IRC logs for #openttd on OFTC at 2011-06-15
⏴ go to previous day
00:14:41 <planetmaker> Terkhen: the remaining wiki pages have been bot-edited. What still might need checking if somewhere content wrt the TTDP wiki is missing. And the usual general review, of course. But everything which has not been looked at should now approx. follow the style
01:15:00 *** fjb is now known as Guest4631
02:27:20 *** rhaeder has joined #openttd
04:14:43 *** DayDreamer has joined #openttd
04:15:23 *** Adambean has joined #openttd
04:56:20 *** Eddi|zuHause has joined #openttd
05:08:10 *** andythenorth has joined #openttd
05:24:54 *** Nite_Owl has joined #openttd
05:56:33 *** sla_ro|master has joined #openttd
06:18:28 *** Br33z4hSlut5 has joined #openttd
06:51:16 <planetmaker> sounds partially duplicate to the - also still draft - wiki page on newgrf recommended behaviour
06:51:35 *** andythenorth is now known as Guest4656
06:51:35 *** andythenorth has joined #openttd
06:52:11 <planetmaker> I'd say: go for it. But possibly split it in at least two: the getting started and the newgrf development recommendations
06:52:34 <andythenorth> planetmaker: I think the recommendation is a bit overwhelming for those entirely new
06:52:44 <andythenorth> I'd see it as information in stages...
06:52:53 <planetmaker> but it's virtually that by your list ;-)
06:53:07 <planetmaker> well, the name doesn't matter
06:55:05 <andythenorth> detailed pages might be: standards | tools | tutorials
06:55:11 <andythenorth> linked by a 'getting started' guide
06:55:56 <andythenorth> standards is more like a check-list
06:59:57 *** Cybertinus has joined #openttd
07:05:56 <Terkhen> planetmaker: great :)
07:10:35 <planetmaker> that was a list of pages like the main page. Auto-generated
07:11:46 <planetmaker> yes, we should. These pages should work via categories
07:12:28 <planetmaker> That actually should probably now come up with a few meaningful categories for pages which make sense to be displayed in a common overview page each for that particular "topic"
07:12:28 <Terkhen> how do you remove a page? :P
07:14:02 <planetmaker> deleted. I suppose you don't have a 'delete' button next to 'edit'?
07:14:57 <planetmaker> then you'll need to ask orudge to also become administrator, I guess
07:15:02 <Terkhen> you should also write a post about newgrf specs being formatted and available again at the news thread :P
07:15:31 <planetmaker> I guess you're right.
07:16:30 <planetmaker> we should still review the pages the bot fixed that nothing got missing in the convert. The bot didn't remove anything but as we know sometimes really stuff got missing in the conversion, esp. with formulas in the page
07:17:13 <Terkhen> it seems that you forgot to add the s/</</g changes
07:17:26 <Terkhen> but mostly they look fine :)
07:17:58 <Terkhen> oh wait, this page was not edited by the bot :P
07:20:17 <Terkhen> there are a few broken tables because of internal links, nothing else :)
07:20:29 <planetmaker> in the bot-edited pages?
07:20:43 <planetmaker> hm... then I missed that before I fixed that
07:21:16 <planetmaker> eventually I did that manually
07:21:44 <planetmaker> as the page got messed up and I didn't want to do a rollback-edit orgy and most was done correctly. Except ... some stuff with tables
07:22:18 <planetmaker> and yes, there that bug was still present
07:23:35 <planetmaker> but... where is that messed up table? I don't see it
07:25:23 <Terkhen> I fixed it, check the diff I pasted
07:38:02 <planetmaker> looks better with, actually
07:38:09 <planetmaker> but it's then the only one
07:40:40 <planetmaker> and... should we use Action0, Action1, VarAction2 (that's what I prefer) or Action 0, Action 1, VarAction 2?
07:45:02 <planetmaker> and... should we keep tasks 'done' in the TODO list? Seems a bit pointless to me
07:46:03 <planetmaker> hm.. but is psychologically good ;-)
07:59:23 *** JVassie_ has joined #openttd
08:03:13 <Terkhen> planetmaker: IMO without space
08:03:27 <Terkhen> otherwise it needs a lot of changes and people might be used to the old format
08:03:37 <planetmaker> hm, I did it now with... but yes, maybe
08:05:44 *** TWerkhoven has joined #openttd
08:16:17 <Zarra> In 1.1.1 problem with AI! Not workin'... who know the solution???
08:18:05 <Yexo> Zarra: how is it not working?
08:18:07 <Yexo> what message do you get?
08:18:23 <Yexo> did you follow the advise in the error message (if any)?
08:20:18 <Zarra> The program does not see a source of AI
08:20:32 <Yexo> did you download them via the online content system?
08:22:21 <Zarra> No. Cuz there's no content... :(
08:22:37 <Yexo> so how did you download them?
08:22:54 <Yexo> if you downloaded any .tar.gz files, you need to extract them to .tar files
08:23:01 <Yexo> you might also be missing the AI libraries
08:23:48 <Zarra> I simply have not waited the response :(
08:25:25 <Zarra> Found them!!! Sorry to bother!!! Odessa - Ukraine = OUT!
09:04:17 *** andythenorth has joined #openttd
09:12:22 *** Progman has joined #openttd
09:37:47 *** hgnmu128 has joined #openttd
09:38:30 * andythenorth is going to have tea
09:38:45 <andythenorth> tea is more available than ponies
09:47:53 *** keky___ has joined #openttd
10:03:32 <Terkhen> fine, my project is finally working as it should :)
10:04:19 * JVassie_ is doing planning for his project
10:08:50 <JVassie_> unfortunately my lack of understanding how stations are coded
10:09:06 <JVassie_> has meant me pixelating stuff into objects which can be broken down and reused
10:18:39 <JVassie_> a thought, not sure if you'd knwo the answer Terkhen
10:18:56 <JVassie_> is there anyway I can plan to support multiple tracksets, for non track tiles?
10:19:08 <JVassie_> ie use their 'underlay' but not the rails
10:19:49 <Terkhen> sorry, I have never touched stations or railtypes :/
10:20:03 <Terkhen> the station pages are scary :P
10:20:58 <JVassie_> Im hoping someone will write some more examples and/or tutorials in the appropriate places :D
10:21:10 *** Adambean has joined #openttd
10:21:19 <JVassie_> more NFO developers can only be a good thign I guess
10:21:53 <planetmaker> lobby for station specs included in NML, JVassie_ ;-)
10:22:04 <planetmaker> that probably would make it easier for the newgrf author ;-)
10:22:23 <JVassie_> how easy is NML to setup on a Win7 box?
10:22:33 <JVassie_> for someoen with no Python experience
10:22:48 <planetmaker> easy. As it comes with an all-inclusive installer / exe download
10:22:51 <Yexo> you don't need python, there are windows binaries
10:23:47 * JVassie_ wonders if he should do a new version of his grfmaker guide but for NML
10:23:56 <JVassie_> .doc with lots of explainign and pictures
10:24:06 <planetmaker> I'd very welcome that
10:24:15 <JVassie_> gotta learn NML then I guess :p
10:24:17 <Yexo> me too, although I'd prefer html with pictures
10:24:23 <planetmaker> tt-wiki now would be a good place. IMHO it s/should get an NML section anyway
10:24:24 <JVassie_> html is easier for me tbh
10:24:27 <Yexo> that way it could be included in the nml documentation
10:24:39 * JVassie_ wonders off to find NML .exe
10:24:41 <planetmaker> either way is fine
10:26:13 <JVassie_> no donotreadme included
10:27:28 <JVassie_> no docs folder in the win32 bundle
10:28:50 <JVassie_> is the .exe included an installer?
10:29:02 <planetmaker> George: I ran the bot on the list of pages you supplied me with. I did not check the result on every page though, only those where the diff looked suspicious. Most likely there will be some pages where it failed in one way or another to fix the tables completely
10:29:03 <Yexo> no, it's the actual executable
10:29:28 <planetmaker> George: but many of the translation pages seem to not exist. Thus they were not touched
10:29:41 <planetmaker> probably they're named slightly differently
10:30:02 <JVassie_> so looks like in business
10:32:55 <JVassie_> the docs are much more in depth than i thought :D
10:33:22 <hgnmu128> I think he created NML himself by reading the docs.
10:35:16 <Eddi|zuHause> "in the beginning was the docs"
10:35:55 <planetmaker> JVassie_: even though, there's still lots of room for improvement
10:36:17 <JVassie_> I was planning on working through a tutorial for coding a locomtive + coach
10:36:23 <JVassie_> similar to the grfmaker tutorial I did
10:36:29 <JVassie_> and then add more advanced topics
10:36:41 <JVassie_> like livery overrides, refitting, etc
10:37:13 <JVassie_> planetmaker, which features work in NML so far, trains and?
10:37:34 <Yexo> JVassie_: all vehicles, industries, airports (as far as they are supported by the spec)
10:37:43 <orudge> planetmaker: or would you desire a list, I guess, first
10:37:56 *** JVassie_ is now known as JVassie
10:38:01 <Yexo> I think everything except stations and houses
10:38:01 <planetmaker> orudge: yes, plain text list of page names as they appear in the URL
10:38:08 <JVassie> Stations is what I want most :D
10:38:12 <planetmaker> will be no big problem
10:41:49 <orudge> since they've come up quite a bit in the TTDPatch manual while I was doing it manually
10:43:40 <planetmaker> most. I remove(d) the +- stuff
10:43:47 <JVassie> One thing I've noticed thats missing in the docs is a quick guide on how to actually run NML to make a grf
10:43:56 <Eddi|zuHause> shouldn't you do s/ / /g?
10:43:57 <planetmaker> but easy to change
10:44:03 <orudge> Eddi|zuHause: not generally, they're usually redundant.
10:44:06 <planetmaker> that's what I do, Eddi|zuHause
10:44:12 <orudge> Eddi|zuHause: e.g., people double-spacing things.
10:44:21 <orudge> at least in my experience
10:44:35 <orudge> planetmaker: if you could keep the <tt> style, I'd prefer that
10:44:45 <Eddi|zuHause> orudge: but double spaces should be handled by the software itself?
10:44:53 <orudge> Eddi|zuHause: well, yes. I just like to tidy things up and remove them. :p
10:45:10 <Eddi|zuHause> orudge: s/ +/ /g?
10:45:52 <Eddi|zuHause> orudge: imho better than suddenly having no space somewhere
10:46:32 <orudge> inevitably there'll be more later ;)
10:47:21 <George> planetmaker: George: but many of the translation pages seem to not exist. Thus they were not touched <- ask orudge
10:47:29 <orudge> they've not been imported
10:47:32 <orudge> need to figure out how best to handle them
10:47:34 <Yexo> JVassie: "nmlc filename.nml" will create filename.grf
10:48:01 <orudge> planetmaker: you could also add Tutorials, PreSignalsTutorial, GRFTutorial and DrawingSprites to that list
10:50:01 <planetmaker> that's the literal page names, not sections with several pages, right?
10:51:30 <orudge> planetmaker: yep, unless I missed any.
10:52:12 <planetmaker> orudge: I remove lines with ''''' as start (the first line), too. Ok?
10:53:09 <planetmaker> and any {toc} and {maketoc}
10:58:45 *** [WOODY] has joined #openttd
11:13:57 *** HerzogDeXtEr has joined #openttd
11:20:11 *** amkoroew has joined #openttd
11:20:37 <planetmaker> hm... the bot waits 9.6s between two subsequent page changes...
11:20:59 <orudge> Seems to be working, though :)
11:22:13 <Yexo> planetmaker: if you use the same python framework as I do, pass "-pt:0" as argument
11:26:35 <planetmaker> that was an excellent tip, Yexo
11:27:43 <planetmaker> It's quite an easy-peasy framework to use for automatic wiki edits
11:27:51 <planetmaker> could certainly also be abused as spam bot :-P
11:28:29 <Yexo> luckily the bot interface is not open for anonymous edits :)
11:29:02 <planetmaker> thus it needs "only" a wrapper to register spam accounts ;-)
11:29:22 <orudge> well, hopefully, the challenge of having to register on the forums rather than the wiki will deter spammers
11:30:08 <planetmaker> orudge: I just discovered that I might have accidentially replaced a <pre>in some instances by a one-space indentation.... it looks 95% the same, though
11:30:11 <Yexo> planetmaker: can normal accounts without the bot flag even use the api?
11:30:26 <orudge> planetmaker: ah, well, I'll be looking through everything slowly but surely anyway
11:30:33 <orudge> still have to transfer graphics over and ideally categorise pages
11:30:35 <planetmaker> Yexo: I think so.
