4 min 9 sec: https://jenkins.openttdcoop.org/job/opengfx/125/
18:24:51  <frosch123> hmm, did i break it?
18:26:25  <frosch123> oh, the Makefile fails
18:28:04  <frosch123> oh well, i'll just run make clean
18:28:28  <frosch123> which throws errors :p
18:30:51  <frosch123> bah, so i actually cannot more sprites around like that :(
18:32:59  <frosch123> so i would need separate files gui1, recolour1, gui2, recolour2, houses1, recolour3, houses2 :/
18:42:12  <planetmaker> eh... I don't exactly understand nml's error message
18:42:21  <frosch123> i moved some sprites around
18:42:33  <frosch123> apparently they must be in order of the sprites in the grf
18:42:37  <frosch123> which makes sense :)
18:42:49  <planetmaker> well :) base sets are rather rigid
18:43:04  <frosch123> the stupid thing is just that the makefile dependencies are complete bollocks, so it compiled fine all the time, without actually compiling it :p
18:43:17  <planetmaker> :D
18:43:36  <planetmaker> rather :(
18:55:36  <frosch123> if there is one thing about annoying me as hell, it's that it prints paths in hg status from the repository root, instrad of from the local dir
18:58:03  <planetmaker> never annoyed me tbh. Where does it come in unhandy?
18:58:22  <frosch123> when i want to c&p the names to open in an editor
18:58:26  <frosch123> or use them in some shell command
19:00:38  <Alberth> hg st .
19:00:50  <frosch123> oh :p
19:01:21  <Alberth> although I agree it would be useful if that was the default :)
19:05:23  <planetmaker> you can easily make yourself an alias in your ~/.hgrc
19:05:58  <DevZone> Yippee, build fixed!
19:05:59  <DevZone> Project OpenGFX build #126-push: FIXED in 5 min 27 sec: https://jenkins.openttdcoop.org/job/opengfx/126/
19:06:37  <frosch123> custom aliases are a thing of last resort, you will be totally screwed if you at someone else's computer :p so i am happy with adding a "."
19:06:47  <planetmaker> :)
19:08:36  <planetmaker> today I actually for the first time really used hgview. Quite useful
19:12:58  <frosch123> sounds like some gui
19:13:32  <planetmaker> yup
19:13:52  <planetmaker> but very nice to navigate, find where things changed, see them in comparison, see files at revs etc
19:14:35  *** Alberth has quit IRC
20:10:19  *** ODM has joined #openttdcoop.devzone
20:15:34  <DevZone> Project xussrset - Trains from Russia build #203-push: SUCCESS in 2 min 42 sec: https://jenkins.openttdcoop.org/job/xussrset/203/
20:54:40  *** frosch123 has quit IRC
21:06:47  *** ODM has quit IRC
21:08:22  *** yorick has joined #openttdcoop.devzone
21:20:51  <DevZone> Project 2ccts build #19-push: SUCCESS in 30 sec: https://jenkins.openttdcoop.org/job/2ccts/19/
21:31:44  <DevZone> Project zBase - the 32bpp base set build #11-push: SUCCESS in 1 hr 9 min: https://jenkins.openttdcoop.org/job/zbuild/11/
23:55:32  *** Supercheese has joined #openttdcoop.devzone

Powered by YARRSTE version: svn-trunk