Log for #openttdcoop.devzone on 13th November 2010:
Times are UTC Toggle Colours
00:08:49  *** ODM has quit IRC
00:44:55  *** KenjiE20 has quit IRC
00:48:06  *** frosch123 has quit IRC
01:31:25  <Brot6> AdmiralAI - Revision 75:00069000d893: Fix: r74 also for trainmanager (yexo) @
03:45:50  *** Lakie has quit IRC
07:39:14  *** andythenorth has joined #openttdcoop.devzone
07:49:46  *** andythenorth has quit IRC
07:58:45  *** ODM has joined #openttdcoop.devzone
08:12:12  <planetmaker> moin
08:34:58  *** andythenorth has joined #openttdcoop.devzone
08:41:18  *** andythenorth has quit IRC
08:55:56  *** andythenorth has joined #openttdcoop.devzone
10:19:38  * Terkhen wonders why the bulk trucks don't appear ingame anymore
10:19:53  <planetmaker> train or RV?
10:19:57  <Terkhen> RV
10:20:29  <Terkhen> I also tried recompiling an old revision, but they still don't appear
10:21:08  <planetmaker> which one should I watch out for?
10:22:10  <Terkhen> hmmm... the trams don't appear either
10:22:13  <Terkhen> only the flatbed trucks
10:22:36  <planetmaker> Bulk Truck Mk I + Mk II ?
10:23:06  <Terkhen> Bulk Truck MkI + MkII + Mk III + all trams are missing
10:23:22  <planetmaker> it's the refit probably
10:23:42  <planetmaker> I have the MkI + MkII in 1970 with FIRS 0.5.2
10:23:54  <Terkhen> hmmm... I see
10:23:59  <planetmaker> But they transport only vegies, scrap metal and lumber
10:24:10  <Terkhen> but they used to appear in vanilla temperate
10:26:59  <planetmaker> ah
10:27:09  <planetmaker> bitmask(CC_BULK) and so on
10:27:42  <planetmaker> that changed
10:27:47  <planetmaker> in NML
10:28:02  <planetmaker> all cargo class definitions need revision
10:28:10  <planetmaker> by making them a bitmask
10:28:32  <Terkhen> oh
10:28:42  <Terkhen> okay, thank you :)
10:28:52  <planetmaker> no problem :-)
10:29:43  <planetmaker> nasty thing as it didn't break the newgrf
10:33:41  *** andythenorth has quit IRC
10:44:31  *** andythenorth has joined #openttdcoop.devzone
10:49:35  <Brot6> OpenGFX+ Road Vehicles - Revision 25:cabd5b906450: Update: Some vehicle properties are now set us... (Terkhen) @
10:57:20  <Brot6> OpenGFX+ Road Vehicles - Revision 26:e6d6a8fece61: Fix: Allow the bulk truck and the flatbed truc... (Terkhen) @
11:05:22  <Brot6> Berries - Feature Request #1837 (New): Password only requestable by known people (Spike) @
11:06:30  * Rubidium waits for the bananas berry
11:07:10  <Rubidium> although that would be a tautology
11:07:25  <Brot6> FIRS Industry Replacement Set - Feature #1404 (Closed): Copy Sand staithe sprites and modify for ... (andythenorth) @
11:13:34  *** frosch123 has joined #openttdcoop.devzone
11:44:16  *** KenjiE20 has joined #openttdcoop.devzone
11:49:56  <Terkhen> <-- this code fails to change the capacity for cargos that are being used as the default for the truck in a given climate; for example, fruit in tropic or sugar in toyland will use the vehicle capacity instead of the one at the callback
11:50:15  <Terkhen> I have no clue about why this happens
11:50:49  <planetmaker> Terkhen: there's a nasty difference between cb36 and cb15(?)
11:50:51  <frosch123> Terkhen: because of the cb15/36 mess
11:50:53  <planetmaker> both change capacity
11:51:27  <frosch123>
11:51:42  <planetmaker> and if the primary cargo in a climate should differ, then it fails, if you don't use both CB.
11:52:02  <Terkhen> wow, that is indeed a nasty mess
11:52:31  <Terkhen> okay, I'll try using both callbacks then
11:52:52  <frosch123> note that they return different things
11:52:59  <Terkhen> :/
11:53:11  <frosch123> cb15 returns the plain capacity, cb36 returns the capacity before applying cargo multipliers
11:54:48  <frosch123> alternatively change ottd behaviour to always call cb15 (also in purchase list), but i have no idea what might break by that, and whether everything needed is available :)
11:55:57  <Terkhen> thank you, I'll see if I can fix this mess
12:08:16  <frosch123> <- there is also some old thought, but some plans got changed, and i lost track of what was already archieved and what needs still doing :)
12:08:51  <frosch123> but iirc i managed to fix the refit-to-zero-capacity-deadlock, and the inconsistencies between purchase list and ingame-refitting
12:12:16  <Terkhen> oh, behaviour changes even between different vehicle types?