11:30:37 *** Chris_Booth has joined #openttd
11:31:55 <planetmaker> last pages are now updated, orudge
11:32:12 <orudge> now mb really has nothing to complain about ;)
11:32:26 <orudge> there may be a few other pages I've missed
11:32:29 <orudge> but I'll look through later and see
11:32:43 <planetmaker> and a few where the diff looked ok, but the result is ugly ;-)
11:32:56 <planetmaker> I judge the commit / not commit by the diff I get
11:33:15 <planetmaker> though I presses 'y' in all but 3 cases or so
11:34:25 <Ammler> <tt> and <pre> needs to be there?
11:35:00 <Ammler> the page looks quite broken
11:35:20 <orudge> yes, something went a bit wrong there
11:35:27 <JVassie> <Yexo> JVassie: "nmlc filename.nml" will create filename.grf
11:35:51 <JVassie> any old text editor will do i presume?
11:36:26 *** ashledombos has joined #openttd
11:37:27 <Chris_Booth> older text editors are more fun :P
11:38:59 <planetmaker> page fixed, Ammler
11:43:37 <planetmaker> orudge: Yexo : this is basically what I use wrt pywikipediabot's trunk (r9294)
11:43:54 <planetmaker> the fixttdp.py is a modified version of replace.py
11:44:28 <planetmaker> :-P Meant to include the line above :-P
11:45:06 <orudge> oh hey, it must have just been Patch Tuesday
11:45:08 <orudge> 18 new Windows Updates
11:50:04 <planetmaker> hm... this Hgnmu128 guy has no real clue what he writes about... and what NewGRF variables are and still 'fixes' the town radius information...
11:56:37 <Yexo> which as discussed yesterday is copyrighted
11:57:35 <planetmaker> also that. But the info isn't helpful beyond what the var description said but even mis-leading or wrong
11:58:08 <Yexo> "The first zone is where station activity contributes to growth of the town," that might be wrong, the rest looks correct to me?
11:58:23 <planetmaker> and even from the source I don't understand quite what the difference between, say 0x94 and 0x95 is
11:58:44 <planetmaker> Yexo: to me it implies it's something which can be changed by the newgrf. Which is stupid at least
11:59:05 <planetmaker> These values are frequently updated by TTD, so changing them has little or no effect in the long run.
11:59:06 <Yexo> var 0x94 is the size of the first town zone, 0x95 the size of the second town zone
11:59:28 <Yexo> not really, but I thought something like that
11:59:36 <planetmaker> lampToU16(t->squared_town_zone_radius[0]); and GB(ClampToU16(t->squared_town_zone_radius[0]), 8, 8);
11:59:48 <planetmaker> thus it's something related to the same zone. But I don't know the difference
11:59:57 <planetmaker> that's 10 vars as opposed to the 5 I'd expect
11:59:58 <Yexo> and keep in mind that the info was copied from a page about the TTD savegame format. Changing values made perfect sense for that page (=binary changes in the savegame), but indeed not for a varaction2 page
12:00:30 <Yexo> planetmaker: oh, it's 5 word values
12:00:40 <Yexo> so var 94 is a word, var 95 is the high byte of var 94
12:00:50 <planetmaker> hm, yes. But why?
12:00:56 <Yexo> which again has to do with how ttdpatch works, in ttdpatch all 80+ vars are direct memory access
12:01:18 <Yexo> you can read var 94 as double-word, shift right 16 bits and you have var 96
12:01:33 <Yexo> in openttd that obviously doesn't work, but it's emulated for the high bits
12:01:42 <Yexo> so you can read just var 95 (=high word of var 94)
12:02:20 <planetmaker> he... I wonder whether those vars need special mentioning actually. Just the word ones are enough
12:02:43 <Yexo> only 94, 96, 98, 9A and 9C should be used, at least for new newgrfs
12:02:46 <Yexo> if they need those at all
12:02:47 <andythenorth> it's not clear what town radii do anyway - I only know because frosch explained me
12:02:54 <andythenorth> I read src and am still not much wiser
12:02:59 <andythenorth> picture = 1k words?
12:03:12 <Yexo> I know approximately what they do, but I have no idea about the values of those vars
12:03:25 <andythenorth> I wouldn't be able to write newgrf using those values
12:03:33 <andythenorth> if I could, it would be used in FIRS already :P
12:03:50 <Yexo> you could check the newgrf debug gui in openttd to see what the values are
12:03:56 <Yexo> hmm, if those vars are shown there
12:04:19 <andythenorth> only if you have a town newgrf active I guess
12:06:00 <Yexo> if you have a house newgrf active you can see those values
12:06:18 <Yexo> town with size 566: values are 0x40, 0, 4, 0, 0
12:07:07 <Yexo> hmm, other town, size 730, values: 0x31, 0, 4, 0, 0
12:07:45 <Yexo> town with 2115 inhabitants: 0x51, 0x24, 0x19, 0x10, 0x09
12:08:43 <Yexo> planetmaker: at any rate I'd revert those changes by hgnmu128
12:09:07 <planetmaker> yes, I'm thinking of just editing it. Or do you mean rollback?
12:09:39 <Yexo> that stuff is copyrighted, so if it's not too much work I'd rollback and write your own description
12:14:38 <JVassie> planetmaker, isnt FF a valdi language ID for any language?
12:15:27 <planetmaker> JVassie: it's the default language 7F and FF
12:15:41 <planetmaker> each language has two IDs: value and value + 0x80
12:15:46 <JVassie> Im just working through a language file
12:15:53 <JVassie> and FF isnt in the list of IDs is all
12:16:01 <planetmaker> thus valid language IDs are 0x00 ... 0x7F
12:16:08 <planetmaker> 0x7F is default and thus also unavailable
12:16:28 <JVassie> ill stick with 0x01 then i guess
12:16:58 <Yexo> JVassie: what do you need it for?
12:17:11 <Yexo> if you don't have any translations, you should only use 7F / FF
12:17:33 <JVassie> its my lang file called default.lng
12:18:02 <JVassie> it should use the same strign in every language
12:18:56 <Yexo> oh, for an nml language file, use 0x01 there :)
12:20:01 <JVassie> might have my first NML-made grf shortly :D
12:21:34 <JVassie> I use ConTEXT normally, but thanks :D
12:21:59 <JVassie> hmm does the .nml file need to be in the same folder as nmlc.exe?
12:22:07 <JVassie> or can it be in a subfolder?
12:24:16 <Ammler> Terkhen: maybe could be added to nml repo as contrib
12:25:04 <Terkhen> Ammler: that would mean that someone will maintain it :P
12:25:15 <Yexo> planetmaker: wow. I still don't see how that would be useful, but at least it's complete :)
12:25:31 <planetmaker> I don't either. But I found it interesting ;-)
12:25:41 <planetmaker> And for things like TAI it might indeed be interesting
12:25:43 <Terkhen> I created it copying the tokens by hand myself, and I don't code nml in windows anymore
12:27:06 <planetmaker> Yexo: what was new to me is that actually at _every_ town size there's at least one town zone which is not available
12:27:46 <planetmaker> hm... not true. For 52 houses... there are all around ;-)
12:28:02 <planetmaker> but that possibly up to 55 ;-)
12:28:02 <Terkhen> planetmaker: really nice :)
12:28:35 <Ammler> Terkhen: maybe you make one for kate :-)
12:29:00 <Ammler> and there you have it?
12:29:49 <welshdragon> rcon psswd 'setting network.server_name' fails while running the game
12:30:33 <Terkhen> it seems to use scintilla lexers, I don't know what are those
12:30:36 <planetmaker> try without "network."
12:30:53 <Terkhen> but at least it seems more widely used than notepad++'s own format
12:31:14 <welshdragon> Error: Command not found, planetmaker
12:32:07 <Terkhen> although... notepad++ is listed as supported
12:32:58 <welshdragon> Error: Command not found, planetmaker
12:33:30 <planetmaker> welshdragon: please read the wiki page on how to use rcon. You probably also miss " somewhere
12:34:15 <welshdragon> planetmaker, it's ok, I'll edit the config :)
12:34:27 <planetmaker> that won't help you ingame a single bit
12:34:49 <planetmaker> nor for any existing map or scenario
12:35:04 <JVassie> nmlc: default language file "lang\english.lng" doesn't exist
12:35:48 <planetmaker> JVassie: then... how's your file called? :-)
12:36:03 <planetmaker> I *think* it needs to be indeed english.lng
12:36:19 <Yexo> there is a command-line flag to use another name for the default language
12:36:30 <planetmaker> but that's tedious in the long run ;-)
12:36:34 <JVassie> well, my nml file is in a folder called Project1, within that folder is a subfolder called lang, in which tere are two files (both identical) called default.lng and english.lng
12:37:00 <planetmaker> and where do you call nmlc?
12:37:30 <JVassie> C:\Games\Tt\NML is the folder nmlc.exe is in
12:37:41 <JVassie> and im using the command nmlc Project1/BB27000.nml
12:38:10 <planetmaker> JVassie: you need to go into Project1 dir
12:38:13 <Yexo> JVassie: better do "cd Project1" first
12:38:13 <planetmaker> and then call nmlc
12:38:22 <Yexo> than do "../nmlc BB27000.nml"
12:40:16 <JVassie> i copied the files to the main NML directory
12:40:26 <JVassie> well until it gets to my syntax error(s)
12:40:34 <Yexo> yes, but that breaks as soon as you start your second project
12:40:41 <Yexo> as both will try to use the same lang/ directory
12:40:41 <Ammler> wouldn't adding nml path to %path% suiffice?
12:40:57 <Yexo> Ammler: yes, but in general that's harder than using ..\
12:41:08 <JVassie> <Yexo> oh, try ..\ instead
12:41:14 <JVassie> same syntax error shown
12:41:23 <JVassie> I guess weight doesnt need a 't'
12:43:00 <JVassie> Syntax error, unexpected token "t"
12:43:10 <JVassie> on the line where id set the property
12:43:25 *** Chris_Booth has joined #openttd
12:43:47 <JVassie> same idea with 30years and not 30 years i presume
12:43:49 <planetmaker> units must follow their numbers immediately
12:44:17 <Yexo> hmm, didn't even know about that
12:44:36 <JVassie> settign tractive effort
12:44:47 <Yexo> <planetmaker> units must follow their numbers immediately <- actually that's not true
12:44:53 <JVassie> i presume tractive_effort_coefficient isnt for this purpose as it has a range of 0..1
12:45:02 <planetmaker> hm... maybe it's then tons instead of t
12:45:28 <Yexo> both "ton" and "tons" are valid
12:46:03 <JVassie> i get an invalid token s if i put tons
12:46:13 <JVassie> and it doesnt like years
12:46:29 <JVassie> easiest to skip the units alltogrther for these? :D
12:46:36 <planetmaker> leave out the years... if a var is called availability_years no point to add a unit
12:46:57 <planetmaker> not all properties have units. coefficients intrinsically cannot have, years, ...
12:47:10 <planetmaker> and TE is physically a value between 0 and 1
12:47:14 <planetmaker> Thus that's how NML handles it
12:47:28 <planetmaker> forget NFO values. They're... arbitrary at best
12:47:32 <JVassie> flipping it on its head
12:47:38 <JVassie> how would i set a max TE of 300 kN then?
12:47:39 <Yexo> JVassie: "years" nor "year" is a valid unit
12:47:47 <Yexo> "tons" should work, but doesn't, that's a bug in nml
12:49:16 <Yexo> the "tons" issue is fixed in r1405
12:49:21 <planetmaker> JVassie: TE is weight [tons] * 10 * TE_coeff
12:49:35 <planetmaker> just simple physics
12:49:45 <planetmaker> and NFO also only knows the TE_coeff
12:49:47 <JVassie> thanks for explaining though
12:50:00 <JVassie> which does that calc for you
12:50:08 <JVassie> perhaps addition for future NMl version?
12:50:54 <JVassie> TE is weight [tons] * 10 * TE_coeff
12:51:09 <JVassie> you are allowed in grfmaker to input a TE value, not the coefficient
12:51:17 <JVassie> it calculates the coefficient for you
12:51:23 <planetmaker> are you sure it's labelled correctly there?
12:51:23 <JVassie> for example in this case
12:51:32 <JVassie> it willingly takes 300 as an input
12:52:21 <JVassie> for example in my case, with weight 90 tons, and aiming for a TE of 300, the coefficient is 0.033333 etc
12:53:00 <planetmaker> factor 10 difference
12:53:49 <planetmaker> @calc 90*1000* 10 / 300000
12:55:10 <JVassie> now I have to sort out my .png not having a palette
12:55:31 <JVassie> there an easy way to add one in GIMP?
