Log for #openttd on 14th March 2012:
Times are UTC Toggle Colours
00:02:31  *** cyph3r [] has joined #openttd
00:08:23  *** cypher [] has quit [Ping timeout: 480 seconds]
00:12:26  *** KritiK [] has quit [Quit: Leaving]
00:13:37  *** xiong [] has joined #openttd
00:14:47  *** Devroush [] has quit []
00:20:12  *** grzes [] has joined #openttd
00:21:52  *** DoubleYou [] has joined #openttd
00:22:06  *** grzes [] has quit []
00:34:44  *** xiong [] has quit [Remote host closed the connection]
00:52:07  *** xiong [] has joined #openttd
00:53:57  *** kkb110 [] has joined #openttd
00:54:02  *** xiong [] has quit []
00:54:28  *** xiong [] has joined #openttd
01:02:05  *** xiong [] has quit [Remote host closed the connection]
01:05:21  *** Biolunar [] has quit [Quit: All your IRC are belong to us]
01:15:13  *** pjpe [] has joined #openttd
01:35:03  *** cyph3r [] has quit [Quit: Miranda IM! Smaller, Faster, Easier.]
01:38:42  *** Pinkbeast [] has joined #openttd
01:58:49  *** Zuu [] has joined #openttd
02:01:13  *** Elukka [] has quit []
02:03:52  *** Phoenix_the_II [] has quit [Ping timeout: 480 seconds]
02:14:35  *** Zuu [] has quit [Read error: Connection reset by peer]
02:15:14  *** HerzogDeXtEr1 [] has quit [Read error: Connection reset by peer]
02:32:23  *** DoubleYou [] has quit [Ping timeout: 480 seconds]
02:33:21  *** DoubleYou [] has joined #openttd
02:34:08  *** DoubleYou [] has quit []
02:40:01  *** Pikka [] has quit [Quit: ChatZilla [Firefox 10.0.2/20120215223356]]
03:15:05  *** Markavian` [] has quit [Read error: Connection reset by peer]
03:25:21  *** glx [glx@2a01:e35:2f59:c7c0:5178:8126:b750:e291] has quit [Quit: bye]
03:55:28  *** xiong [] has joined #openttd
04:08:08  *** arakasii [] has joined #openttd
04:10:03  <arakasii> good morning. anyone have a moment to test my server connection. ive setup port forwarding but apparently it still shows offline when my brother tries to connect
04:10:30  <arakasii> wondering if its still something ive done or problem his end
04:14:24  *** arakasii [] has quit [Quit: ajax IRC Client]
05:05:09  *** Pikka [] has joined #openttd
05:45:06  *** tokai|noir [] has joined #openttd
05:45:07  *** mode/#openttd [+v tokai|noir] by ChanServ
05:45:07  *** tokai|mdlx [] has quit [Read error: Connection reset by peer]
05:54:15  *** Eddi|zuHause [] has quit []
05:54:35  *** Eddi|zuHause [] has joined #openttd
06:16:29  *** andythenorth [] has joined #openttd
06:20:00  <Pikka> andy!
06:20:06  <Pikka> it works! :D
06:21:06  *** tokai|mdlx [] has joined #openttd
06:24:46  *** tokai|noir [] has quit [Read error: Operation timed out]
06:25:24  *** Prof_Frink [] has quit [Ping timeout: 480 seconds]
06:33:11  *** JVassie [~James@] has joined #openttd
06:33:29  *** sla_ro|master [slaco@] has joined #openttd
06:36:56  *** Pikka [] has quit [Quit: ChatZilla [Firefox 10.0.2/20120215223356]]
06:56:13  *** Pikka [] has joined #openttd
06:57:10  *** jo2k [] has joined #openttd
07:02:20  *** jo2k [] has quit [Quit: jo2k]
07:09:28  *** kkb110 [] has quit [Ping timeout: 480 seconds]
07:16:30  *** andythenorth [] has quit [Quit: andythenorth]
07:20:55  *** Maarten_ [] has joined #openttd
07:21:34  *** JVassie [~James@] has quit [Ping timeout: 480 seconds]
07:25:14  *** Maarten [] has quit [Ping timeout: 480 seconds]
07:36:14  *** TWerkhoven[l] [] has quit [Quit: He who can look into the future, has a brighter future to look into]
07:36:48  *** andythenorth [] has joined #openttd
07:39:57  *** Phoenix_the_II [] has joined #openttd
07:40:06  *** Twofish [] has joined #openttd
07:40:06  *** Twofish [] has quit [Read error: Connection reset by peer]
07:47:52  <andythenorth> Pikka: shiny
07:48:02  <Pikka> :]
07:51:43  *** Pixa [~pixa@] has joined #openttd
07:52:02  *** George [~George@] has quit [Read error: Connection reset by peer]
07:55:33  <planetmaker> results look MUCH better, Pikka :-)
07:55:46  <planetmaker> as you're here, I can also ask here: how did you achieve the result?
07:56:48  <planetmaker> (did I ever mention that I like very much the transparent view of TAI for its great clarity?)
07:56:52  *** LordPixaII [] has quit [Ping timeout: 480 seconds]
07:57:13  *** TWerkhoven[l] [] has joined #openttd
07:57:20  <planetmaker> no newgrf gives a better overview to a tycoon and help where to build new connections
07:57:23  <Pikka> planetmaker: it builds a house which it then immediately removes
07:57:29  <planetmaker> ok :-)
07:58:02  <Pikka> I had my doubts that it would work, but it seems to
07:58:30  <SpComb> Chris Sawyer needs to do a kickstarter pledge for TTD2..
07:58:33  <SpComb> oh, no, wait..
07:58:44  <andythenorth> :)
07:58:47  <planetmaker> how do you unbuild it?
07:59:03  <planetmaker> technically speaking?
07:59:05  <Pikka> callback 21
08:00:05  <Pikka> it's not a /perfect/ solution, because there's a chance you could bump into one of the non-houses (which for TaI will look like rough ground sprites) while building, for the <256 ticks of the house's existence.  but it's unlikely to happen often in practice.
08:00:07  *** George [~George@] has joined #openttd
08:00:16  *** Eddi|zuHause [] has quit [Remote host closed the connection]
08:00:31  *** Eddi|zuHause [] has joined #openttd
08:02:00  <planetmaker> yeah. You could use a special ground sprite for it though which features some <whatever>... maybe some advertisement or so
08:02:23  <planetmaker> thus the kind of transitional things which appear in outskirts sometimes
08:02:34  *** TheMask96 [] has quit [Ping timeout: 480 seconds]
08:02:41  <Pikka> possibly, although I think drawing attention to them might just make things more confusing for players.
08:03:23  <planetmaker> maybe. 256 ticks is just short of 4 days...
08:03:24  <Pikka> I'll see how things go once I get it into the set and people can play with it.
08:03:31  <planetmaker> ^^
08:03:35  <Pikka> yeah
08:03:49  <Pikka> but that's the maximum period they'll live for, I think.  It can be shorter.
08:04:02  <planetmaker> that's how I understand it, yes
08:07:13  *** TheMask96 [] has joined #openttd
08:09:46  <planetmaker> Now all it needs is a prominent warning in the readme "Do not use this NewGRF with town growth scripts" so that we all can say "rtfm you dork!" :-P
08:10:00  <Pikka> lol
08:10:12  <Pikka> well, I don't know how much a town growth script would hurt
08:10:45  <Pikka> plenty of people use it at the same time as other house newgrfs, which totally defeats the growth limit thing. :P
08:10:49  <planetmaker> it would not hurt. But it'll be two things trying the same thing with different limits and conditions
08:11:09  <planetmaker> thus people will complain that the script doesn't work as intended
08:11:27  <Pikka> true
08:11:31  <planetmaker> reports you most likely won't get but the script writer
08:11:35  <Pikka> although is anyone ever going to make such a script? :)
08:11:47  <planetmaker> they do exist already. More than one. On bananas
08:12:10  <Pikka> hmm
08:12:12  <planetmaker> basically 75% of the game scripts are town growth scripts so far
08:12:20  <Pikka> speaking of bananas, it would be good if I could update my grfs :)
08:12:31  <planetmaker> did you try again?
08:12:40  <Pikka> yep, today
08:12:51  <planetmaker> with what error?
08:13:14  <Pikka> same as last time, unknown error uploading or something along those lines.
08:13:23  *** PhoenixII [] has joined #openttd
08:13:24  *** Phoenix_the_II [] has quit [Read error: Connection reset by peer]
08:14:13  <planetmaker> can you send me one of the grfs you tried? I don't have time to look now. I have a suspicion that it's something with the bananans DB and I might look later or ask Rb
08:14:58  <planetmaker> but a more exact error would be helpful ;-)
08:15:44  <Pikka>
08:16:39  <planetmaker> ok, thx. So that zip could go there in its entirety as posted there, yes?