12:56:19 <glx> adding the palette is easy, but usually your colors will be wrong :)
12:57:53 <Ammler> there are ttd palettes for gimp on the devzone
12:58:44 <Terkhen> the scintilla lexers seem to be compiled with the program, so not really an option :)
12:58:46 <JVassie> theres links to them in the NML doc
12:59:03 <JVassie> I presume the ttd palettes are indexed palettes
12:59:20 *** Chris_Booth_ has joined #openttd
12:59:47 * JVassie cant fidn the button to even open the dialog..
13:01:15 <JVassie> *sigh* this is painful :x
13:02:17 *** Chris_Booth_ is now known as Chris_Booth
13:02:48 <planetmaker> JVassie: window -> dockable dialoges -> colour palette
13:03:09 *** Vikthor has joined #openttd
13:03:11 <JVassie> Colourmap or Palettes?
13:04:18 <JVassie> now to import the .gpl?
13:04:58 <Eddi|zuHause> make sure you take a palette without magic colours
13:05:05 <Eddi|zuHause> or the conversion may do weird stuff
13:05:25 <planetmaker> well: different: image -> mode -> indexed
13:06:25 <planetmaker> use own palette -> (choose one) -> open palette selction menu. right click bottom of that window -> import palette
13:07:14 <JVassie> I have no import palette option :x
13:12:08 <JVassie> in the palette selection window I have the dos palette imported
13:12:25 <JVassie> but it doesnt come up in the list on the indexed conversion window
13:15:19 <JVassie> colours dotn look screwed
13:16:13 <JVassie> now NML think the palette doesnt contain 256 entries :p
13:16:23 <JVassie> i unchecked the 'dotn remove unsused colours'
13:16:43 <planetmaker> JVassie: use the ttdviewer to check for animated colours where they shouldn't be
13:16:56 <planetmaker> you find it also on the bundles server
13:17:06 <planetmaker> it's a nice sweet java app for that
13:18:06 <andythenorth> JVassie: sorry just noticed that :|
13:18:16 <andythenorth> 'wrong' is possibly subjective
13:18:28 <JVassie> which bit particularly?
13:18:35 <andythenorth> the roofs stand out
13:18:59 <JVassie> light comes from top right, no?
13:19:06 <andythenorth> bottom right, about 4.30pm
13:19:07 <JVassie> well, 4:30PM as I recall xD
13:19:31 * andythenorth is annoyed at self for not noticing earlier
13:19:37 <andythenorth> and is now on a mission to help other artists :P
13:19:47 <JVassie> how would you change the shading then?
13:19:54 <andythenorth> about 1 year ago I told Irwe something like "oh Simon Foster clearly made a mistake in this sprite" :P
13:20:06 <andythenorth> JVassie: darken the rear of the roof, lighten the front
13:20:26 <JVassie> read meaning top left of it, front meaning bottom right?
13:20:51 <andythenorth> it makes the correct shading quite clear
13:21:02 <andythenorth> it was very stupid of me to tell Irwe that Simon Foster had drawn it wrong :P
13:21:28 <welshdragon> can I load a scenario in a VPS?
13:21:38 <JVassie> it would look weird if the roof went from ligth to dark, then light to dark again when put next to another canopy
13:21:40 <welshdragon> (Dedicated Server)
13:21:41 <planetmaker> Yexo: the town index... in OpenTTD it's rather word-sized than byte-sized, right?
13:22:02 <Ammler> welshdragon: if in doubt, just rename the sencario to .sav
13:22:18 <planetmaker> then that needs fixing, too :-)
13:22:26 <Ammler> afaik, loading doesn't care about extensions anymore
13:22:41 <Yexo> as var 81 is not the high byte of var 80 in that case
13:23:08 <JVassie> andythenorth: wouldnt you say?
13:23:51 <Ammler> welshdragon: also try first then ask :-P
13:23:54 <JVassie> <JVassie> it would look weird if the roof went from ligth to dark, then light to dark again when put next to another canopy
13:24:13 <JVassie> unless i misunderstand you
13:25:44 <planetmaker> Yexo: you mean me wrt to town index about the 'not easy'?
13:26:15 <Yexo> it needs clarification that 80+x vars are not direct memory accesses in openttd
13:26:19 <planetmaker> well. That var simply returns the index. Which is a pool index which is 0 ... 64000
13:26:23 <andythenorth> JVassie: I was trying to retouch it for you, but it's taking a bit long :)
13:26:46 <planetmaker> what else does the newgrf author need?
13:27:05 <Yexo> nothing, I just doubt a newgrf author needs that at all
13:27:21 <JVassie> andythenorth: perhaps a simple outline in paint instead of a retouch? just so i can understand the concept clearly perhaps? :)
13:31:48 <JVassie> the 'top' of the canopies
13:32:16 <JVassie> not like my example then :p
13:32:41 <andythenorth> sorry distracted by failing web server :D
13:32:46 <JVassie> so in essence its flipped?
13:32:59 <andythenorth> it's a very annoying discovery :|
13:33:10 <andythenorth> I had to redraw a lot of FIRS :(
13:33:17 <JVassie> so many newgrf sets are based on light from top right
13:33:34 <andythenorth> it's a silly direction to light isometric sprites from though
13:33:47 <andythenorth> for the art to look good, at least one face needs to be fully lit
13:33:53 <andythenorth> otherwise everything is gloomy
13:34:20 <andythenorth> not suggesting you have to repaint it :o
13:34:44 <andythenorth> some have made the case that changing the lighting direction improves the art
13:35:42 <JVassie> for now I guess will keep them as is, and once the set is eventually in the coding stage
13:35:54 <JVassie> I can see how well it interfaces in different situations in ttd
13:35:58 <JVassie> and make the decision then
13:36:02 <JVassie> thanks for the heads up :)
13:36:02 <andythenorth> it's quite easy to retrofit the lighting ;)
13:36:09 <andythenorth> and mostly it's just roofs that need to change
13:36:11 <Ammler> wow, that firs shed looks like lighting is from west
13:36:32 <andythenorth> only if the default coal mine looks like it's lit from the west :P
13:36:49 <Ammler> andythenorth: you mean original :-P
13:37:04 <andythenorth> I think default is ambiguous
13:37:08 <andythenorth> depending on context :P
13:37:13 <andythenorth> I should say original TTD
13:37:22 <Ammler> original is ttd, default can be both
13:37:24 *** Adambean has joined #openttd
13:37:25 <peter1138> 3d render everything
13:38:11 <peter1138> (are the 32bpp crowd still insisting that everything needs to be modeled and rendered with a 3d package, and also 1000X larger than normal...?)
13:38:30 <Ammler> peter1138: there is no crowd
13:38:59 <JVassie> planetmaker: how exactly do I use ttdviewer to check for the animated colours?
13:39:18 <planetmaker> and load your file
13:39:24 <Ammler> play around with the options
13:39:39 <planetmaker> enable / disable the animation colours in the rhs menu
13:39:52 <planetmaker> it nothing changes: all is fine ;-)
13:39:56 <JVassie> have done, none changes
13:40:06 <JVassie> so saving that file and using it in NML should work then?
13:40:35 <planetmaker> well, you need to have it saved to view it in ttdviewer in the first place, right?
13:41:08 <JVassie> Image file BB27000.png: Palette is not recognised as a valid palette >.<
13:41:38 <planetmaker> hm... JVassie nml/docs has afaik two palettes
13:41:53 <JVassie> ill try importign the windows one
13:41:54 <Ammler> andythenorth: JVassieis from right up, I thought you said, it needs to be right down, but your shed is from left down
13:42:03 <planetmaker> JVassie: if you can use the DOS one, do that
13:42:10 <planetmaker> more colours for you there
13:42:17 <JVassie> the sprites are already drawn
13:42:28 <JVassie> in MS paint I believe he did them in
13:42:29 <planetmaker> it has nothing to do with what OS you like
13:42:31 <andythenorth> Ammler: it's not
13:42:37 <andythenorth> you can zoom in and pick RGB values
13:42:44 <planetmaker> and if you don't have a valid palette... then using windows palette is stupid
13:42:49 <andythenorth> the shed is lit from about 4.30pm
13:43:00 <JVassie> <JVassie> doesnt like DOS - i meant that NML doesnt like the file when its using the DOS palette
13:43:15 <planetmaker> maybe it's a wrong palette file. There have been some...
13:45:06 <JVassie> when I imported the palette
13:45:14 <JVassie> GIMP picked up it had 256 colours in 16 columns
13:45:29 <Yexo> that doesn't mean the palette is correct
13:45:46 <JVassie> but at one poitn I had an error saying palette didnt have 256 colours
13:46:15 <Ammler> andythenorth: with your lighting, around 90% of every newgrf is wrong
13:46:17 <planetmaker> yes... both can be errors. But NML checks for the palette to also have the right colours
13:46:23 <JVassie> liek for example now, ive just tried it with the windows palette
13:46:27 <JVassie> days there arent 256 entries
13:46:27 <andythenorth> Ammler: my lighting?
13:46:36 * andythenorth will do something else
13:46:37 <Ammler> well, the lighting you think is right :-)
13:46:43 <andythenorth> arguing about lighting is stupid
13:46:50 <andythenorth> the newgrfs are wrong
13:46:56 <Ammler> well, you know what I mean
13:47:03 <Ammler> yes, I said that, 90% of them are
13:47:03 <andythenorth> assuming they want to be compliant with original graphics, they're wrong
13:47:04 <peter1138> what lighting is correct?
13:47:19 <planetmaker> andythenorth's of course ;-) :-P
13:47:22 <andythenorth> if they want to invent their own art, then its's subjective and all bets are off
13:47:31 <andythenorth> therefore there is no argument or debate to be had :P
13:48:20 <Ammler> well, on the example above, the difference is quite much
13:48:27 <peter1138> ttrs3 has some odd lighting
13:49:15 <andythenorth> we should add lightning :P
13:49:26 <andythenorth> as advanced options
13:49:40 <JVassie> should add ability for png's to be used withotu a palette!
13:49:42 <peter1138> unless you just want to lighten everything
13:49:44 <andythenorth> 'lighting struck your station'
13:49:48 <Chris_Booth> lol that could crash planes
13:49:51 <andythenorth> 'lightening strike, your game turned white'
13:49:52 <peter1138> JVassie, it's called 32bpp
13:49:58 <Chris_Booth> or stops Erail trains
13:50:43 <JVassie> this is annoying me :(
13:50:50 <JVassie> all i want is to make a wee little newgrf!
13:52:07 <Yexo> as far as I know nml handles palettes in image files exactly the same as grfcodec
13:52:14 <Yexo> ie it has to be the correct palette
13:52:49 <JVassie> shame the windoes NML doesnt support PCX
13:53:05 <Ammler> JVassie: if you convert colors with gimp, you need to disable the option "remove unused colors" or so
13:53:23 <JVassie> Ammler: I did so when I converted to an indexed picture
13:53:42 <Ammler> but you need to use the "full" ttd palette
13:53:52 <Ammler> not just the one without action/company colors
13:54:06 <Ammler> that one is just for converting :-)
13:54:23 <JVassie> full emaning the oen with 256 colours?
13:54:27 <Ammler> before you export, you need to use the full palette
13:54:54 <JVassie> is what I used to import the DOS palette into GIMP
13:55:30 *** Chris_Booth has joined #openttd
13:55:53 <Ammler> he, thought you use the windows palette?
13:56:18 <JVassie> on sepereate 'fresh' versions of the file
13:56:24 <Yexo> can you upload your new png file?
13:56:32 <Yexo> than I'll take a look at what exactly is wrong with the palette
13:56:34 <JVassie> after beign indexed by DOS?
13:57:01 <Ammler> your original and your convert
13:58:37 *** DayDreamer has joined #openttd
14:00:23 <JVassie> all 3 look identical to my untrained eye
14:03:45 <Yexo> JVassie both the -dos and the -win image have a palette with 255 entries, not 256
14:14:37 <JVassie> so there was an error with the .gpl I imported?
14:15:05 <Terkhen> I have been using the dos and win palettes from openttdcoop a lot
14:15:24 <JVassie> for importing into GIMP?
14:15:25 <Eddi|zuHause> there are options like "remove unused palette colours", which need to be disabled on saving
14:15:37 <JVassie> Eddi|zuHause: have made sure thats unchecked
14:19:14 <JVassie> awesome ill try those instead
14:23:29 <JVassie> ah hah this one comes up with (256) in the palette window
14:24:16 <JVassie> NML didnt output any message that time
14:25:33 <JVassie> hmm, not appearing in the newgrf window though :x
14:25:35 <DorpsGek> Eddi|zuHause: 3.24339751003
14:26:24 * JVassie cant remember how Yexo said to fix that :p
14:26:34 <Terkhen> JVassie: it's probably something wrong on the grf block
14:31:04 <peter1138> did nml ever support setting the same property for multiple ids at once?
14:31:11 <JVassie> how do you set the version of (o)ttd(p) its compatible with?
14:31:49 <JVassie> like 3 * 108 08 07 "JC" 00 01 "British Station Set v0.0.4 "
14:32:12 <Yexo> JVassie: that sets the nfo version, it's done automatically by nml
14:32:57 <JVassie> because those ones worked :p
14:34:07 <JVassie> first one works, 2nd doesnt weirdly
14:35:05 <Yexo> JVassie: compiling a file with that nml code and it shows up fine
14:36:02 <Ammler> your browser seems to convert the hg file somehow
14:36:06 <Ammler> 20ca38855c09d94d439e056094246ae6 ttd-newgrf-dos.gpl
14:36:07 <Ammler> 20ca38855c09d94d439e056094246ae6 ttd-newgrf-dos-hg.gpl
14:37:00 <JVassie> hmm, in terms if alignment issues, if i want to move it up a pixel, I add or subtract one to the y-offset?
14:37:06 <JVassie> and vice versa for the x offset
14:37:33 <Ammler> JVassie: use the ingame grf debugger
14:37:56 <DorpsGek> Eddi|zuHause: 23.0497990699
14:38:13 <Ammler> you need to enable newgrf_develper config
14:38:28 <Terkhen> JVassie: in your openttd.cfg, set "newgrf_developer_tools = true"
14:38:30 <Ammler> then you get a GUI to check sprites ingame
14:39:03 <Terkhen> with the newgrf debug gui you can also check stuff related to a newgrf item such as variables, active callbacks and so on
14:39:23 <Ammler> start a game with your newgrf
14:39:28 <Eddi|zuHause> hm... how about NewGRF-currencies?
14:39:53 <Ammler> Eddi|zuHause: newgrf-locales :-)
14:40:09 <Eddi|zuHause> Ammler: entirely different issue :p
14:40:17 <Terkhen> JVassie: IIRC the sprite aligner was in the last menu
14:41:10 <Eddi|zuHause> the problem: a vehicle set may want to have historically correct pricing, but this depends on start year and inflation settings
14:41:11 <Ammler> Eddi|zuHause: well, currencies are locales
14:41:22 <Eddi|zuHause> Ammler: not in the game, they are completely separate
14:41:26 <JVassie> so i can use the arrow buttons
14:41:31 <JVassie> will that move the sprites ingame?
14:41:47 <Terkhen> yes, you can use that to get the right offsets
14:42:32 <Ammler> but it shows the alignment just ingame, it doesn't write those to the grf :-P
14:44:27 <Eddi|zuHause> the problem (2): assume a vehicle set is modeled for 4% inflation, and a start year 1920, and sets correct prices for 1:1 conversion factor. then a game start in 1950 would divide all "historically correct" prices by 3.24, in order to counter that, the newgrf could set the conversion factor to 3.24 higher by an action 6 checking the start year
14:45:42 <JVassie> got 8 sets of offsets written down :)
14:49:08 <Ammler> why does a newgrf set calc with inflation at all?
14:50:08 <Ammler> that could be task of openttd
14:53:41 * JVassie makes mental note to do alignment with wagons in future straight away
14:53:46 <JVassie> rather than 'egine solo'
14:54:04 <JVassie> thanks Ammler / Yexo / Terkhen / planetmaker for all yuor help :D
14:58:10 <planetmaker> hm... there's a pre-made wiki syntax beautification script... :-)
15:04:13 <JVassie> * ********************************************
15:04:13 <JVassie> * Adjust y-offset in depots
15:04:13 <JVassie> * ********************************************
15:04:15 <JVassie> traininfo_y_offset = 2;
15:05:36 <planetmaker> looks like from ogfx+trains?
15:06:19 <JVassie> so that sets all sprites in the depot up 2 px?
15:06:50 <planetmaker> or rather down iirc
15:07:06 <JVassie> mmm would make more sense
15:07:24 <Yexo> negative values are allowed
15:07:35 <JVassie> mine float 1 or 2px when put next to ukrs2 coaches in depot for example
15:07:44 <JVassie> even though they line up ingame
15:09:26 <JVassie> I presume adding stations to NML involves a hefty amount of work?
15:13:05 <planetmaker> it's said to be not exactly straight forward
15:13:51 <JVassie> I wish i knew enough about the nfo involed to help out
15:14:18 <JVassie> what would you say is the most complex stations grf released so far? IRS?
15:16:27 <planetmaker> CHIPS is probably among them
15:16:41 <Yexo> ISR is more complex for sure
15:16:45 <planetmaker> ISR is not bad either, though
15:16:56 <planetmaker> much of ISR is single tiles
15:17:34 *** andythenorth has joined #openttd
15:18:05 <JVassie> just realised I have a lot of completed French Set sprites laying around on my pc
15:21:34 <planetmaker> a set without coder, eh?
15:21:53 <planetmaker> well, there's always enough to do for NewGRF coders
15:25:59 <Terkhen> JVassie: do you know what is the status of the spain set licensing?
15:26:15 <JVassie> not the foggiest mate
15:26:23 <JVassie> not been involved in the dev for years :(
15:26:25 <Terkhen> I'm guessing "none", but I have not read the hwole thread
15:27:03 <Terkhen> the sprites are nice, but the set has many bugs and the current coder does not seem to have much time to work on it
15:27:51 <JVassie> I should really choose a license for it
15:28:08 <JVassie> I guess it will be GPL
15:29:06 <JVassie> seeign as there are sprites posted on the website
15:29:12 <JVassie> I should probably make it clear asap
15:29:52 <Terkhen> clear licensing is always a good thing :)
15:31:54 <andythenorth> unrelated to that, it's fun when GPL is described as 'less restrictive'
15:32:03 <andythenorth> some might say it's more restrictive :)
15:33:48 <Terkhen> I said clear, not less restrictive :P
15:34:23 <Terkhen> it ensures that development can continue
15:34:56 <planetmaker> yes. IMHO having a GPL or CC-BY license is very important. Or work can be lost for the community
15:35:31 <planetmaker> I'd have been quite sad, if I'd not have been allowed to add parameters to TTRS3
15:35:46 <planetmaker> but they did the right thing [TM]. And so did ISR actually
15:36:07 <planetmaker> and JapanesSets... though there it's hard(er) to find sources
15:36:29 <planetmaker> though... it might be a similar development style like ECS
15:45:28 <peter1138> just like the old days
15:46:08 <planetmaker> men can perfectly well communicate with quite a reduced set of letters ;-)
15:49:29 <Eddi|zuHause> <JVassie> what would you say is the most complex stations grf released so far? IRS? <-- modern stations is quite overcomplicated
15:51:39 <Eddi|zuHause> (and buggy, and borderline broken)
15:51:52 <Terkhen> it does not sound like a great example :P
15:58:56 <planetmaker> Terkhen: you like GRFID over grfID?
15:59:27 <planetmaker> I somehow prefer the latter
16:00:05 <Terkhen> I actually prefer GrfID
16:00:13 <Terkhen> but GRFID seems to be more used
16:01:40 <planetmaker> well, I guess it doens't matter. I just find GRFID hides the GRF while grfID makes it slightly more visible. But might be me. So...
16:01:54 <Terkhen> go ahead with grfID, I don't really mind :P
16:02:18 <planetmaker> war about proper capitalzation!
16:02:30 <planetmaker> one spelling certainly is copyright protected
16:03:50 <Eddi|zuHause> i'm totally inconsistent wrt this as well :p
16:04:32 <orudge> I would say GRFID, probably.
16:05:22 <orudge> we could rename it to something new and snazzy
16:05:43 <orudge> "Update the GRF's Splomr to uniquely identify it"
16:06:37 <planetmaker> this is just nice... run regex on all wikipages is so simple :-)
16:06:51 <orudge> For what it's worth, the original specs, and the wiki, use "GRF ID"
16:07:04 <planetmaker> well. it's w/o space now
16:07:32 <Terkhen> I saw it without spaces in some places before writing that entry in TermConsistency
16:08:24 <planetmaker> hm, yes, there are places with space, too
16:08:41 *** frosch123 has joined #openttd
16:10:09 <planetmaker> changed the spaced versions also to GRFID
16:11:15 <Terkhen> frosch123: what do you think that should be the next step regarding town control?
16:11:48 <andythenorth> add cb 15E: provide ponies
16:12:00 <frosch123> Terkhen: maybe additional text?
16:12:11 <Terkhen> I'm thinking on callbacks, probably additional text as it is required by the other ones
16:12:42 <Terkhen> not exactly required... but these features should have a way of providing feedback
16:13:03 <planetmaker> callback: rename town: 1948: s/Chemnitz/Karl-Marx-Stadt/; 1990: s/Karl-Marx-Stadt/Chemnitz/
16:13:11 <planetmaker> (or whenever that was)
16:13:17 <Terkhen> ok, if we agree I'll start reading other similar callbacks :)
16:17:41 *** Prof_Frink has joined #openttd
16:22:36 <planetmaker> I always confuse files and documents ;-)
16:25:04 <Terkhen> hmm... generic callbacks need a feature number, so towns would need to reserve a feature number
16:25:21 * Terkhen wonders if any kind of action0 makes sense for towns
16:26:52 <JVassie> Introduction date, town life?
16:27:03 <JVassie> date of random destruction
16:27:13 <planetmaker> Terkhen: it could.
16:27:23 <planetmaker> like town growth cargo
16:27:30 <planetmaker> growth speed indeed
16:27:43 <planetmaker> and a potential town type
16:27:53 <Terkhen> for stuff like growth speed, it would also make sense to have a callback
16:27:53 <planetmaker> which house newgrfs could read
16:28:05 <Terkhen> and if we have a callback... why do we need an action0 property?
16:28:29 <JVassie> Hmm, planetmaker; what sort of block would I need for a livery override for the engine itself, im thinking in terms of year built
16:28:55 <planetmaker> JVassie: a livery override block is what you need
16:29:26 <planetmaker> but... livery override is NOT the graphics depending on build year
16:29:35 <planetmaker> that's a simple switch sequence
16:29:39 <JVassie> you need to combien with a switch block?
16:29:46 <planetmaker> livery override is ONLY when a wagon adopts its view to an engine
16:29:53 <frosch123> Terkhen: towns need a feature byte for long
16:30:05 <frosch123> nforenum has quite some troulble with stuff without a feature byte
16:30:33 <frosch123> (internally it fakes feature 8 to be 'town'; but that is no nice way)
16:30:42 <planetmaker> JVassie: check out how I handle the bulk wagon, coal cargo
16:30:51 <JVassie> which src? opengfx+ ?
16:30:52 <planetmaker> that has different graphics depending on build year
16:31:36 <Terkhen> towns could take number 0x13 (reserving 0x12 for airports)
16:32:01 <Terkhen> or maybe a number in a different, higher range if we want to make clear it is a fake feature
16:32:33 <Terkhen> unless it makes sense to have a real action0 for towns
16:33:08 <planetmaker> they also have a sort-of dummy feature byte
16:33:12 <JVassie> /* Steam / Diesel / Electric wagons */
16:33:12 <JVassie> disable_item(FEAT_TRAINS, 29, 53);
16:33:12 <JVassie> /* Monorail wagons */
16:33:12 <JVassie> disable_item(FEAT_TRAINS, 59, 83);
16:33:15 <JVassie> disable_item(FEAT_TRAINS, 91, 115);
16:33:30 <Terkhen> planetmaker: true, 0E
16:33:33 <frosch123> Terkhen: airports already have 0D
16:34:00 <planetmaker> JVassie: it's even cooler, when you use if (disable_trains == 1) { ... }
16:34:04 <planetmaker> with a named parameter
16:34:20 <Yexo> <Terkhen> towns could take number 0x13 (reserving 0x12 for airports) <- airports already ahve 0D
16:34:30 <planetmaker> it was not possible when I wrote that. But is now
16:35:24 <Yexo> that makes me think we should slowly head towards disallowing the "param[x]" syntax
16:35:25 <planetmaker> Terkhen: then reserve feature 12 now. Today there's a special offer: buy this feature and get a wiki entry for free
16:35:29 <Yexo> as there is no need for it anymore
16:35:45 <JVassie> ive found the spriteset/groups for coal wagon
16:35:49 <JVassie> just need to fidn the item
16:37:39 <planetmaker> hm, JVassie seems I put that thing in a template
16:37:54 <planetmaker> templates_wagon:27ff
16:38:07 <planetmaker> template_wagons.pnml
16:38:22 <planetmaker> should be understandable also there
16:39:02 <planetmaker> switch(FEAT_TRAINS, SELF, name ## _year_switch, build_year < year_for_modern) { \
16:39:04 <planetmaker> 1: name ## _early_group; \
16:39:05 <planetmaker> name ## _modern_group; \
16:39:30 <planetmaker> that's it basically
16:39:36 <Terkhen> for using templates you need to use a makefile system though
16:39:56 <JVassie> im happy not to bother with them and the .pnml files
16:40:43 <planetmaker> JVassie: pnml is also nml. Just with a few added short-cuts which the compiler fills in for me. Thus I need to type less. E.g. this template is used for each bulk cargo. But I only typed it once
16:41:03 <Yexo> and edit it once, should there be a bug in it
16:41:18 <Terkhen> if you don't want to bother with templates, grab the generated nml file for ogfx-trains and look for the code generated from the template
16:41:31 <Terkhen> but... what they said :P
16:57:57 *** planetmaker was kicked by DorpsGek (Wrong channel. Retry in #openttdcoop.)