08:16:44  <Pikka> yep
08:16:51  <Pikka> that is the zip I was trying to upload
08:17:11  <Pikka> btw, "required version" on bananas still only goes up to 1.1.3 :)
08:17:17  <planetmaker> true
08:17:22  <planetmaker> stupid ;-)
08:20:32  *** Scuddles [] has joined #openttd
08:20:55  <Pikka> no scrubblecars
08:21:01  <Scuddles> pikka is a butt
08:21:17  <Pikka> go and play your bad game
08:40:20  *** andythenorth [] has quit [Quit: andythenorth]
08:59:59  *** Rhamphoryncus [] has joined #openttd
09:07:24  *** welshdragon [] has joined #openttd
09:13:11  *** pjpe [] has quit [Quit: ajax IRC Client]
09:13:17  *** andythenorth [] has joined #openttd
09:34:32  *** Pikka [] has quit [Read error: Connection reset by peer]
09:34:52  *** Pikka [] has joined #openttd
09:35:03  *** DDR [] has quit [Quit: ChatZilla [Firefox 10.0.2/20120216100510]]
09:39:22  <dihedral> hello
09:47:15  *** tokai|noir [] has joined #openttd
09:47:18  *** mode/#openttd [+v tokai|noir] by ChanServ
09:48:41  *** Mazur [] has quit [Remote host closed the connection]
09:52:07  <Pikka> where's roadtypes, then?
09:52:29  *** tokai|mdlx [] has quit [Ping timeout: 480 seconds]
09:54:48  *** Pikka [] has quit [Quit: ChatZilla [Firefox 10.0.2/20120215223356]]
10:00:26  *** tokai|mdlx [] has joined #openttd
10:03:36  *** Mazur [] has joined #openttd
10:06:29  *** tokai|noir [] has quit [Ping timeout: 480 seconds]
10:07:24  *** glevans2 [] has joined #openttd
10:31:24  *** smoovi [] has joined #openttd
10:31:41  *** tokai|noir [] has joined #openttd
10:31:45  *** mode/#openttd [+v tokai|noir] by ChanServ
10:36:28  *** tokai|mdlx [] has quit [Ping timeout: 480 seconds]
10:41:03  *** cypher [] has joined #openttd
11:00:53  *** sla_ro|master [slaco@] has quit [Quit: DANGER is OFFLINE DANGER]
11:06:18  *** Mazur [] has quit [Remote host closed the connection]
11:15:38  *** Mazur [] has joined #openttd
11:16:17  *** KouDy1 [~KouDy@] has joined #openttd
11:22:07  *** KouDy [~KouDy@] has quit [Ping timeout: 480 seconds]
11:24:49  *** cypher [] has quit [Quit: Miranda IM! Smaller, Faster, Easier.]
11:37:16  *** tokai|mdlx [] has joined #openttd
11:43:04  *** tokai|noir [] has quit [Ping timeout: 480 seconds]
11:43:49  *** welshdragon [] has quit [Killed (NickServ (Too many failed password attempts.))]
11:44:18  *** welshdragon [] has joined #openttd
12:02:43  *** glevans2 [] has left #openttd []
12:38:19  *** HerzogDeXtEr [] has joined #openttd
12:51:10  *** evdvelde [~evdvelde@] has joined #openttd
12:54:10  *** Pikka [] has joined #openttd
12:57:50  <evdvelde> hi all, question of the day :)
12:58:09  <evdvelde> i seem unable to fund buildings or build statues, any clues?
12:58:25  <Mark> not enough cash?
12:58:44  <Pikka> I believe you have to have a high enough rating in the town and/or the town needs to be a certain size before you get those options.
12:59:55  <evdvelde> ah ok thanks Pikka
13:00:14  <evdvelde> Mark: i have enough and see other options i cannot pay atm, but thanks too :)
13:03:37  <andythenorth> Pikka: did you find roadtypes anywhere?
13:03:50  <Pikka> nope :[
13:04:00  <andythenorth> non moi aussi
13:04:24  <andythenorth> allemagne dix points
13:05:02  <Pikka> it's a little bit sad working on 19th century towns and industries, but not being able to do anything about those bitumen roads
13:05:19  <Pikka> atsa lotta points'
13:06:13  <Eddi|zuHause> by my experience, that almost never happens
13:06:18  <Eddi|zuHause> it's either 12 or no points
13:08:09  <andythenorth>
13:08:21  <andythenorth> Pikka: how's your c++ ? :P
13:08:29  <andythenorth> I have a roadtypes repo :|
13:08:36  <Pikka> poor to average :P
13:08:48  <andythenorth> better than mine
13:09:19  <andythenorth> it's not just the C++, it's knowing the game structure 'and now you need to do xyz'
13:09:23  <Eddi|zuHause> <andythenorth> I have a roadtypes repo :| <-- how many steps is that above "i have a thread in the suggestions forum"? :p
13:09:31  <andythenorth> Eddi|zuHause: 1.15
13:09:43  <andythenorth> I think I have a commit to flip map bits :P
13:10:25  <andythenorth> I could learn more, but then I become a slow (bad) game dev, instead of a reasonably quick (moderately adequate) newgrf dev
13:11:39  <Pikka> hmm
13:11:41  <andythenorth>
13:11:52  <Pikka> are "roadtypes" and "tramtypes" two separate things?
13:12:01  <andythenorth> technically?  or for the player?
13:12:08  <Pikka> technically
13:12:24  <Eddi|zuHause> Pikka: that was kinda undecided
13:12:24  <andythenorth> there are (iirc) three built-in roadtypes currently: road, tram, hway
13:12:50  <andythenorth> in a roadtypes spec, 'tram' is a subset of roadtypes
13:13:00  <andythenorth> it's a flag, and is drawn on top of any road surface
13:13:12  <Pikka> ok
13:13:20  <Eddi|zuHause> Pikka: there were two philosophies, one philosophy where a tile should have one road-like type, and one tram-like type, and one philosophy where they should be exchangable
13:13:20  <andythenorth> you only get two kinds of route per tile iirc
13:13:39  <Pikka> yep, tricky :)
13:14:15  <Eddi|zuHause> Pikka: and one (more or less strong) opinion that there should be 3 independent road types per tile (which raises a "not enough map bits" issue)
13:14:44  <andythenorth> fwiw, there was a lot of discussion, not entirely a lot of agreement, and the result I had was this:
13:15:16  <andythenorth> if you want 'road with tramway' that's a single type
13:16:38  <Pikka> hmm
13:16:52  <Pikka> that's probably actually a better way to do it
13:17:17  <Eddi|zuHause> andythenorth: but then you need crossings between "road with tram" and "road without tram"
13:17:31  <andythenorth> I didn't say it was good
13:17:36  <Eddi|zuHause> so essentially you get the same problems, but you wasted one roadtype
13:17:49  * andythenorth shrug :)
13:17:58  <andythenorth> the alternatives weren't better
13:18:04  <andythenorth> extend the map array?
13:19:49  <Eddi|zuHause> entirely unrelated:
13:20:30  <andythenorth> Pikka: v0.2.0 \o/
13:20:41  <andythenorth> cement tank?
13:20:41  <andythenorth> :)
13:21:04  <Pikka> haven't put in anything that wasn't already coded yet.  but sure, I can get that in...
13:21:23  <andythenorth> hmm
13:21:37  <andythenorth> what else did I do?
13:21:42  * andythenorth has no brain
13:21:47  <andythenorth>
13:21:49  <Pikka> if you dropped the second roadtype, are there enough bits for each quadrant of the tile to have a different roadtype? :)
13:28:19  <Pikka> if the number of possible roadtypes is reduced to 15, then we could get away with replacing the "4 bits for type" and "4 road bits" with 4*4 bits for the road type in each quadrant of the tile (or 0/F for none).  Or keep the "road bits" too and have 16 road types.
13:30:36  <Pikka> anyway, it's late and I should go to bed :)
13:34:17  <Rhamphoryncus> Track types are stored separately?
13:34:35  <Eddi|zuHause> Rhamphoryncus: there is only one tracktype per tile
13:35:25  <Eddi|zuHause> Pikka, andythenorth: the whole discussion might be meaningless, if michi_cc's map-layering stuff makes it to trunk
13:35:28  <Rhamphoryncus> Yeah, but's a related bit of data.  Might be some good options.
13:35:48  *** Pikka [] has quit [Quit: ChatZilla [Firefox 10.0.2/20120215223356]]
13:36:27  <Rhamphoryncus> What's that?  Google isn't telling me
13:36:30  <Eddi|zuHause> then each MP_ROAD "tile" would only have one roadtype, but you could layer multiple MP_ROAD "tiles" over each other
13:37:34  <Rhamphoryncus> Hum.  Does that create a new layer on the map for each type loaded?  Or is it for bridges/tunnels?