16:58:07 *** planetmaker has joined #openttd
16:58:07 *** ChanServ sets mode: +o planetmaker
17:09:45 *** Vikthor has joined #openttd
17:18:58 *** |Jeroen| has joined #openttd
17:26:02 <frosch123> hmm, planetmaker: did you mark the wrong item in the todo list as "done" ?
17:26:16 <frosch123> i.e. page-to-be-created instead of varaction2-towns
17:38:11 <planetmaker> probably, frosch123
17:40:39 <planetmaker> for what it's worth: The wiki now uses consistently towns over cities, TTDPatch instead of TTDP, OpenTTD instead of OTTD, more often makes clear that some version infos refer to TTDPatch,
17:42:20 <planetmaker> oh, and Size is always linked in page tables ;-)
17:42:50 <frosch123> planetmaker: there are lot of cases, where internal links [[ | ]] have two || in the middle
17:42:57 <frosch123> thus breaking tables
17:43:31 <frosch123> (i also fixed lots of those in the pages i converted)
17:43:43 <frosch123> does your bot handle that?
17:44:04 <planetmaker> it did. But all pages should be treated - those which had not been touched up to last night
17:44:17 <planetmaker> thus everything *should* be readable now
17:44:38 <planetmaker> but some pages where the bot did not yet work, with the old script, might have that failed conversion you pointed out
17:44:54 <planetmaker> and... if pages had formula - they'll be broken most likely
17:45:23 <CIA-10> OpenTTD: translators * r22590 /trunk/src/lang/ (indonesian.txt welsh.txt):
17:45:23 <CIA-10> OpenTTD: -Update from WebTranslator v3.0:
17:45:23 <CIA-10> OpenTTD: indonesian - 1 changes by adjayanto
17:45:23 <CIA-10> OpenTTD: welsh - 19 changes by welshdragon
17:46:57 * andythenorth has to work out advanced sprite layout
17:47:03 <andythenorth> getting up at 5am not ideal for this :)
17:53:35 <planetmaker> you'll get that with NML :-P
17:54:00 * andythenorth is too impatient for that :D
17:58:10 <Lakie> You might want to ask frosch123 about those, andythenorth.
17:58:53 <Lakie> If you are refering to what I think you are, using buffers and such to control how the action2 layout is used.
18:04:55 <andythenorth> I have the spec in my browser
18:05:00 <andythenorth> I'm just a bit slow
18:16:39 <Terkhen> hmm... openttd is using generic callback 18, only for stations though
18:21:33 <Yexo> yes, that's the only useful thing from that callback
18:21:56 <andythenorth> but that's not essential?
18:21:59 <Terkhen> I'll document it as implemented for stations only then :)
18:22:19 <frosch123> andythenorth: no, it is not essential
18:22:51 * Terkhen is happy to have generic callback code already :P
18:23:18 <frosch123> it won't help you much with the texts :p
18:23:58 <frosch123> iirc it only chains through the grfs until it has a result; you would need to concat the texts
18:24:16 <frosch123> but, well, it already loads the texts
18:24:30 * andythenorth sees why procedures are used
18:24:57 <Terkhen> the generic callback code will need many codechanges anyways; right now it is only prepared to deal with a single callback
18:25:13 <Terkhen> or at least some parts of the generic callback code
18:26:34 <frosch123> btw. the industry window already concats the texts from multiple callbacks into a single string; so maybe that helps you
18:26:54 <frosch123> (cargo suffixes and such)
18:29:22 <Terkhen> thanks, I'll look into it :)
18:29:40 <planetmaker> hm... generic callback. What will it enable? The persistant storage access?
18:30:22 <frosch123> planetmaker: industries and object grfs displaying additional text in the town gui
18:31:27 <planetmaker> that sounds like a quite difficult change
18:31:59 <frosch123> i doubt that scares terkhen :)
18:32:25 <Terkhen> hmm... can town generic callbacks access the town's persistent storage?
18:32:57 <Terkhen> it is a bit scary, yes :P
18:33:13 <Terkhen> but I have been doing only small stuff for months
18:33:20 <frosch123> Terkhen: they have to, what else should they base their output on?
18:33:47 <Terkhen> does that mean that they should also have access all town variables?
18:33:57 <Terkhen> I'm a bit confused about how generic callbacks work, each one seems different :P
18:34:13 <frosch123> however, changes to persistant storages by gui callbacks are reverted after the callback for obvious desync reasons
18:34:35 <Yexo> access to the town persistent storage makes it sound like a town callback, not a generic callback
18:34:41 * Terkhen makes a note to test that once this is done
18:34:51 <Yexo> a generic callback should not reference to any object at all, or it wouldn't be a "generic" callback anymore
18:34:53 <frosch123> Yexo: a generic callback for the town feature, yes
18:35:05 <frosch123> "generic callback" only means, it is not assigned to any ID
18:35:20 <Yexo> if you see it that way, yes
18:35:22 <frosch123> as towns are no grf defined entities, they have no ID
18:35:47 <planetmaker> they have an index
18:35:54 <frosch123> the ai callback exists for all vehicle- and the station feature afaik
18:36:18 <frosch123> planetmaker: but a grf won't set properties for town 0 :p
18:36:28 <Yexo> yes, but the ai callback can't access any vehicle or station specific variables
18:36:31 <frosch123> there are no 'town classes'
18:37:07 <frosch123> Yexo: but only because the station/vehicle does not exist yet
18:38:06 *** Intexon has joined #openttd
18:40:46 <Terkhen> I wonder if town classes would make sense
18:41:32 <Terkhen> if callbacks are good enough, it should be possible to implement town classes just using callbacks and persistent storage
18:41:47 <Terkhen> so probably not a good idea (as action0)
18:42:10 *** Chruker has joined #openttd
18:42:32 <planetmaker> town classes... rather in the form of regions or so. But... all that can be done via houses or so
18:42:46 <Terkhen> but they could be done via callbacks too
18:43:25 <andythenorth> I can't compile advanced action 2 :)
18:43:37 <andythenorth> maybe I can suppress those warnings?
18:44:56 <andythenorth> can't disable fatal errors :(
18:45:59 <andythenorth> might actually be valid, but I have no way to be sure
18:46:01 <frosch123> Terkhen: nforenum not being ready for advanced sprite layout
18:47:27 <frosch123> the autocorrect feature of action2 gives me trouble
18:47:39 <frosch123> esp. as i do not see how it can actually fix something meaningful
18:51:51 *** supermop has joined #openttd
18:52:25 <CIA-10> OpenTTD: terkhen * r22591 /trunk/src/newgrf_callbacks.h: -Doc: Callback 0x18 (AI purchase selection) is implemented for stations.
18:54:00 <CIA-10> OpenTTD: frosch * r22592 /trunk/src/bmp.cpp: -Fix [FS#4645]: Reading of heightmaps with uncommon BMP formats failed due to uninitialised variables. (Parody)
19:02:32 <planetmaker> wrt variables 18 and extra_callback_info_1 and ..._2: what's the better naming for consistency? I prefer extra_callback_info - but where to put the info that it's vars 10 and 18 then?
19:03:03 <frosch123> well, they are not specific to callbacks
19:03:11 <frosch123> they are also used for normal sprites
19:03:24 <frosch123> e.g. station custom foundations or advanced sprite layout
19:03:36 <frosch123> so, how about "extra info 10" and "extra info 18" ?
19:05:05 <planetmaker> hm... but people need to know also that it's a variable. I'm talking about the newgrf wiki
19:06:51 <frosch123> imo "extra info 10/18" on the variationalaction2 page, and "variable 10/18" in the individual descriptions
19:06:55 <frosch123> everything else is too long imo
19:07:16 <frosch123> "extra info variable 10" or "extra callback info variable 10" :p
19:09:04 <planetmaker> hm... I like the extra callback info. But... I'm prejudiced by NML :-P
19:10:03 <frosch123> it would be nice if the name contains 10 / 18 instead of 1 / 2. the rest is not that important imo
19:10:38 <frosch123> i think most of the callback descriptions currently refer to variable 10 / 18
19:10:50 <frosch123> without any "extra", "callback" or "info"
19:11:08 <planetmaker> one or two mention it additionally as extra_callback_info 1/2
19:11:22 <frosch123> but putting "variable 10" in the description of the variables on the variationalction2 page is silly :p
19:25:11 <andythenorth> autocorrect feature?
19:25:21 * andythenorth has not really read nforenum docs
19:27:59 <frosch123> it is quite handy when decoding .grf into .nfo
19:28:17 <frosch123> it knows where to put escapes, strings, or plain hex in the nfo
19:28:26 <frosch123> grfcodec only dumps random junk compared to that
19:29:12 <Yexo> frosch123: are you talking about the autocorrect feature or the beautifier?
19:29:26 <frosch123> but it also tried to fix others stuff, like differing feature bytes in action 1 and 2
19:29:28 <andythenorth> I'll just hard-code snow ground tile into the graphics
19:29:37 <frosch123> but i do not know how useful that is
19:30:28 <frosch123> Yexo: i use the -a command line option to turn grfcodec qutoed text/hex mess into the right stuff
19:31:06 <frosch123> the description says you can specify -a multiple times to activate more stuff, but i never did that :)
19:31:41 <Yexo> frosch123: I only ever used -b
19:32:34 *** Chris_Booth is now known as Guest4728
19:32:36 <Terkhen> hmm... I wonder if we should do a post regarding town control specs before implementing further stuff
19:32:36 *** Chris_Booth has joined #openttd
19:34:50 <frosch123> Terkhen: first, we should have an idea what to do at all :)
19:36:13 <frosch123> other than that, i guess everyone who would comment constructively is already in this channel
19:36:54 <Terkhen> when would be the Cargo requirement information callback called?
19:37:00 <Terkhen> I don't understand that one
19:37:06 <frosch123> yes, the variables are questionable (esp. the city/town difference). but that cargo-requirement callback is questionable as well
19:37:13 <frosch123> esp. due to ai information
19:37:41 <frosch123> Terkhen: it's a gui callback
19:37:48 <frosch123> called by the window and by ais
19:38:00 <Terkhen> hmm... so it is called for each cargo with a town effect? or for all cargos?
19:38:01 <frosch123> the real decision is made by the growth callback
19:38:26 <frosch123> but the idea is, that cargo delivery does not necessarily affect town growth, but could also affect industries, houses or objects in the town
19:39:04 <frosch123> Terkhen: iirc for all cargos
19:39:51 <frosch123> but the information can become quite complex for display :p
19:39:57 <Terkhen> it will be difficult if not impossible to make available to AIs all of that info :)
19:40:26 <frosch123> that is the easy part
19:40:44 <Terkhen> because if town growth can access town variables then it can for example depend on the number of industries of type X around the town
19:40:48 <frosch123> just add an api function which takes a town and a cargo as paramater, which returns the requirements
19:41:17 <frosch123> oh, you mean general growth information
19:41:29 <frosch123> yes, that will be difficult like stockpile limits :)
19:42:06 <frosch123> a grf may as well decide to not make towns grow at all
19:43:09 <frosch123> but, i guess knowing the exact growth requirements is actually not that important
19:43:30 <frosch123> the ai only transports cargos anyway, so a hint what cargo might be useful, should be enough
19:43:41 <frosch123> or, should an ai start to fund industries to make the town grow?