13:38:12  *** cypher [] has joined #openttd
13:38:38  <Eddi|zuHause> it allows things like this
13:39:11  <Rhamphoryncus> oooh
13:39:50  <Eddi|zuHause> potentially also custom bridgeheads and signals on bridges, but it's way deeper fundamentals
13:40:10  * Rhamphoryncus nods
13:41:57  <Rhamphoryncus> If that covers both road and rail I wonder if it could be used for different traffic patterns too, ie curved tracks
13:42:27  <Eddi|zuHause> nothing to do with that
13:43:49  *** glx [glx@2a01:e35:2f59:c7c0:8d11:84a6:49e8:4318] has joined #openttd
13:43:52  *** mode/#openttd [+v glx] by ChanServ
13:43:56  <Rhamphoryncus> aww
13:47:59  <michi_cc> Rhamphoryncus: Curved tracks isn't complicated, you only have to deal with a mild combinatorial explosion :)
13:48:15  <Rhamphoryncus> Yeah, that's why I'm asking :)
13:49:00  <Rhamphoryncus> It's nothing magical.  Just more of the same we already do
13:49:13  <Eddi|zuHause> and you have to draw them :p
13:49:27  <Rhamphoryncus> that too
13:49:42  <Eddi|zuHause> they don't even need map bits
13:49:51  <Rhamphoryncus> but on that note I'm volunteering to make a newgrf.  It'll be hideous, but I'll make it.  :)
13:49:55  <Eddi|zuHause> only some variables exposed to newgrf
13:50:03  <Rhamphoryncus> huh
13:51:26  <Rhamphoryncus> Hmm.  If I actually put together a set of graphics would that help much in getting the code changed?
13:52:38  <michi_cc> Yes.
13:52:59  <planetmaker> <andythenorth> there are (iirc) three built-in roadtypes currently: road, tram, hway <-- there's no highway anymore
13:53:57  <Rhamphoryncus> No hard proposal on curve radiuses that I should conform to/argue with?
13:54:07  <michi_cc> The biggest problem for curved tracks is to decide on a good schema how to provide the track graphics. E.g. should all tracks on a tile be drawn as one sprite or composited like they are now?
13:54:19  <planetmaker> some visible work generally helps getting a good discussion started :-)
13:54:42  <michi_cc> Or is it easier for NewGRFs if the adjacent track bits are realtive to what is drawn or absolute.
13:54:48  * Rhamphoryncus nods
13:55:05  <Rhamphoryncus> What about diagonal roads? :)
13:55:08  * planetmaker prefers the combinatorical approach. For no good reasons
13:55:20  <Eddi|zuHause> Rhamphoryncus: need movement pattern, entirely different issue.
13:55:20  <Rhamphoryncus> Or curved for that matter
13:55:31  <planetmaker> Rhamphoryncus, I have curves in my roads ;-)
13:55:38  <Eddi|zuHause> <-- design documents about the rail images needed
13:56:22  <michi_cc> planetmaker: Personally I don't think that would look good because for a switch the sleepers of the top-drawn track might hide a lot of the bottom track.
13:56:47  <planetmaker> that's the problem with that approach. Or can be
13:56:55  <Eddi|zuHause> basically, you need 6 sprites per rail bit, "straight", "left curve" and "right curve", each in an "a" (from corner of tile to center) and "b" (from center to other corner of tile) variant
13:57:36  <Eddi|zuHause> michi_cc: maybe sleepers and rail need to be split in two layers
13:57:46  <michi_cc> If you do whole tile at once you can avoide the overlap by doing fancy compositing on built time (where you can have more layers than just track and balast).
13:57:53  <Rhamphoryncus> I want multiple radiuses, so you can have both wide sweeping and tight
13:58:23  <Eddi|zuHause> Rhamphoryncus: not feasible on a per-tile basis
13:58:52  <michi_cc> Rhamphoryncus: You get track bits of adjacent tiles, nothing more.
13:59:10  <Rhamphoryncus> Then we shall see what I can come up with *g*
13:59:53  <Eddi|zuHause> Rhamphoryncus: important design constraint: in the center of the tile, all rail bits must be at the exact same position, to allow arbitrary combinations of "a" and "b" sections
14:00:33  <Rhamphoryncus> Yeah, alignment will be my first task
14:00:53  *** TGYoshi [~TGYoshi@] has joined #openttd
14:01:00  <Rhamphoryncus> I may not go with fully arbitrary.  We'll see.
14:01:38  <Eddi|zuHause> Rhamphoryncus: the combination would be done automatically. you only have to provide the sprites for the 6 half-bits
14:01:57  <Eddi|zuHause> for each of the 6 track directions
14:02:08  <Eddi|zuHause> so 36 sprites
14:03:47  *** tokai|noir [] has joined #openttd
14:03:51  *** mode/#openttd [+v tokai|noir] by ChanServ
14:04:14  <Rhamphoryncus> That sounds too easy and boring.  I'm going for insane, pie-in-the-sky :)
14:04:33  <Rhamphoryncus> I never said they'd be GOOD sprites, so I need to do something meaningful, heh
14:05:01  <Eddi|zuHause> "meaningful" is "fits in our concept"
14:05:08  <Eddi|zuHause> everything else is basically useless
14:05:38  <michi_cc> Eddi|zuHause: Automatically for you means by OTTD or by some build process? I wouldn't like to do it in OTTD because it would directly limit what's possible.
14:05:52  <Rhamphoryncus> I did ask if there was a hard proposal
14:05:53  <Eddi|zuHause> michi_cc: i meant the build process
14:06:16  <Eddi|zuHause> Rhamphoryncus: and i did tell you my proposal
14:06:29  <Rhamphoryncus> If any suggestions I make will be ignored unless they fit exactly what you want then there's no point in me working on it
14:06:55  <Eddi|zuHause> Rhamphoryncus: that's not what i said
14:07:31  <Rhamphoryncus> It sounds like your "curved" is just a graphical tweak over what we have now
14:07:39  <Eddi|zuHause> yes
14:07:53  <Rhamphoryncus> That's not what I'm interested in
14:08:15  <Eddi|zuHause> (imho) everything "bigger" will likely not be included, as it is a too large "step"
14:08:39  <michi_cc> Rhamphoryncus: The proposal by Eddi is based on what is possible to implement. "Real" arbitrary curved tracks means not only rewriting how rail is stored on the map but also rewriting all of the vehicle movement code.
14:08:42  *** Elukka [] has joined #openttd
14:08:47  <Eddi|zuHause> there is, however, another thing i have been proposing: "traffic objects"
14:08:47  <planetmaker> anything bigger has very quickly the issue of backward compatibility for train sets
14:09:19  <planetmaker> as it will look awkward without further viewing angles for vehicles
14:09:20  <Eddi|zuHause> which means it's like a NewObject (i.e. multi-tile), but has a state-machine (airport-like)
14:09:33  *** tokai|mdlx [] has quit [Ping timeout: 480 seconds]
14:09:47  <Eddi|zuHause> thus you may lead a vehicle on arbitrary paths through this object
14:10:16  <Eddi|zuHause> but that's rather a "far future" thing
14:10:44  <Eddi|zuHause> as nobody has figured out yet how to generalize airport-like state machines for (articulated) vehicles
14:10:44  <Rhamphoryncus> Well, wait and see if I can come up with a decent alignment, alright?  That's not too insane to store
14:13:02  *** Nat_aS [] has quit [Quit: Konversation terminated!]
14:14:48  <Rhamphoryncus> I'll code it
14:15:24  <Rhamphoryncus> And yes, not having enough rotation sprites for vehicles will be ugly, but IMO it's already ugly.  Nothing will be lost.
14:15:35  <Eddi|zuHause> *hÃŒstel*
14:16:30  <Rhamphoryncus> If you actually look at how trains go around corners they do this odd jumpy thing
14:18:19  <Eddi|zuHause> i meant:
14:19:43  *** kkb110 [] has joined #openttd
14:20:06  <Rhamphoryncus> ooh
14:20:19  <Rhamphoryncus> So some work has already been done.  Nice.
14:23:00  *** Biolunar [] has joined #openttd
14:25:09  <V453000> I think the way how cets does it looks great, but it is just so much more to draw than original :z
14:49:17  *** tokai|mdlx [] has joined #openttd
14:55:13  *** tokai|noir [] has quit [Ping timeout: 480 seconds]
14:58:17  *** tokai|noir [] has joined #openttd
14:58:21  *** mode/#openttd [+v tokai|noir] by ChanServ
14:59:43  *** andythenorth [] has quit [Quit: andythenorth]
15:04:13  *** tokai|mdlx [] has quit [Ping timeout: 480 seconds]
15:10:08  *** planetmaker [] has quit [Quit: ...und tschÌß!]
15:10:56  *** Ammler [] has quit [Ping timeout: 480 seconds]
15:11:55  *** SmatZ [] has quit [Ping timeout: 480 seconds]
15:12:40  *** Ammler [] has joined #openttd
15:13:04  *** Nat_aS [] has joined #openttd
15:13:28  *** planetmaker [] has joined #openttd
15:13:31  *** mode/#openttd [+o planetmaker] by ChanServ
15:15:29  *** _maddy [] has joined #openttd
15:15:59  *** SmatZ [] has joined #openttd
15:18:20  <NGC3982> how ..does that work without signals?