19:44:06 *** Chris_Booth has joined #openttd
19:44:07 <Terkhen> not really, that would probably need specific support in the AI for that newgrf
19:44:31 <Terkhen> using towneffects for knowing what cargos have an effect should be enough
19:44:42 <Terkhen> then the callback could give a hint about each towneffect
19:44:49 *** Chris_Booth is now known as Chris_Booth[ph]
19:45:14 <frosch123> towneffect is not that useful
19:45:35 <frosch123> it basically has no meaning at all, it only controls various kinds of sideeffects
19:46:26 <Terkhen> hmm... maybe this needs a new cargo variable, that would be responsible for giving the town growth hint
19:46:43 *** Chris_Booth[ph] has quit IRC
19:46:49 *** Chris_Booth[ph] has joined #openttd
19:46:56 <Terkhen> but using a cargo variable creates a big mess between town growth and industry newgrfs :)
19:47:43 <Terkhen> so it is probably not a great option
19:47:46 *** Guest4728 is now known as Chris_Booth
19:48:17 <planetmaker> cargo delivered and transported, thus what's going on on stations would not interfer with industries directly
19:49:54 <frosch123> maybe it would also be a good idea to clean up the subsidy mess in ottd
19:50:15 <frosch123> and to detect what stuff is accepted by industries or houses
19:50:50 <frosch123> Terkhen: the town effect defines whether the subsidy is for town->town, industry->town or industry->industry transportation
19:51:22 <frosch123> depending on that it will search for an industry, which produces/accepts it
19:51:29 <JVassie> is it possible to change a station tile based on where abouts a train is within the platform?
19:51:34 <frosch123> or for a big town (independent whether it actually accepts it)
19:51:37 <JVassie> or on its 'ready to depart' state?
19:51:54 <frosch123> JVassie: the station knows the pbs reservation stage
19:52:11 <frosch123> so you can open doors when a train reserves a path into the station
19:52:13 <Terkhen> I start to understand the mess :)
19:52:32 <JVassie> frosch123: you know how stations now have a somewhat built in PBS signal?
19:52:46 <Terkhen> it would be better to scan industries and towns to know for sure what subsidies can be created instead of relying on that property
19:52:48 <frosch123> Terkhen: that mess is likely ottd specific. i guess ttdp just dropped subsidies support for new cargos
19:53:07 <Terkhen> yes, I checked towneffects on the specs, this behavior is openttd specific
19:54:02 <Terkhen> after fixing subsidies town effects could be deprecated in favour of something more flexible
19:54:37 <frosch123> JVassie: the station does not know where the train is exactly, whether it is slowing, loading/accelerating, speeding up, or just passing through
19:55:06 <frosch123> Terkhen: when the subsidy thing is removed, town effects could be used for growth again
19:55:18 <frosch123> e.g. take a look at the ttdpatch towngrowth switches
19:55:46 <frosch123> they can define cargo requirements for some default cargos
19:55:48 <JVassie> how about at the point when the train sets a PBS reservation leavign the station?
19:55:56 <JVassie> will the station know at that point?
19:56:20 <frosch123> JVassie: it only knows about the reservation of station tiles
19:56:37 <frosch123> enable the "show track reservation" in advanced settings, and see yourself
19:57:05 <JVassie> how does the built in signal work then out of itnerest?
19:59:49 *** Chris_Booth[ph] has quit IRC
20:00:30 <Terkhen> frosch123: then we would need a way to assign a "growth value" to each towneffect
20:02:52 <supermop> So I haven't really been paying attention the last few months
20:03:03 <frosch123> Terkhen: there already is
20:03:18 <frosch123> cargos have such a property
20:04:26 <Terkhen> but that is a multiplier for the cargo, not for the town effect itself
20:05:16 <frosch123> hmm, where would that be needed?
20:07:13 <JVassie> planetmaker: found the switch :p
20:07:16 <CoreUK> me and my friends are having trouble setting up a server
20:07:18 <JVassie> } switch(FEAT_TRAINS, SELF, bulk_wagon_coal_year_switch, build_year < 1970) {
20:07:18 <JVassie> 1: bulk_wagon_coal_early_group; bulk_wagon_coal_modern_group;
20:07:23 <CoreUK> are there some common problems ?
20:07:43 <Terkhen> frosch123: cargo multipliers for town growth would be like vehicle specific costs, multipliers for the town effect would be like basecosts
20:07:46 <CoreUK> we have port forward on 3979, ubuntu 11.04, openttd version: 1.1.1
20:08:10 <frosch123> Terkhen: but who would use them besides the growth callback?
20:08:34 <Terkhen> frosch123: oh, true, you can multiply them by whatever you want at the callback :P
20:08:55 <Terkhen> a global multiplier is not needed then
20:09:55 <Terkhen> hmmm... then the server is probably set correctly
20:10:28 <Chris_Booth> then you need to load a map and play
20:10:36 <glx> if you see it on the site it's started
20:10:43 <JVassie> i can see it on the server list
20:10:45 <CoreUK> other people cannot connect
20:11:01 <Chris_Booth> then its there issue not yours
20:11:09 <Chris_Booth> can they see other servers?
20:11:27 <Chris_Booth> you can give then the IP to your server
20:11:39 <Chris_Booth> 81.105.215.131:3979
20:11:46 <Chris_Booth> and they can manualy add you
20:11:51 <Terkhen> CoreUK: I can connect fine to that server
20:12:05 <planetmaker> JVassie: sure, it's no rocket science ;-)
20:12:19 <Terkhen> my guess is that it is not a server side issue
20:12:21 <CoreUK> possible connection loss no data received in x seconds
20:12:40 <JVassie> planetmaker: instead of the spritegroup i reference as the graphics set
20:12:45 <JVassie> i specificy the switch name, correct?
20:13:04 <andythenorth> it's more interesting here when people are doing stuff :P
20:13:19 * andythenorth states the obvious ;)
20:13:20 <JVassie> andythenorth: ever since I started being active!
20:13:22 <Chris_Booth> I am on your server now CoreUK
20:13:51 <Terkhen> frosch123: maybe the town growth callback should avoid using specific cargos and rely on town effects only
20:13:51 <CoreUK> Chris_Booth: ok thank you
20:14:16 <CoreUK> Chris_Booth, Terkhen, thank you I will ask they check out stuff at their end, regards
20:14:18 <JVassie> planetmaker: awesoem thanks
20:14:33 <andythenorth> cargos and towns
20:14:59 <Terkhen> meh, now I have to do something boring :P
20:15:14 <andythenorth> the cleaner interface is to rely on town effect flag on cargos
20:15:33 <andythenorth> it's reliable, but makes for quite a 1-dimensional gameplay effect
20:15:52 <andythenorth> so towns only see cargos with flag set, and all those cargos are same
20:16:02 <andythenorth> it's not very interesting
20:16:24 <frosch123> andythenorth: there are multiple town effects, not one
20:16:33 <JVassie> planetmaker: works like a dream <3
20:16:39 <Terkhen> andythenorth: the town growth callback will take care of that linearity :)
20:16:48 <Terkhen> once we know what it should do :P
20:16:51 <planetmaker> JVassie: you're building up expectations :-P
20:16:54 <JVassie> checking off the features here :p
20:17:06 <andythenorth> how are town sets decoupled from specific industry sets?
20:17:09 <planetmaker> next week we expect the first release :-P
20:17:22 <JVassie> im using graphics someoen drew back in 2005 :D
20:17:37 <Terkhen> andythenorth: town growth callback should IMO use only town effects
20:17:45 <planetmaker> JVassie: digital art doesn't rot ;-)
20:17:46 <JVassie> i invented a 2nd livery to test the override :)
20:17:49 <andythenorth> the existing town effects?
20:17:51 <Terkhen> then the specific cargos don't matter
20:17:53 <andythenorth> or some new scheme of lables?
20:17:56 <JVassie> by drawign a green stripe :D
20:18:16 <JVassie> next thing is livery refits
20:18:33 <JVassie> then ill move onto MUs :)
20:18:45 <Terkhen> andythenorth: in which regard would labels be better?
20:19:08 <andythenorth> how many current TE flags are there?
20:19:12 <Terkhen> hmm... a cargo can't have two town effects at once
20:19:24 <andythenorth> the existing system is a PITA when I tried to work with it
20:19:37 *** Chris_Booth_ has joined #openttd
20:19:40 * andythenorth will look at actual spec instead of guessing :P
20:20:20 <Terkhen> "how many" is not an issue, empty slots could be added
20:20:41 <andythenorth> but it would need to be defined what they do
20:20:44 <andythenorth> classes or labels or such
20:20:45 <Terkhen> heh, that's where labels start making sense, right? :P
20:21:03 <andythenorth> in the current spec, 05 is TTDP only isn't it?
20:21:08 <planetmaker> andythenorth: labels have also a backdraw... if not well prepared.
20:21:11 <frosch123> if you have to many labels, you tie the town growth to the industry grf
20:21:19 <frosch123> which is the only one providing those labels
20:21:27 <planetmaker> Like railtype labels happily defined in each grf another one
20:21:37 <andythenorth> I don't like tying town growth to industry grf at all
20:21:44 <andythenorth> but there's no clean interface I can think of
20:21:53 <planetmaker> town persistant storage
20:21:57 <andythenorth> cargos need to be defined by industry grf (is my assumption)
20:22:03 <planetmaker> that's already 65% there
20:22:19 *** Chris_Booth_ is now known as Chris_Booth
20:22:31 <Terkhen> planetmaker: that still needs that the town growth newgrf and the industry newgrf work along
20:22:41 <andythenorth> the assumption is that industry + town growth should *not* be coupled, yes / no?
20:22:53 <andythenorth> there's no case for saying they are 'economy'
20:22:56 <Terkhen> which is fine if someone wants to do it, but we should also have a way of working transparently
20:23:02 <andythenorth> and requiring to define all economy in one grf?
20:23:18 <Terkhen> andythenorth: they can be coupled if you want too; as tightly as you want
20:23:38 <planetmaker> Terkhen: special grfs slots: we could define one in a CTT type similar to the station delivery thing
20:23:41 <Terkhen> but IMO we should have a clean way of working without knowledge of what is the other newgrf doing
20:23:45 <andythenorth> well houses are quite decoupled from town growth nicely as I understand it
20:23:55 <andythenorth> so we don't need to think about house sets
20:23:59 <planetmaker> the 0xFEFFFFFF range allows to pre-define some of those
20:24:15 <Terkhen> andythenorth: everything with access to town persistent storage can modify town growth :P
20:24:25 <andythenorth> there will need to be conventions
20:24:25 <Terkhen> as long as the town growth grf is made that way
20:24:42 <andythenorth> otherwise you're just coding something that newgrf authors will turn into a cluster fuck :D
20:24:43 <Terkhen> hmm... not really, they can't write in the register of the town growth grf
20:24:54 <Terkhen> but the town growth grf could check the register of the house/industry newgrf
20:24:57 <planetmaker> mostly it needs a indeed well-defined and specific-grf independent interface
20:25:05 <planetmaker> at least that's desirable
20:25:20 <planetmaker> railtypes unfortunately don't have that with vehicles
20:25:24 <Terkhen> so the point is: why are town effects not enough in this regard?
20:25:37 <planetmaker> well... not as good as I'd like, but I've no better idea. anyway...
20:25:49 <andythenorth> this is more complicated when I try and consider YACD
20:26:27 <andythenorth> because a town may demand certain cargos for growth, but the industry grf doesn't locate destinations for them and/or YACD doesn't route any cargo there
20:26:29 <Terkhen> because YACD has no support for customizing its behaviour regarding newgrf cargos?
20:26:40 <andythenorth> it's a lot of stuff that can go wrong
20:26:52 <planetmaker> Terkhen: in principle I could want to make the growth callback depend the amount and type of cargo. Thus an enhanced town growth effect
20:26:53 <Terkhen> well, that's IMO an issue of YACD that can also happen with standard town growth
20:27:00 <andythenorth> so we ignore YACD
20:27:17 <planetmaker> but basically defining towns newgrf-able would IMHO superseed the cargo's property
20:27:20 <frosch123> andythenorth: houses might also want to access the cargo history
20:27:23 <Terkhen> andythenorth: not ignore it, but ignore known issues
20:27:24 <andythenorth> for both ECS and FIRS (and I think TAI), the industry grf has to take care of providing town cargo sink industries
20:27:42 <andythenorth> so it would be unwise for towns to demand specific cargos for growth
20:27:46 <Terkhen> planetmaker: so it would depend on specific cargos?