15:18:27  * NGC3982 i still new to the more advanced pbs.
15:21:11  *** cl8 [] has joined #openttd
15:22:29  *** tokai|mdlx [] has joined #openttd
15:28:07  *** tokai|noir [] has quit [Ping timeout: 480 seconds]
15:31:53  *** andythenorth [] has joined #openttd
15:38:02  *** _maddy [] has quit [Quit: leaving]
16:14:15  *** cypher [] has quit [Ping timeout: 480 seconds]
16:15:13  *** cypher [] has joined #openttd
16:23:40  *** Prof_Frink [] has joined #openttd
16:30:09  *** PhoenixII [] has quit [Ping timeout: 480 seconds]
16:30:11  *** smoovi [] has quit [Ping timeout: 480 seconds]
16:39:47  *** smoovi [] has joined #openttd
16:52:16  *** Doorslammer [] has joined #openttd
16:52:47  *** Doorslammer [] has quit []
16:52:54  *** Doorslammer [] has joined #openttd
16:54:48  * andythenorth ponders
16:55:19  <andythenorth> place a bet: disappearance of BROS forum....fuckup or flouncing drama?
16:55:23  <andythenorth> welshdragon?
16:55:27  <andythenorth> :)
16:55:59  <planetmaker> lol
16:56:37  <Doorslammer> Whatever it is, it's a bloody facepalm yet again
16:56:43  <planetmaker> and then they say that history doesn't repeat itself :-)
16:57:05  <Eddi|zuHause> who said that? :p
16:57:18  <planetmaker> I'm also sure one could make a good monty python sketch out of it meanwhile
16:57:52  <andythenorth> doorslammer...never work with brits.  disorganised, crappy infrastructure, prone to infighting :P
16:59:04  <Doorslammer> Tell me about it
17:01:13  <Eddi|zuHause> so what now, they spend 90% of the time with discussing their forum, 10% of the time drawing, and 0% coding
17:04:12  <andythenorth> no, they spend 50% of their time responding to me baiting them :P
17:04:16  <andythenorth> I blame the tories personally
17:04:36  <planetmaker> red herring is a nice thing to do ;-)
17:04:55  <Doorslammer> Although the baiting isn't exactly... helpful, this has been an enormous waste of time for me
17:05:02  <planetmaker> or not. Depends on view
17:05:29  <planetmaker> (sorry, was no reply to DS)
17:08:04  <andythenorth> Doorslammer: is your work now lost in some broken forum?
17:08:08  <andythenorth> :|
17:09:18  <andythenorth> if so that's very disheartening
17:10:08  <Doorslammer> I have it still, but it's more of the fact that I still don't know if it's a) up to date and b) any good (still plenty of colour errors in it)
17:11:07  <andythenorth> the devzone no help?
17:11:55  <Doorslammer> I never really followed it
17:12:18  <planetmaker> tt-forums and devzone are a bit more reliable than... seemingly the bros forum
17:13:07  <Doorslammer> What did Oberhumer or WWW use?
17:13:28  <planetmaker> oberhumer uses devzone
17:13:32  <planetmaker> www uses gibhub
17:17:02  <Doorslammer> Not that I even knew who we used half the time anyway
17:19:27  <planetmaker> "who we used" sounds.... like the real issue :-P
17:21:41  <Doorslammer> Well, I spend long periods of time on other projects and things, so something had to be neglected
17:23:51  <andythenorth> it was community managed, you never stood a chance ;)
17:24:36  <andythenorth> it's a synonym for no leadership, no direction
17:25:56  *** |Jeroen| [] has joined #openttd
17:26:58  <andythenorth> there's also this
17:27:48  <Doorslammer> Not authorised to access
17:28:10  <andythenorth> really :o
17:28:58  <andythenorth> oh yes
17:29:01  <andythenorth> how odd
17:29:24  <andythenorth> it's not public
17:29:41  <planetmaker> err, hm
17:30:10  <andythenorth> planetmaker: does devzone have a policy on private repos?
17:30:20  <andythenorth> repos / projects
17:30:22  <andythenorth> I'd vote for 'not allowed'
17:30:23  <planetmaker> yes: none
17:30:39  <andythenorth> Doorslammer: try again
17:30:44  *** Progman [] has joined #openttd
17:30:45  <planetmaker> ^^
17:30:52  <andythenorth> there's nothing there anyway, no repo
17:30:58  <andythenorth> only one issue
17:31:01  <andythenorth> it's probably dead
17:31:39  <planetmaker> I think that might be the reason... not remove. But 'hide'. As it never got any real content
17:31:55  <Doorslammer> Didn't even know about that
17:33:37  *** HerzogDeXtEr1 [] has joined #openttd
17:35:53  *** sla_ro|master [slaco@] has joined #openttd
17:40:10  <Doorslammer> The thing is, the first time this happened was because there was no GPL, no public access and just as much progress
17:40:16  *** oskari89 [] has joined #openttd
17:40:20  *** HerzogDeXtEr [] has quit [Ping timeout: 480 seconds]
17:40:46  <Doorslammer> The difference now is the sprites are covered by GPL and that the project was more accessible
17:41:41  <Doorslammer> So... someone is in rather serious trouble for killing it
17:45:12  <andythenorth> assuming it's not just a billing failure
17:45:45  <Doorslammer> I was under the assumption it was a university site
17:45:55  <oskari89> Can someone code stations, non-track tiles?
17:46:07  <planetmaker> they can have issues at university computers, too, Doorslammer
17:46:26  <planetmaker> and... I could basically plug-in my own box here and do there whatever (as long as legal)
17:46:50  <oskari89> Finnish Stations Set needs one..
17:46:52  <planetmaker> oskari89, 'yes'. someone surely can
17:47:21  <planetmaker> do you code it and have a specific question about it?
17:47:55  <oskari89> No, i'm not into coding but i have tracking table with sprites and information :P
17:48:03  <oskari89>
17:48:17  *** Devedse [] has joined #openttd
17:48:26  <planetmaker> oskari89, post about it in the tt-forums
17:48:36  <planetmaker> in the newgrf development section or so
17:48:42  <oskari89> Posted already..
17:49:00  <oskari89> Here:
17:50:37  <oskari89> But so far, no-one hasn't replied about coding..
17:51:45  <oskari89> Signals and rails have been coded, but no stations :P
17:53:25  <planetmaker> those who code stations need to venture into nfo land. And there an island which is special from other features. Thus not many people roam there
17:54:55  <oskari89> I see..
17:56:39  <oskari89> Hmm, can looping sound be implemented on stations?
17:56:56  <planetmaker> looping as in repetitive?
17:57:07  <oskari89> Yes.
17:57:15  <oskari89> Non-stopping :P
17:57:15  <planetmaker> not directly. Sounds are called via callback on certain occasions
17:57:54  <oskari89> Ok.
17:59:09  <oskari89> It would be nice to have a stations taster, but so far very quiet on that...
18:04:10  <planetmaker> oskari89, the fastest way probably is: do it yourself
18:04:29  <planetmaker> with the ISR's code, there's plenty of example code present (provided you also release your set under the GPL)
18:05:33  *** evdvelde [~evdvelde@] has quit [Quit: Konversation terminated!]
18:05:50  *** |Jeroen| [] has quit [Quit: oO]
18:12:03  <oskari89> Hmm..
18:12:52  *** frosch123 [] has joined #openttd
18:35:36  *** FLHerne [] has joined #openttd
18:37:30  *** Doorslammer [] has quit [Quit: ajax IRC Client]
18:38:46  *** Chris_Booth[ph] [~chrisboot@] has joined #openttd
18:39:06  <oskari89> I've never coded anything, not even smallest things.
18:39:35  *** FLHerne [] has left #openttd []
18:39:42  <oskari89> Just done some modifications at my own set's name, nothing more.
18:40:11  <oskari89> NFO isn't just my thing.
18:41:41  *** TheMask96 [] has quit [Ping timeout: 480 seconds]
18:42:47  <oskari89> And the statement "Stations is the hardest thing to code as far as I know!" isn't very motivating.
18:43:24  *** valhallasw [~valhallas@] has joined #openttd
18:44:01  <CIA-1> OpenTTD: translators * r24028 /trunk/src/lang/ (hungarian.txt latvian.txt):
18:44:01  <CIA-1> OpenTTD: -Update from WebTranslator v3.0:
18:44:01  <CIA-1> OpenTTD: hungarian - 2 changes by Brumi
18:44:01  <CIA-1> OpenTTD: latvian - 11 changes by Parastais
18:45:21  <frosch123> yup, not motivating, but maybe challenging :p
18:45:40  *** TheMask96 [] has joined #openttd
18:46:02  <frosch123> V453000: did you saw that 73129 requested a user name change to 47407 ?
18:46:20  <frosch123> maybe also an idea for you :p
18:48:53  *** Chris_Booth[ph] [~chrisboot@] has quit [Quit: Colloquy for iPhone -]
18:49:53  <andythenorth> my two year old requests a patch: when drag-dynamiting land, explode animation shows on every tile, not just drag-box corners
18:51:47  <TWerkhoven[l]> checkerboard (-ish) could work too
18:52:19  *** Chris_Booth[ph] [] has joined #openttd
18:53:53  *** Devedse_ [] has joined #openttd
18:54:13  *** Devedse [] has quit [Read error: Connection reset by peer]
18:55:17  *** Chris_Booth[ph] [] has quit [Remote host closed the connection]
19:01:40  <planetmaker> andythenorth, did you notice that explosion graphics differ if you explode one tile or more than one?
19:02:59  <xiong> No no no! It's fine as is. No explosion graphics would be even better. We don't want *more*.
19:04:22  <xiong> If you're dynamiting something it's likely because you want to build something else there. Balls of fire do not help with this; they do not dissipate while paused and take too long to go away when the game is running.
19:06:15  *** JVassie [~James@] has joined #openttd
19:17:53  <Rhamphoryncus> If build-while-paused in enabled they should dissipate, just like construction costs and income do.  Patch it.  :P
19:20:57  <V453000> frosch123: :D
19:23:52  *** lmergen [] has joined #openttd
19:23:59  <Terkhen> hello
19:25:56  <planetmaker>  kyosuke1989 <-- oskari89 is that you? Then you could as well go initially for object tiles for those non-track station parts. It's easier as it can be done in NML
19:26:18  <oskari89> Yes that's me.
19:26:29  *** krinn [] has joined #openttd
19:26:43  <krinn> hi all
19:27:12  <oskari89> Is there some tutorial for nml object coding?
19:27:23  <krinn> i have a trouble with a vehicle that doesn't answer to AIEngine.IsArticulated and act like one, newgrf can also lie about articulated status ?
19:27:26  <oskari89> NML seems much more simpler than that NFO.
19:27:27  <frosch123> Rhamphoryncus: i think 1.2 even behaves likes that
19:27:49  <Yexo> <krinn> i have a trouble with a vehicle that doesn't answer to AIEngine.IsArticulated and act like one, newgrf can also lie about articulated status ? <- what do you mean exactly?
19:28:05  <Yexo> how does a vehicle "act like one"?
19:28:09  <Rhamphoryncus> Hrm.  I'm not seeing any fireballs
19:28:12  <krinn> the engine answer false, but i cannot use it with "classic" station
19:28:23  <Yexo> oskari89:
19:28:28  <Rhamphoryncus> Yeah, the fireballs show normally but aren't created when paused
19:28:28  <krinn> AIOrder always refuse then
19:28:31  <oskari89> Thx :)
19:28:33  <Yexo> krinn: talking about a bus/truck?
19:28:39  <planetmaker> oskari89, there might not be an exact tutorial for objects. But of course there's the ^
19:28:45  <krinn> Yexo, yes
19:29:10  <krinn> wait a sec i'll gave you the message if i try by hands to order it to a station (my ai cannot get that error)
19:29:29  <planetmaker> and for objects you can find something in ogfx+landscape... though that might be way over the top compared to what you need
19:30:00  <planetmaker> it works basically the same as industry tiles or airport tiles which you find in FIRS or ogfx+airports repos
19:30:04  <krinn> Yexo, translate from french to english it says "cannot add order, that vehicle cannot goes in that station"
19:30:11  <planetmaker> or ogfx+industries for that matter
19:30:18  <Yexo> krinn: which AIOrder function are you using?
19:30:31  <planetmaker> krinn, bus vs truck?
19:30:35  <Yexo> and what error message does your AI get?
19:30:52  <krinn> the classic one, but the AI cannot get any error, just it fail, the error message comes from me switching to play as my ai and trying to set the order with the GUI
19:30:56  <krinn> bus
19:31:02  <planetmaker> and the station is also bus?
19:31:02  <krinn> ikarus 180
19:31:07  <Yexo> an AI can always get an error
19:31:08  <krinn> from ikarus set
19:31:21  <krinn> Yexo, no, the ai get the failure, but ERR_NONE
19:31:40  <Yexo> are you using AppendOrder() ?
19:32:03  <krinn> 2s checking how i add them, but must be that
19:32:25  <krinn> if (!AIOrder.AppendOrder(veh, homedepot, AIOrder.AIOF_STOP_IN_DEPOT))
19:32:26  <krinn> 	{ DError("Vehicle refuse goto depot order",2,"cCarrier::VehicleSetDepotOrder"); }
19:32:35  <krinn> Derror display the message + the last error string
19:32:44  <krinn> and the result is boolean = false error=NONE
19:33:08  <Yexo> are you sure "homedepot" is correct?
19:33:09  <krinn> and that engine bypass my valuate(AIEngine.IsArticulated) keepvalue(0)
19:33:18  <krinn> not refusing depot order
19:33:22  *** KritiK [] has joined #openttd
19:33:26  <krinn> only station order
19:33:38  <krinn> and it's a bus station, not a truck one :)
19:33:50  <krinn> just a classic station, not one "on the road"
19:34:47  <Yexo> AIOrder.AIOF_STOP_IN_DEPOT <- that does not make sense for a "goto station" order
19:34:47  <krinn> that's why i'm lost, it really answer false to AIEngine.IsArticulated but looks like it is
19:35:12  <krinn> oh sorry Yexo this one a sample of how i set order
19:35:19  <krinn> didn't see first it was for a depot
19:35:32  <krinn> but this one work, setting a depot work, only station are refuse
19:35:39  <Yexo> and you can manually reproduce the problem? Without any AI I mean?
19:35:47  <planetmaker> I think you should paste your relevant code. And maybe your testgame
19:35:47  <krinn> yes
19:35:52  <planetmaker> this is getting quite confusing
19:36:06  <krinn> when i play as my AI i try by hands setttings order and it says "that vehcile cannot use that station"
19:36:13  <planetmaker> ah
19:36:27  <krinn> there's no really a problem with code, just the newgrf and NOAI api
19:36:50  <Yexo> "ikarus 180" looks like an articulated engine
19:37:01  <krinn> it looks like yes
19:37:09  <krinn> but my query to IsArticlated says false
19:37:22  <krinn> so it isn't remove from my engine list of possible bus to use
19:37:26  <krinn> and so the problem comes
19:37:55  <krinn> it really looks (the grphic) and act like articulated engine
19:38:04  <krinn> but AIEngine.IsArticalted says false
19:38:25  <Yexo> it's definitely articulated
19:38:42  <krinn> but it says false
19:39:03  <krinn> hence my question: can newgrf also lie to IsArticulated :)
19:39:11  *** Scuddles [] has quit []
19:39:17  <Yexo> no
19:39:28  <krinn> then i found a bug :)
19:39:31  <krinn> because this one lie
19:39:37  <planetmaker> in your code? :-P
19:39:51  <krinn> no, just create a simple ai if you wish
19:40:04  <krinn> you just have to show result of AIEngine.IsArticulated result you'll see
19:40:08  <Yexo> there is at least some bug in the openttd code
19:40:18  <Yexo> CountArticulatedParts will return 0 when no vehicle can be allocated
19:40:24  <Yexo> which is wrong since it's used in the AI API
19:41:20  <Yexo> though my guess would be the newgrf might be returning inconsistent information too
19:41:41  <krinn> if you wish test, i have in openttdcoop post an aivehicletest
19:41:48  <krinn> but i can't find the link :)
19:41:57  <krinn> the search engine there is a bit weak
19:42:13  <Yexo> ?
19:42:21  <Yexo> google works fine with
19:42:22  <krinn> oh yes
19:42:26  <krinn> that one
19:42:28  <krinn> run it
19:42:49  <krinn> switch to the AI player and edit the X label with Ikarus 180
19:43:18  <krinn> if you have the ikarus set, you'll see the answer for many AI query, and answer for AIEngine.IsArticulated will show a false
19:44:18  <krinn> oh wait i could paste there the result from the console if you wish
19:44:28  <Yexo> I've run it already
19:44:56  <krinn> it's a simple AI that just query the noai api and display result
19:46:21  <krinn> btw using 1.1.5, i think noai api 1.0
19:50:59  <krinn> (i'm sooooo skill at finding bugs...)
20:07:20  <Yexo> krinn: as long as the engine is not yet build, it's not articulated
20:07:34  <Yexo> but any vehicle build from that engine type will be articulated
20:07:49  <Yexo> it's not strictly a bug in the newgrf (it works according to the spec)
20:08:12  <Yexo> however there is some ambiguity there, we'll try something different in openttd
20:08:52  <krinn> so, we end with INVALID_ENGINE ?