20:27:52 <frosch123> e.g. they might want to build different houses depending on whether the town deals a lot with valueables or rather with iron ore or steel
20:28:06 <Terkhen> oh, I never thought of that :O
20:28:07 <frosch123> and it makes no sense to tie those things to a town effect
20:28:15 <andythenorth> Towns would have no way of guaranteeing that a required cargo can be delivered :o
20:28:23 <planetmaker> Terkhen: the town could itself define the cargos it wants... but indeed... specific cargos is an issue
20:28:27 <andythenorth> because they can't control the building of specific houses or industries
20:28:37 <andythenorth> so no guarantee of a destination
20:28:45 <Terkhen> andythenorth: industry newgrfs have no way of guaranteeing that their cargos can be transported, because they have no control over vehicle newgrfs
20:28:49 <planetmaker> so... it has access to a town's CTT which only gives back anonymously the growth cargos?
20:29:01 <andythenorth> Terkhen: that just reintroduces labels :P
20:29:15 <Terkhen> planetmaker: that might be an option, but we need also a way that is completely independent of present cargos
20:29:19 <frosch123> so, imo there are some cases where a grf might be interested in a specific cargo, and not in classes
20:29:28 <frosch123> just because noone defined those classes in advance
20:29:35 *** welshdragon has left #openttd
20:29:40 <andythenorth> I wonder if the destination should write to a specific set of registers the town can use for growth
20:30:04 <andythenorth> how much storage is there?
20:30:22 <Terkhen> frosch123: true, it should be able to access cargos by label
20:30:22 <andythenorth> so how many slots? 65,356 or so?
20:30:24 <planetmaker> terkhen knows :-P
20:30:35 <planetmaker> andythenorth: pools are limited to exactly 64k
20:31:05 <Terkhen> planetmaker: right now OpenTTD can have up to 0xFF000 persistent storages
20:31:25 <planetmaker> he... that's even more
20:31:25 <andythenorth> I'm wondering if we could reserve 256 specific slots for town effects
20:31:34 <Terkhen> they are limited to 24 bits (as vehicles are)
20:31:41 * andythenorth is thinking out loud
20:31:42 <michi_cc> YACD could force a demand link for each growth cargo (if accepted), just like it is implemented for industries right now.
20:31:46 <Terkhen> most pools only use 16 bits
20:32:12 <andythenorth> for the cases I was envisaging
20:32:22 <andythenorth> it's not necessarily specific cargo delivered that should count
20:32:23 <michi_cc> Of course, some people will complain about too many destinations of those cargos only have very few producers.
20:32:33 <andythenorth> it might be a specific industry receiving some input cargo
20:32:44 <frosch123> so, maybe we can agree on town effect for growth, and specific cargo for other visual stuff?
20:32:49 <andythenorth> whether that cargo has town effect is up to the receiving industry
20:32:56 <planetmaker> nice wiki page, Terkhen :-)
20:33:31 * andythenorth feels like he is onto something but can't quite write it down
20:33:55 <planetmaker> andythenorth: town effect is a cargo property. Not industry
20:34:10 <andythenorth> consider electricity
20:35:09 <andythenorth> in PBI, wood delivered to a fuel depot might be town growth
20:35:18 <andythenorth> but not if delivered to a sawmill in a town radius
20:35:40 <andythenorth> coal to power station could be town growth
20:35:48 <andythenorth> but not coal to steel mill
20:36:04 *** welshdragon has joined #openttd
20:36:04 <andythenorth> town growth shouldn't be tied to specific cargos
20:36:32 <Terkhen> frosch123: do you mean using both?
20:36:58 * andythenorth needs to consider default case (no industry newgrf)
20:37:01 <Terkhen> I'm now convinced that specific cargos must be present
20:37:18 <Terkhen> but I'm not sure if town effects are enough for handling "generic" town growth
20:37:34 <frosch123> imo it is not only about town growth
20:37:44 <welshdragon> Gentlemen, I have a server on ubuntu, can I get openttd to look at a seperate config file?
20:37:45 <frosch123> there are various other reasons to check for a cargo
20:38:15 <frosch123> welshdragon: there is a command line option for that
20:38:35 <Terkhen> frosch123: for example?
20:39:14 <welshdragon> frosch123, is it a startup command?
20:39:15 <planetmaker> like the difference between manchaster and london
20:39:20 <frosch123> 22:27] <frosch123> house grfs might want to build different houses depending on whether the town deals a lot with valueables or rather with iron ore or steel
20:39:45 <frosch123> welshdragon: yes, run with option --help to see what is available
20:39:51 <Terkhen> yes, I was thinking only about the town growth callback
20:39:59 <planetmaker> mining town vs. commercial centre vs. farming "town"
20:40:10 <Terkhen> and that house grf would be awesome :P
20:40:20 <andythenorth> we definitely envisage standalone grfs for town control?
20:40:22 <frosch123> of course you could also try to attach that to industry types
20:40:25 <Terkhen> so... maybe a 60+ var with cargo label as parameter?
20:40:37 <Yexo> frosch123: so something like station var 69 but for towns?
20:40:38 <frosch123> like primary or secondary industries, but that is really up to the grf
20:40:38 <planetmaker> andythenorth: would IMHO make sense to some extend
20:40:52 <planetmaker> much better scenario control possible than mixing it with houses
20:40:53 <supermop> like those big silk drying towers in san giamino for a town that handles a lot of textile?
20:41:12 <planetmaker> meanwhile I don't fancy too much the close mixture of cargos and industries :-P
20:41:54 <welshdragon> oh, bugger, I can't start another dedicated server while running a dedicated
20:42:03 <frosch123> they were meant to somehow categorise the town without knowing about particular industries or whatever
20:42:35 <frosch123> sure, they could also be used for towngrowth
20:42:35 <andythenorth> it seems quite horrible to make the town have to know about n cargos from those already defined
20:42:45 <andythenorth> but to get those effects described, there's no alternative
20:43:13 <andythenorth> how many cargo labels do we already have?
20:43:14 <planetmaker> it's less horible as it sounds, I think
20:43:31 <andythenorth> I guess vehicle grfs deal with it
20:43:32 <planetmaker> frosch123: esp. 61 sounds exactly like the common interface
20:43:47 <frosch123> andythenorth: a grf can have 255 labels in the cargo trans table
20:43:48 <planetmaker> vehicle grfs are only on the receiving end, they're easy
20:44:21 <andythenorth> I would *not* want to be an author testing a town control gr
20:44:22 <welshdragon> can I tell openTTD to use port 3978 or something?
20:44:35 <Terkhen> welshdragon: edit your openttd.cfg and change the port
20:44:57 <andythenorth> is there not some way we could do it simpler for once?
20:45:16 <frosch123> andythenorth: if the author makes it too complicated, it is his own fault
20:45:19 <Terkhen> what would be simpler than towneffects?
20:45:27 <Terkhen> IMO they might be too simple :P
20:45:44 <Belugas> pick a microphone, plug it in the computer and say "Openttd, use port xxxx instead of yyyy"
20:46:22 <planetmaker> a sensitive enough accelerometer will do, too
20:46:22 <frosch123> Belugas: but you have to say it with a very clear dutch accent
20:47:37 *** Twerkhoven[L] has joined #openttd
20:47:38 <Terkhen> after this chat I'm convinced about vars 60-66
20:49:34 <planetmaker> he. During road reconstruction: no growth :-P
20:50:05 <planetmaker> hm... with growth control there, TAI could actually work...
20:50:14 <planetmaker> if growth callback could just fail then
20:51:53 <Terkhen> yes, pikka should check that document too
20:52:03 <Terkhen> not sure if he checked it before I started being interested on this
20:53:06 <welshdragon> Ok, can somebody (other than frosch123) tell me how to start a server which reads a config file in another directory?
20:53:33 <frosch123> Terkhen: the cargo history has a draw back though
20:54:01 <frosch123> it requires that the stuff is actually serviced, or you cannot detect the presence of secondary industries producing steel
20:54:25 <frosch123> so, maybe one would rather know potential production of cargos instead of actual production
20:54:28 <Terkhen> welshdragon: what else do you need to know?
20:54:52 <Terkhen> frosch123: hmm... can we know that?
20:55:12 <frosch123> you know the output cargos of industries
20:55:15 <andythenorth> hence industry writes to the town
20:55:28 <andythenorth> but that means industry sets have to know about town growth spec
20:55:28 <frosch123> you could e.g. count the number of industries producing goods or something like that
20:55:57 <frosch123> andythenorth: no, we do not want to require industry sets to supply additional data
20:55:59 <Terkhen> but we would not know the amounts
20:56:03 <frosch123> there is plenty available already
20:56:26 <frosch123> Terkhen: you only know the amounts, if they are actually serviced
20:56:42 <andythenorth> no electric pony for you then :P
20:58:11 <Terkhen> frosch123: hmm... so number of industries that produce/accept the cargo label passed by parameter?
20:58:57 <frosch123> maybe something like that.
20:59:12 <frosch123> however currently all variables expect a index to the cargo trans table
20:59:43 <Terkhen> hmmm... how does that work?
20:59:49 <Terkhen> only for the newgrf using it?
21:00:06 <Terkhen> would be the same for town growth newgrfs anyways, they would need their cargo table too
21:00:13 <frosch123> the cargo translation table of the grf accessing the variable
21:00:32 <frosch123> same as for the station cargo variable
21:02:02 <welshdragon> openttd@welshdragon:~/Chipp_v13_5$ openttd -D -c /home/openttd/Chipp_v13.5/openttd.cfg -d net=0
21:02:21 <welshdragon> will that find the config file specified?
21:03:12 <Terkhen> should this variable have a towneffect version too?
21:04:10 <frosch123> maybe for consistency, no idea
21:04:40 <Terkhen> makes sense for consistency, yes
21:05:47 <andythenorth> town growth gives me a headache :)
21:05:55 <welshdragon> am I being ignored? :(
21:07:37 <Yexo> welshdragon: you keep asking questions you can easily find the answer to yourself in a few minutes
21:08:47 <welshdragon> Yexo, I prefer to ask a person instead of reading
21:09:36 <Yexo> and I prefer to only spend time answering questions when the person asking them has already tried to find the answer himself
21:09:45 <Chris_Booth> welshdragon: what I think Yexo is trying to say is this isn't a help line to walk you through setting up a server
21:09:48 <Yexo> for your last question: yes, that should work
21:18:13 <frosch123> yeah, i guess there is no point in counting houses
21:18:34 <frosch123> esp. as the house cargos are not known properly
21:18:38 <glx> welshdragon: and if you move -D after -c ?
21:19:11 <Ammler> welshdragon: why don't you simply try?
21:19:21 <Terkhen> I'll start with subsidies anyways
21:20:56 *** welshdragon has left #openttd
21:23:34 <Terkhen> after that I guess that it would make sense to implement support for listing all cargos sharing the same town effect instead of showing a single one at the town gui
21:23:59 <andythenorth> town effect should be abolished :P
21:24:15 <Terkhen> what should be used instead of it?
21:24:19 <andythenorth> keeping it is going to messy
21:24:29 <andythenorth> town growth will be the preserve of the town grf
21:24:34 <Terkhen> and why is it going to be messy?
21:24:45 <andythenorth> keeping town effect implies industry sets can control town growth
21:25:03 <Terkhen> a town growth newgrf can ignore town effect if it chooses to
21:25:45 <andythenorth> I know it will work
21:25:59 <andythenorth> industry grfs will be setting properties that are then ignored
21:26:04 <andythenorth> it just makes everything harder to work with
21:26:45 <Terkhen> they don't have to be ignored, other town growth newgrfs could use them too
21:26:55 <Terkhen> it's a way of providing a hint to the town newgrf
21:27:19 <andythenorth> maybe I just got up too early :)
21:27:29 <Terkhen> honestly I don't think a cleaner way is possible; with communication you also get this kind of problems :)
21:27:41 <andythenorth> this seems like it's going to be heading the same way as inflation, and economy
21:27:55 <andythenorth> inflation is broken so sensible newgrfs ignore it
21:28:13 <andythenorth> trying to make an industry set work with economies is near impossible
21:28:21 <andythenorth> so industry grfs over-ride it
21:28:38 <frosch123> a industry newgrf should just not care about towngrowth
21:28:52 <andythenorth> in which case why leave a messy bit of spec in place?