20:09:07  <krinn> and i=1 while return i - 1; == 0
20:09:13  <Yexo> yep
20:09:44  <krinn> strange only that engine does that
20:09:46  <Yexo> the newgrf spec doesn't specify that a newgrf must implement the articulated callback for the "build menu chain" which is called when a vehicle doesn't exist
20:10:01  <krinn> we might have others too i never seen, but others looks to works
20:10:40  <krinn> newgrf bug so, not really newgrf, but the specs should enforce callback usage
20:11:01  <Yexo> but they don't, hence it's currently nog a bug in the newgrf
20:11:11  <krinn> but how can openttd return its capacity ?
20:11:14  <Yexo> and in general we try to avoid changes to the spec that break existing newgrfs
20:11:33  <krinn> i mean i saw openttd get the capacity for each articulated part : so engine is also not built, but capacity is count
20:11:42  <Yexo> the newgrf can fake the capacity for the first part
20:12:01  <Rhamphoryncus> I need a break from thinking about all these curves.. time for some minecraft :D
20:12:30  <krinn> Yexo, this also explain why newgrf can lie about the engine length
20:12:44  <Rhamphoryncus> Doh, server is down
20:12:48  *** [1]Mark [] has joined #openttd
20:12:50  *** Markavian [] has joined #openttd
20:12:50  *** Mark is now known as Guest6283
20:12:50  *** Markk is now known as Mark
20:13:31  <Rubidium> Rhamphoryncus: my (and OpenTTD's) server work perfectly okay
20:13:42  <Rhamphoryncus> Minecraft server
20:13:58  <Rhamphoryncus> And it's merely turned off :P
20:14:14  <Rhamphoryncus> (But openttd is MUCH more reliable than minecraft..)
20:14:30  *** DOUK [] has joined #openttd
20:14:34  <krinn> Yexo, how about returning non zero for if (!HasBit(EngInfo(engine_type)->callback_mask, CBM_VEHICLE_ARTIC_ENGINE)) return 0;
20:14:51  <Yexo> we've been discussing that :)
20:14:55  <krinn> this way, openttd must assume non callback = articulated
20:14:59  <krinn> oh :)
20:15:41  <krinn> and just writing it down here i realize why: classic engine newgrf makers certainly doesn't set the callback if they build a non articulated engine :(
20:18:35  *** mahmoud [] has quit [Ping timeout: 480 seconds]
20:18:46  *** Guest6283 [] has quit [Ping timeout: 480 seconds]
20:19:15  <krinn> i think CountArticulatedParts should be 1 base and not 0 base, so we could see if >1 articulated if 1 non articulated and if 0 missing the callback (no info)
20:19:45  <Yexo> missing callback means non-articulated
20:19:48  <Yexo> there is no way around that
20:20:10  <krinn> how about then forcing the engine creation?
20:20:19  <Yexo> the change that we would make: if callback flag is set: assume the engine is articulated
20:20:31  <krinn> by telling return CountArticulatedParts(engine_id, false) != 0;
20:20:58  <Yexo> technically a newgrf can set the articulated_engine flag but not return any articulated parts, so it wouldn't really be articulated
20:21:15  <Yexo> in this case the AI would think the engine was articulated while actually it's not (=the reverse of your current problem)
20:21:34  <Yexo> I don't think that situation would lead to problems, since anywhere you can use an articulated vehicle you can use a non-articulated one too
20:21:41  <krinn> the false shouldn't force openttd to build an engine to get a real answer ?
20:21:55  <krinn> i see a v = new Vehicle() if false
20:22:08  <krinn> for the bool purchase_window
20:24:15  <krinn> i love the : "so it doesn't matter how many articulated parts there are" :D
20:24:55  *** lmergen [] has quit [Ping timeout: 480 seconds]
20:25:02  <krinn> it looks like the bill gates "640k will be enough..." kind of citation :)
20:25:19  <Yexo> that is not the problem
20:25:28  <Yexo> and it's not that kind of citation, this one is correct
20:26:13  <krinn> looks like a dead end, it doesn't matter for openttd as no one wish see how many articulated parts are there in the gui, but that's different for noai
20:26:44  <Yexo> it's like if you want to know if you have enough money to buy a book, you go to the store and try to buy it. At home you first check if your wallet isn't empty. If it's empty, you say "I don't have to go to the store, it doesn't matter how cheap the book is, I don't have money anyway."
20:27:33  <krinn> clear now
20:28:01  <krinn> but if you cannot enforce articlated specs of newgrf, this is mostly unsolvable
20:30:43  <Hirundo> Currently a newgrf can make vehicles bought on Wednesdays articulated and those bought on all other weekdays not
20:31:31  <krinn> yes, but currently noai cannot know if it's wednesdays or not, and if the newgrf wish do that...
20:31:34  <krinn> totally blind
20:32:20  <Hirundo> NoAI can get the current game date and then determine if it's a (in-game) Wednesday
20:33:06  <Hirundo> but that newgrf may also decide to change its behaviour depending on a parameter, or the presence of another grf, or even a parameter of that other grf
20:33:18  <Hirundo> So indeed, for all practical purposes you're blind :-)
20:33:27  <krinn> :)
20:34:03  <krinn> we can still create a list of engine and blacklist any engine that doesn't accept the AIOrder, but any ai then must build and use its own engine table
20:34:04  <andythenorth> same issue for NoGo too btw
20:34:13  <krinn> NoGo is just an ai
20:34:21  <krinn> if i get it right :)
20:34:37  <planetmaker> not quite. It has different powers. But conceptually yes
20:34:39  <Yexo> there are differences in the API (some functions are only available for NoGo, some only for NoAI)
20:38:06  <krinn> you wish a filebug for it ? it looks a simple issue, with a real hard solve
20:38:23  <Yexo> <- seems simply enough
20:39:21  <krinn> hu?
20:39:34  <krinn> this one should return false if HasBit isn't set right ?
20:39:56  *** Chris_Booth[ph] [] has joined #openttd
20:39:57  <andythenorth> Yexo: enforce the buy menu cb in newgrf v9 :P
20:40:08  <krinn> so the AI will still get false and think the vehicle isn't articulated
20:40:32  <Yexo> krinn: IsArticulatedEngine returns true if the CBM_VEHICLE_ARTIC_ENGINE flag is set
20:40:47  <krinn> lol but the problem is that this flag isn't set
20:40:50  <Yexo> that flag is set for the Ikarus 180
20:40:55  <krinn> oh
20:41:05  <Eddi|zuHause> Yexo: but the artic callback could return "not articulated", even if the bit is set
20:41:09  <Yexo> if the flag isn't set the vehicle is definitely not articulated
20:41:21  <Yexo> Eddi|zuHause: yes, but for the NoAI API this is the lesser evil
20:41:34  <krinn> lol wonder how many newgrf are lying around non articulated with that flag set :)
20:41:47  <Yexo> returning "is articulated" when actually it "is not articulated" does not lead to as many problems as the other way around
20:42:21  <Chris_Booth[ph]> Yexo: if it's an AI issue is in not best to force
20:42:22  <Eddi|zuHause> krinn: it could just be something generic, like "i set all vehicles to have this callback, whether i use that is determined later"
20:42:24  <krinn> yep as my ai will just clear a non articulated vehicle, not a big issue to lost one
20:42:44  <Chris_Booth[ph]> Then they can only use drive through stops
20:43:02  <Eddi|zuHause> Yexo: why can't it just run the callback without being able to allocate the vehicles?
20:43:12  <Yexo> Eddi|zuHause: it does that
20:43:26  <Yexo> but in this particular case the newgrf doesn't have the callback chain in the main menu
20:44:01  <Eddi|zuHause> Yexo: so call it as the purchase menu?
20:44:14  <Eddi|zuHause> oh, i see what you mean
20:44:32  <Eddi|zuHause> it branches differently in the action3
20:45:02  <Eddi|zuHause> and there's probably no foolproof method
20:45:34  *** LordPixaII [~pixa@] has joined #openttd
20:45:36  <Yexo> calling the "normal" chain, but that expects vehicle variables to be there, so you'd need to really allocate the vehicles, only to delete them later
20:45:52  <Yexo> and even that can/will fail for your wednesday example
20:46:17  <Eddi|zuHause> it's certainly not "my" example...
20:47:01  <Yexo> oh, sorry, Hirundo gave it above
20:48:05  <CIA-1> OpenTTD: yexo * r24029 /trunk/src/ (3 files in 2 dirs): -Change: [NoAI] make AIEngine::IsArticulated return true if the articulated callback flag is set, don't try to run the callback.
20:49:27  <krinn> thank you guys for the solve
20:50:03  *** Pixa [~pixa@] has quit [Ping timeout: 480 seconds]
20:50:11  *** Zuu [] has joined #openttd
20:52:24  *** Chris_Booth [] has joined #openttd
20:53:34  <planetmaker> <andythenorth> Yexo: enforce the buy menu cb in newgrf v9 :P <-- I'd rather say: "no special purchase stuff except graphics"
20:53:53  <andythenorth> nah, that won't work :)
20:53:57  <planetmaker> why not?
20:54:09  <andythenorth> because purchase menu doesn't know consist capacity etc
20:54:17  <andythenorth> you have to tell it such things
20:54:31  <andythenorth> well I do anyway :P
20:54:34  <planetmaker> run the actual consist capacity CB for that. But no special branch
20:54:49  <planetmaker> though... yes, difficult
20:54:55  <Yexo> it doesn't know currently, but the code can be changed to build a vehicle (including all articulated parts), compute consist capacity/other information, then destroy the vehicle again
20:54:55  <planetmaker> bad idea :-)
20:55:12  <planetmaker> (mine)
20:55:22  <krinn> some newgrf authors could build buggy engine for AI if they could switch properties like that no ?
20:55:33  <andythenorth> it has to handle cargo subtypes and such
20:55:39  <andythenorth> you can see HEQS nfo code for examples
20:55:41  <planetmaker> krinn, why "could". They can
20:55:45  <andythenorth> or probably BANDIT nml
20:55:46  <krinn> i mean showing capacity of 1000 and then when brought setting it to 0...
20:55:55  <planetmaker> and they sometimes do
20:55:58  <krinn> planetmaker, i was thinking that
20:56:04  <andythenorth> also wagon power, weight etc has to be handled
20:56:43  <planetmaker> andythenorth, I know that all that needs handling. But when just summing over all articulated parts as if built, that's much more fail-proof
20:57:54  <andythenorth> shrug :)
20:58:04  <andythenorth> I really don't mind how it's implemented
20:58:20  <krinn> andythenorth, must be why we get bug like that next :)
20:58:31  <andythenorth> I'm not too bothered if the spec changes either, as long as changes are aware of why current approach is used
20:58:54  <planetmaker> why is the current approach used?
20:59:19  <planetmaker> only reason I can imagine is "I don't want to test-build a vehicle for the purchase list"
20:59:32  *** Chris_Booth[ph] [] has quit [Quit: Colloquy for iPhone -]
21:00:05  <krinn> Yexo, did you check ikarus 280 ?
21:00:15  <krinn> looks like gfx also show an articulated bus
21:00:24  <krinn> and engine return false for that one too
21:01:28  <Eddi|zuHause> Yexo: maybe generally building "test" vehicles may be useful to clean up the whole purchase-list callback mess
21:02:15  <andythenorth> planetmaker: probably hysterical raisins :P
21:03:00  <Eddi|zuHause> people have been complaining that the purchase list stats for complicated articulated chains are "wrong" for ages
21:03:16  <andythenorth> they should man up and fix it
21:03:27  <andythenorth> it only took me a day to find a good non-insane solution :P
21:03:44  <andythenorth> where's the problem? :|
21:05:09  <krinn> andythenorth you're newgrf maker no ?
21:05:23  <andythenorth> yup
21:06:42  <krinn> if you did like that author did, would you change to handle the callback ?
21:08:17  <krinn> just to know if i could take time to find who made that newgrf with ikarus to tell him about the issue
21:09:25  <planetmaker> andythenorth, IMHO it is indeed conceptually wrong that there's a separate chain where the NewGRF author decides what to return as display values for the properties. Which can differ totally from what you get when actually building it
21:09:42  <andythenorth> planetmaker: no argument
21:09:57  <planetmaker> as it both, is more work for the NewGRF author and even with the best of intentions and effort can fail
21:10:00  <andythenorth> I am having to provide quite a lot of brittle code (in HEQS) to make the buy menu match what is built
21:10:12  <andythenorth> in BANDIT, life is much easier, the build script sorts it out
21:10:25  <andythenorth> one tram in HEQS took an afternoon
21:15:39  <krinn> i think newgrf authors only design/code their newgrf specs base on openttd experience, they should be more aware of noai issue
21:16:03  <andythenorth> krinn: your point is well made
21:16:08  <andythenorth> no idea how to solve it though
21:16:21  <andythenorth> there aren't enough hours per day to deal with all the complexity
21:16:43  <andythenorth> and for hysterical raisins, bits of the API best messy
21:17:00  <krinn> as long as they are aware noai exist, this should remove many errors/tweaks needs
21:18:25  <andythenorth> what are the issues?
21:18:46  <krinn> until the fix hit the noai api any ai will fail on that bus
21:19:55  <krinn> and this mean handling all engine and a blacklist, blacklisting any "ikarus" in engine name... something not really nice or easy to do as first sight, like remove an articulated vehicle should do
21:20:54  <andythenorth> you should refuse to handle such grfs, and post errors
21:21:12  <andythenorth> noai authors have been a good source of bug reports / change requests for FIRS
21:21:17  <andythenorth> it makes for cleaner newgrf
21:21:18  <planetmaker> krinn, normally NewGRF authors should not need care about AI
21:21:25  <andythenorth> +1
21:21:34  <andythenorth> I don't / won't do anything special to support AI
21:21:39  <krinn> newgrf authors should care about openttd: and noai like newgrf are openttd
21:21:48  <planetmaker> AI should get all required info by means of the API
21:21:48  <andythenorth> but cleaning up mistakes is not special support
21:22:28  <planetmaker> andythenorth, support for AI can be part of writing NewGRFs. Like special information on categorizing stuff
21:22:34  <krinn> planetmaker, even the fix now exist this mean for me per example, the noai will never be able to query how many articulated parts a vehicle have
21:22:57  <krinn> i don't see any usage right now for that, just an example what this fix will imply
21:22:57  <planetmaker> krinn, that's also not important. Nor accessible to the human player
21:23:15  <krinn> not important for human isn't the same for an AI
21:23:34  <planetmaker> where does it matter? The only imporant info is arcticulated yes/no
21:23:53  <planetmaker> or rather: can use normal road stops
21:24:10  <krinn> right now yes, i already said i don't see the usage for that, just that because of that, i see the noai cannot get it
21:24:43  <frosch123> krinn: when we get vehicle construction sandboxes, ais and humans will be able to query everything :)
21:25:25  <krinn> just to cite you planetmaker >> and AI should get all required info by means of the API
21:25:28  <krinn> see my point ?
21:25:47  <krinn> even not useful i know, this cite cannot be true
21:25:53  <planetmaker> no. "required info" cannot be more info than available to a human player
21:26:12  <krinn> oh, and AI should act like human player base on 0 info
21:26:21  <frosch123> well, a ai will never be able to understand the readme
21:26:21  <krinn> sorry, human player at least have "visual" info
21:26:32  <krinn> frosch123, right
21:26:44  <planetmaker> krinn, the visual info will not tell you the amount of articulated parts reliably
21:26:47  <frosch123> e.g. it has no chance to detect whether a vehicle has constant running cost, or whether they depend on speed
21:26:54  *** FLHerne [] has joined #openttd
21:27:07  <frosch123> and there will never be an api function to query this from the purchase list
21:27:14  <planetmaker> I know of grfs which use invisbible articulated parts
21:27:35  <krinn> lol oh my planetmaker i wasn't aware of that
21:27:40  <krinn> :)
21:27:49  <frosch123> number of articulated parts? what is that good for
21:28:01  <frosch123> i think what krinn wants to know is the length of the vehicle
21:28:12  <planetmaker> that's a different thing entirely
21:28:13  <frosch123> there is no point in the number of articulated parts, is there?
21:28:20  <planetmaker> ^^ yup
21:28:22  <krinn> frosch123, yep, something as simple as that is a nightmare for an ai
21:28:36  <frosch123> krinn: it's also for humans :p
21:28:42  <krinn> like the wagon that is length 10 and when buy length 15 : nice you've build a train too long
21:29:00  <planetmaker> also for humans
21:29:28  <planetmaker> this stuff has a lot of historical bear traps built into it
21:29:34  *** Progman [] has quit [Remote host closed the connection]
21:29:35  <frosch123> the only solution we can offer for this is a construction sandbox
21:29:55  <frosch123> but you won't be able to cut 90% of grf features just to make it easy for ais and boring for humans :p
21:30:03  <planetmaker> :-)
21:30:15  <andythenorth> there is no practical way that HEQS could provide accurate AI information about the trams
21:30:36  <andythenorth> the API we have simply has no concept for that kind of stuff
21:30:36  <frosch123> andythenorth: construction sandboxes would also make it easier for newgrfs :p
21:30:45  <andythenorth> frosch123: so I see :)
21:30:50  <krinn> i don't want cut 90% newgrf features, but some simple thing could be made to ease thing
21:31:04  <planetmaker> andythenorth, for all it's worth: it suffices to know "carries cargo X". "has length Y" and "has top speed Z" and can be built from 1843 to 1932
21:31:23  <andythenorth> has length Y for refit A
21:31:41  <krinn> and just costs!