21:29:32 <andythenorth> maybe I can make the case better another day :)
21:29:40 <andythenorth> don't let me delay the writing of code ;)
21:29:42 <Terkhen> because town growth newgrfs might want to assign different effects to different cargos, without having to rely on specific labels
21:29:56 <andythenorth> but then industry authors are going to complain
21:29:58 <Terkhen> andythenorth: don't worry, I have to fix a big mess before coding variables
21:30:27 <andythenorth> I only set town growth effect because I have to currently
21:30:37 <andythenorth> with town growth, I wouldn't set those flags
21:30:55 <andythenorth> there's no point, the effect is completely indeterministic
21:31:06 <Terkhen> maybe it is because I'm sleepy already, but I don't understand your point
21:31:23 <andythenorth> by setting the flag, the industry author is trying to control something that is outside their domain
21:31:30 <Terkhen> it is not more indeterministic than the rest of interactions between different newgrfs IMO
21:31:30 <frosch123> andythenorth: you should set the "food" effect if the cargo supplies houses
21:31:38 * andythenorth is also sleepy and may be explaining badly
21:31:42 <frosch123> you should not care whether it makes towns grow
21:32:04 <andythenorth> I don't want to care at all :)
21:32:12 <andythenorth> I want to not have to care about those flags
21:32:47 <andythenorth> anyways, it's not about me :P
21:32:56 <andythenorth> and I'm sure you understand the actual issues better than me
21:34:26 <Terkhen> hmm... I can't think of a way that does not need setting some property for cargos
21:34:43 <Terkhen> except relying only on cargo labels
21:35:14 <Terkhen> good night frosch and andythenorth
21:35:16 <planetmaker> and g'night andythenorth
21:35:22 <andythenorth> Terkhen: good luck solving it ;)
21:35:35 <Terkhen> I can't fix it if I don't understand the issue :)
21:35:38 <andythenorth> relying on labels *might* just be enough
21:35:42 <Terkhen> let's talk about this again other day :P
21:35:42 <andythenorth> but anyway - bed for me
21:35:45 *** andythenorth has left #openttd
21:38:41 *** andythenorth has joined #openttd
21:39:02 <andythenorth> Terkhen: it suddenly was clear to me, might not be able to explain it though :P
21:39:21 <andythenorth> town effect should be preserved, but disabled if a town-control grf is loaded
21:39:38 <andythenorth> that way it's very enforced in the spec about responsibilities of different authors
21:39:43 <andythenorth> but doesn't break default situation
21:40:03 <andythenorth> otherwise town grf and industry grf authors are going to have a horrible time
21:40:27 <andythenorth> I already have enough discussions about 'please change class on xyz to support my town / vehicle set'
21:40:48 <andythenorth> industry authors are going to bitch at town authors and vice versa
21:40:49 <Terkhen> buf if town effects are disabled, that means that a town grf should be able to replicate the default behavior without using town effects
21:41:10 <andythenorth> 'please set town effect growth on ECS cargo xyz so I don't have to bother checking a lot of cargos'
21:41:10 <Terkhen> how will it know what cargos should affect town growth as passengers do without specific support to industry newgrfs?
21:41:19 <andythenorth> check for label PASS
21:41:25 <andythenorth> it's not going anywhere :P
21:41:26 <Terkhen> that will fail for tourists
21:41:30 <andythenorth> so check the class
21:41:56 <andythenorth> those interfaces can't change that much, or all vehicle support breaks. As endlessly discussed :P
21:42:16 <andythenorth> maybe a bad idea, but it seems like some leap is needed
21:42:21 <andythenorth> we can't have all of everything
21:42:25 <andythenorth> something has to give
21:42:25 <Terkhen> hmmm... you can reliabily check for passengers and mail with classes, but I'm not so sure about stuff like water and food
21:42:42 <andythenorth> there are too many nodes in this network - the communication overhead is too high :P
21:42:44 <Terkhen> I'm not against deprecating town effects, but there must be some alternative
21:42:53 <andythenorth> towns have to explicitly check for cargos
21:43:16 <Terkhen> hmmm... it's more consistent, yes
21:43:17 <andythenorth> same as vehicle sets doing graphic support
21:43:20 <andythenorth> it's just a varaction 2
21:43:53 *** hgnmu128 has joined #openttd
21:43:54 <supermop> is town control happening?
21:44:01 <andythenorth> if the spec can be made sane
21:44:12 <andythenorth> and if Terkhen doesn't get bored of it ;)
21:44:18 <Terkhen> check the irc logs of the last 4 hours :P
21:44:22 <supermop> just started reading forums again after 2 onths
21:44:32 <andythenorth> Terkhen: sleep on it :P
21:44:34 <Terkhen> andythenorth: I'll think about it, yes
21:44:41 <supermop> and i am never payig attention on irc because i am at work
21:44:47 <andythenorth> it doesn't have to be my suggestion, but something has to go somewhere
21:45:04 <andythenorth> anyway, for the second time, good night ;)
21:45:05 *** andythenorth has left #openttd
21:45:27 <Terkhen> supermop: I'm half asleep so I can only give you the super abbreviated version :P
21:47:51 <Terkhen> I don't think there is anything about this at the forums, though... I'd prefer to have more clear specs before that
21:49:15 *** Brianetta has joined #openttd
22:02:12 *** CORE_UK has joined #openttd
22:03:02 <CORE_UK> hey guys. our group is wanting to set up some openttd servers anyone here with experience of this?
22:04:10 *** welshdragon has joined #openttd
22:04:51 *** hgnmu128_ has joined #openttd
22:10:34 *** hgnmu128_ is now known as hgnmu128
22:25:01 <hgnmu128> Is there a way to soft-link from TT Wiki to NewGRF-Specs Wiki?
22:27:20 <hgnmu128> Links that doesn't have a trailing arrow-like thingy, I guess.
22:28:09 <JVassie> you either use [ ] or [[ ]]
22:28:16 <JVassie> and [[ ]] is internal only
22:28:30 <JVassie> unless you can setup a class of [ ] which doesnt display the arrow
22:28:36 <Eddi|zuHause> how does wikipedia do this with the different language wikis?
22:28:55 <planetmaker> Eddi|zuHause: those are different wikis
22:29:22 <Eddi|zuHause> planetmaker: yes, exactly like the tt-wiki and the specs-wiki, or am i misinformed?
22:30:02 <hgnmu128> I meant the internal link. The domains may be a problem, won't they?
22:30:15 <planetmaker> and they'rehttp://meta.wikimedia.org/wiki/Pywikipediabot/interwiki.py <-- with something like that
22:30:24 <hgnmu128> Eddi: You're not misinformed, AFAICT
22:31:10 <hgnmu128> planetmaker: It is for different language projects, or so says the page.
22:32:19 <planetmaker> Eddi|zuHause: like that yes. Though that they may contain the same content in different languages
22:32:30 <planetmaker> thus the same things are shown in the language links
22:32:38 <Eddi|zuHause> planetmaker: he means something like: "is it possible to do [[newgrf-specs:Somewhere]]?"
22:33:38 <JVassie> planetmaker: any examples in NML of a) livery refits and b) random liveries ?
22:34:10 <planetmaker> iirc not. But... what's the difference to a normal link? I see no detriment
22:34:16 *** tokai|noir has joined #openttd
22:34:53 <planetmaker> JVassie: yes/no and yes
22:35:11 <JVassie> i presuem the latter is for a wagon?
22:35:21 <planetmaker> best look at the flatbed wagon, I think. There you can refit the wagon to several things
22:35:23 <Eddi|zuHause> planetmaker: imho same argument as against using absolute urls everywhere
22:36:18 <JVassie> does that include random livery too i presume?
22:37:04 <Eddi|zuHause> planetmaker: e.g. things like "what pages link here" could be calculated easier, cross-wiki
22:37:19 <planetmaker> JVassie: the flatbed wagons have randomly different cargo representations
22:38:14 <planetmaker> and a true livery override can be found for all MU trains, they all have each their won pax/mail/valuable wagons
22:38:24 <planetmaker> fitting to the engines
22:38:49 <planetmaker> anyway.... I was about to sleep ;-)
22:39:47 <hgnmu128> I guess Owen should've made a different namespace, or maybe just a directory like structure, instead of different domains, but anyway it's just my opinion of course.
22:41:00 <Yexo> the different domain was intentional, and imo a good choice
22:41:05 <Eddi|zuHause> btw.: what is this /wiki/ intermediate path for? it's just redundant
22:41:18 <Yexo> Eddi|zuHause: a default from mediawiki I believe
22:41:19 <Eddi|zuHause> iirc we had that discussion for openttd wiki before
22:41:22 <planetmaker> that's recommended setup for mediawikis, Eddi|zuHause
22:41:40 <planetmaker> bad things [TM] might happen if not
22:41:59 <planetmaker> quote from a not-bookmarked random mediawiki page
22:42:19 <Eddi|zuHause> no bad things happened on openttd wiki
22:44:47 <hgnmu128> OpenTTD wiki uses the redundant path, doesn't it?
22:44:54 <michi_cc> Eddi|zuHause: try to have a wiki page named robots.txt (well, not very relevant to OTTD, but very relevant to wikipedia, which is probably why /wiki/ is recommended)
22:46:03 <Eddi|zuHause> michi_cc: that's a handful of special cases that can be handled.
22:46:29 <hgnmu128> It redirects to the link sans /wiki, so openTTD wiki does have that structure internally.
22:51:12 <Eddi|zuHause> hgnmu128: it was even worse initially, it used /wiki/index.php/
22:53:08 <hgnmu128> I used to play OpenTTD since 0.5.x or something but didn't give the development or the Wiki a damn before 0.9.x or so. I feel sorry 'bout that. Missed loads of fun it seems.
22:53:28 <Eddi|zuHause> there was no 0.9.x ;)
22:54:12 <JVassie> would it be possible to draw a station tile that extends down over the foundations?
22:54:27 <Eddi|zuHause> JVassie: stations can define custom foundations
22:56:49 <JVassie> you just made Ameecher a happy man
22:57:07 <Yexo> IIRC one of mb's grfs already has such a feature
22:57:19 <Yexo> it shows some shops instead of brick foundations
22:57:20 <Eddi|zuHause> one of the unreleased ones
22:57:52 <JVassie> Ameecher said that too
22:57:58 <JVassie> add it to the feature list
23:01:42 <hgnmu128> "there was no 0.9.x ;)" - Maybe x is large enough to get a carry into the 9... xD
23:02:30 * JVassie trots off to ottd suggestions forum
23:03:50 <JVassie> were planning similar :p
23:04:06 * JVassie wants to request a more advanced stations gui though :p
23:04:07 <Eddi|zuHause> "Wie kommt man denn an solche Bahnhöfe ???" - "An diesen hier? Du fährst zu MBs Haus, und prü^Wredest so lange auf ihn ein, bis er endlich mal ein Release macht."
23:06:00 <planetmaker> and a true answer to the question
23:06:33 <JVassie> planetmaker: going to bed soon? ;p
23:06:49 <JVassie> a /whois reveals a german IP i believe
23:06:58 <hgnmu128> The answer and question, both are in that German, is it?
23:07:45 <hgnmu128> That's what they said they wanted in <to-write>. But going by the page name...
23:08:17 <planetmaker> that's tutorial. not specs
23:11:52 <hgnmu128> planetmaker: You mean "it should be tutorial and it is", or "it has wrongly been made tutorial, you idiot"?
23:12:54 <planetmaker> hgnmu128: it's a tutorial and IMHO does not belong to the newgrf-specs wiki. Thus it's fine in the general tt-wiki
23:13:33 <planetmaker> apropos hgnmu128 I rolled back your one edit as you copied the info from a page which we have no explicit right to copy from
23:13:42 <planetmaker> Or did you ask the person you quoted for permission?
23:13:50 <hgnmu128> Thanks. And you talking in sleep?
23:14:14 <hgnmu128> planetmaker: I thought your name was Mr. Borstel, not Mr.Blunck.
23:14:38 <planetmaker> except that you miss part of my name, what's your point?
23:15:20 <hgnmu128> Owen has copied it too.
23:15:30 <hgnmu128> And I'm sorry for missing part of your name. :(
23:16:55 *** Progman has joined #openttd
23:17:37 * hgnmu128 slaps himself for being such and idiot.
23:19:52 *** Intexon_ has joined #openttd
23:25:29 *** hgnmu128 has joined #openttd
23:26:46 <hgnmu128> planetmaker: There has been a misunderstanding. I couldn't understand you were talking about the 'VarAction2 Towns' page, until later.
23:32:35 <CORE_UK> hey guys. our group is wanting to set up some openttd servers anyone here with experience of this?
23:35:38 *** supermop has joined #openttd
23:37:00 *** ashledombos_ has joined #openttd
23:44:12 <JVassie> hmm how about this, aye?
continue to next day ⏵