21:32:05  <andythenorth> solving this would require 'frameworking' the newgrf spec
21:32:12  <krinn> to see what vehicle is the cheapest to build to carry A, an ai must build and refit all vehicle to compare
21:32:28  <andythenorth> 'frameworking' =  provide a wide range of inflexible built-in functions
21:32:29  <planetmaker> krinn, so does the human player
21:32:37  <planetmaker> or read the readme (if any explaining it)
21:32:45  <andythenorth> krinn: do you also refit to all subtypes for a vehicle?
21:32:58  <frosch123> andythenorth: ais do not know about subtypes :p
21:33:05  <planetmaker> :-D
21:33:07  <krinn> the human player learn it 1 time, the ai learn it (if coded and it was a pain) every time its run
21:33:09  <andythenorth> they're screwed then ;) :D
21:33:35  <planetmaker> krinn, save that info
21:33:58  <krinn> i could, but we're limit by save take too long time, so i won't
21:34:02  <planetmaker> you've to learn it for each game then, but well
21:34:19  <planetmaker> memorizing all NewGRFs also takes too much time. So I don't :-P
21:34:23  <krinn> but the point was that the human save it in his mind, while the ai need to learn at each new party even if it save it in savegame
21:34:40  <planetmaker> you can build that logic into your AI's file
21:34:57  <planetmaker> can AIs query the (exact) NewGRFs?
21:35:04  <Yexo> no
21:35:07  <krinn> no :(
21:35:19  <krinn> same, AI cannot know what newgrf are in use
21:35:20  <planetmaker> hm. Ok. Plan foiled
21:36:32  <krinn> per example if i blacklist any ikarus name, i might blacklist a newgrf with ikarus bus X that have its callback properly set...
21:37:20  <planetmaker> The better way is in any case for the AI to query the game for needed properties of stuff
21:37:24  <planetmaker> way more flexible
21:37:39  <planetmaker> Even when it has to do it for every start
21:38:32  *** DDR [] has joined #openttd
21:38:33  <planetmaker> I'm not 100% sure whether it'd be a good idea to allow direct query
21:38:38  <planetmaker> It has pros. But also cons
21:39:04  <krinn> as long as noai could be cheat, ai author will get lost
21:40:11  <frosch123> planetmaker: you mean a blacklist of bad vehicle grfs for ais on bananas? :p
21:40:52  <krinn> frosch123 bad newgrf for some AI could be good newgrf for more advance ai
21:41:08  <planetmaker> frosch123, not really :-P
21:41:13  <rane> can i generate landscape settings with the guy for a server?
21:41:18  <rane> guy->gui
21:41:27  <Yexo> rane: yes
21:41:28  <rane> are they saved to a cfg?
21:41:31  <Yexo> yes
21:41:34  <Yexo> to openttd.cfg
21:41:39  <Yexo> see the readme where you can find it
21:42:54  <andythenorth>
21:44:52  <krinn> that taco bell answer would be blacklist any ikarus name :P
21:46:19  *** cypher [] has quit [Ping timeout: 480 seconds]
21:46:55  *** collinp [] has joined #openttd
21:56:22  <frosch123> night kids
21:56:25  *** frosch123 [] has quit [Remote host closed the connection]
21:57:33  <rane> landscape = tropic is awfully arid
21:59:31  <krinn> Yexo, that ikarus 180 also do glitches in openttd list of running vehicle :P
21:59:51  <Yexo> so report it to the newgrf author :)
22:00:04  <krinn> the picture looks like it is "eaten" by the vehicle number sometimes
22:01:29  <andythenorth> ha ha
22:01:29  <andythenorth>
22:01:35  * andythenorth codes in his free time, for fun
22:01:41  <andythenorth> good article
22:01:47  * andythenorth goes to bed in his free time, also
22:03:54  <SpComb> how terrible
22:04:02  <SpComb> that guy
22:06:45  <krinn> this guy is mad :)
22:07:21  <Terkhen> good night
22:07:29  *** valhallasw [~valhallas@] has quit [Quit: leaving]
22:07:52  <krinn> it's just logic to ask that to a dev, just like you would prefer a driver that use his car everydays vs one that just got his license to drive your kids at school
22:08:10  <andythenorth> he's slightly trolling
22:08:21  <andythenorth> read the rest of his posts, he's a lot fun
22:08:51  <andythenorth> I employee people who refuse to write code in their spare time
22:08:59  <andythenorth> there's nothing wrong with their code
22:09:19  <andythenorth> I've employed people who loved writing code in their spare time who were just awful
22:09:26  <krinn> there's nothing wrong too with someone who got his license, else he wouldn't have it right?
22:10:01  <krinn> it's just about logic, even logic isn't truth sometimes
22:10:15  <krinn> eheh i suppose alain prost got accident too :)
22:10:53  <Zuu> Well, young men are more probable to have a car accident.
22:11:35  <krinn> yep zuu, just like dev coding in their free time should have stronger knowledge: logic
22:12:13  <krinn> not always true, but you can't blame anyone from assuming that
22:13:01  <krinn> we have a story here (france) that alain prost was caught at 320km/h on one of our route (limit to 130km/h) with a ferrari
22:13:44  <krinn> and cops leave him gone away without ticket seeing it was alain prost (for the fame & the driving skill), who knows!
22:14:35  <krinn> but i admit it's a story :) our cops are bitches
22:15:08  <Yexo> andythenorth: I like his style of posting
22:15:16  <andythenorth> he's funny
22:15:24  <andythenorth> I lost my evening to reading his stuff :P
22:15:26  <Yexo> "Since it's now clear that reading comprehension and critical thinking are not strong suits of the Node.js programmer, I would suggest that all Noders reading this article read it aloud, slowly and loudly, like an American tourist trying to find a train station in Tokyo."
22:15:45  <krinn> i think it's sarcasm, just like the picture use with the cute rabbit eaten
22:16:28  <Yexo> sure, partly it is
22:19:31  *** Devedse_ [] has quit [Quit: Ik ga weg]
22:23:17  <krinn> Suppose you're a less-than-expert programmer, which Node seems to attract in droves for some reason.
22:23:28  <krinn> troll bait :)
22:28:13  <krinn> night all
22:28:17  *** krinn [] has quit [Quit: Quitte]
22:28:57  *** Chris_Booth [] has quit [Quit: ChatZilla [Firefox 11.0/20120310173008]]
22:35:51  <andythenorth> "Don't do image processing work with PIL unless you have proven that command-line ImageMagick won't do the job"
22:35:52  <andythenorth> :D
22:35:56  <andythenorth>
22:41:49  <Zuu> Are 11 stickies in the 32bit forum really needed?
22:41:56  <Yexo> no
22:41:58  <Zuu> Does the one with most stkies win?
22:42:03  <Zuu> :-)
22:42:34  <Yexo> jupix and pikka are moderators there
22:45:10  *** kkb110__ [] has joined #openttd
22:45:10  *** kkb110 [] has quit [Read error: Connection reset by peer]
22:56:15  *** andythenorth [] has quit [Quit: andythenorth]
22:59:54  *** TGYoshi [~TGYoshi@] has quit [Quit: Popidopidopido]
23:01:51  *** oskari89 [] has quit []
23:02:03  *** smoovi [] has quit [Quit: Verlassend]
23:03:52  *** FLHerne [] has left #openttd []
23:04:10  *** Pikka [] has joined #openttd
23:10:50  *** JVassie [~James@] has quit [Ping timeout: 480 seconds]
23:13:48  *** TWerkhoven[l] [] has quit [Quit: He who can look into the future, has a brighter future to look into]
23:15:25  *** George|2 [~George@] has joined #openttd
23:15:26  *** George is now known as Guest6302
23:15:26  *** George|2 is now known as George
23:20:36  *** cypher [] has joined #openttd
23:20:45  *** Guest6302 [~George@] has quit [Ping timeout: 480 seconds]
23:23:04  *** cl8 [] has quit [Quit: Leaving]
23:32:58  *** sla_ro|master [slaco@] has quit [Quit: DANGER is OFFLINE DANGER]
23:42:07  *** Pixa [~pixa@] has joined #openttd
23:47:10  *** LordPixaII [~pixa@] has quit [Ping timeout: 480 seconds]
23:53:49  <peter1138> Pikka, Pikka Pikka Pikka Pikka
23:54:16  <Pikka> peter1138, peter1138, peter1138, peter1139
23:54:39  <peter1138> do you fancy sorting out the 32bit sticky mess?
23:54:52  <Pikka> wokay
23:55:24  <peter1138> there's like eleventy million stickies
23:55:36  <Pikka> just a mo, I'll go look
23:56:01  <peter1138> pretty much all obsolete or just, er, crap, heh
23:56:48  <Pikka> okay, what needs to go?
23:56:59  <Pikka> or what needs to stay?
23:58:24  <peter1138> good question. i'd say all :p
23:58:27  <peter1138> damn
23:58:40  <peter1138> i suppose it should be jupix sorting it out, as he did the deed in the first place
23:59:18  <Pikka> mmhm :P

Powered by YARRSTE version: svn-trunk