12:12:38  <frosch123> some of the most nasty inconsisties are already fixed
12:12:56  <frosch123> like, in the old days you could end up with a different capacity when refitting back to original cargo :)
12:14:15  <frosch123> take a look at Engine::GetDisplayDefaultCapacity() and  GetVehicleCapacity() for differences between vehicle types
12:20:27  <Terkhen> my initial impression is that, ignoring the inconsistencies, CB36 is only required for setting different capacities for both cargos in aircrafts
12:21:16  <Terkhen> that's my principal problem with understanding these callbacks; I don't see the need for two different callbacks that do almost the same thing
12:22:02  <frosch123> hehe, i do not know either why cb36 capacity was every introducted
12:22:11  <frosch123> i think it was a mistake
12:22:32  <frosch123> however, cb36 result behaves like the properties, i.e. with cargo multipliers; while cb15 does not
12:22:57  <frosch123> so, calling always calling cb15 is a good start
12:27:12  <Terkhen> yes, that solution makes sense
12:28:21  <Terkhen> isn't that "capacity multiplier" property equivalent to weight?
12:30:20  <frosch123> maybe, but it is different for different vehicletypes
12:31:11  <frosch123> similiar, but not the same
12:31:58  <frosch123> maybe it's a cominbation of weight and volume or so
12:35:08  <frosch123> anyway, there are some flaws in the capacityadventure proposal. e.g. the proposal allowed to make vehicles refittable/not-refittable by returning capacity 0 in cb36. but that caused quite some trouble, so now it is tied to the simple property from action0
12:42:25  <Terkhen> capacity multipliers are currently only set for the standard cargos you mention on the txt?
12:43:34  <frosch123> yeah, for slot 0, 1, and 5 (which are passenger, mail and goods in most cases)
12:57:41  <Terkhen> IMO they are a bit inconsistent... I guess that we can't just scrap them because existing NewGRFs rely on them
12:59:40  <frosch123> removing them completely gives no advantage, you could use cb15 then
12:59:57  <frosch123> imo the potential of cb36 is that it can deal with unknown cargo types
13:00:15  <frosch123> but that would require that the capacity multiplier can be set by the newcargo-providing grf
13:00:48  <Terkhen> hmm... you are right
13:02:30  <Terkhen> cb15 would be used for known cargo types; if it fails, cb36 would be called and it would let the cargo "decide" using its capacity multiplier
13:04:05  <Terkhen> heh... and to think that I wanted to spend the first part of the day doing something simple :P
13:04:21  <frosch123> :p
14:10:28  *** andythenorth has quit IRC
15:18:09  <Terkhen> would it make sense to evaluate cb15 if we are not in the default configuration OR if the vehicle is set to use the first refittable cargo type?
15:20:07  <planetmaker> what?
15:20:27  <frosch123> Terkhen: i guess just evaluate always :)
15:22:01  <Terkhen> planetmaker: <--- this is what I mean
15:22:31  <Terkhen> let's see what happen if I just always evaluate it
15:23:30  <frosch123> i guess changes in purchase list are more likely to affect newgrfs in some way
15:32:31  *** Alberth has joined #openttdcoop.devzone
15:35:23  <Terkhen> <--- this works for me too, but I wonder how many sets it breaks
15:36:00  <frosch123> doesn't that break mail capacity?
15:36:05  <planetmaker> we'd need to test some...
15:36:20  <Terkhen> probably
15:36:46  <Terkhen> I'll start by coding an aircraft test grf
15:40:02  <planetmaker> hm, that only affects aircraft?
15:40:52  <Terkhen> it affects all vehicle types, but it probably breaks CB36 for passengers/mail in aircrafts
15:41:26  <frosch123> doesn't it even set mail capacity to 0 once cb15 is enabled?
15:42:30  <Terkhen> hmmm... the BAe 146 in aviators aircraft seems to use callbacks to modify both passengers and mail capacities, and it works in the patched version
15:43:23  <Terkhen> frosch123: the mail capacity is set to zero at the beginning of the functions already
15:44:00  <frosch123> yes, that's my point :s
15:44:16  *** Lakie has joined #openttdcoop.devzone
15:44:35  <frosch123> so if cb15 returns early, it is never set to something non-zero
15:45:02  *** thgergo has joined #openttdcoop.devzone
15:45:44  <Terkhen> that it should be breaking mail? I agree... probably aviators aircraft is only using cb36 and that's why it seems to work
16:05:18  *** avdg has joined #openttdcoop.devzone
17:19:43  <Brot6> 2cc train set - Feature #1838 (New): NZR X Class (Voyager1) @
17:19:43  <Brot6> 2cc train set - Feature #1839 (New): NZR X Class (Voyager1) @
17:21:55  <Brot6> ai-admiralai: update from r73 to r75 done -
17:23:28  <Brot6> 2cc train set - Feature #1839 (Closed): NZR X Class (Voyager1) @
17:23:28  <Brot6> 2cc train set - Feature #1839 (Closed): NZR X Class (Voyager1) @
17:23:46  <Brot6> firs: update from r1509 to r1511 done (3 errors) -
17:24:24  <Brot6> Following repos didn't need a nightlies update: 2cctrainset (r651), 32bpp-extra (r39), airportsplus (r69), basecosts (r22), belarusiantowns (r7), comic-houses (r71), fish (r415), frenchtowns (r4), grfcodec (r788), heqs (r504), indonesiantowns (r38), manindu (r5), metrotrackset (r56), newgrf_makefile (r232), nml (r1034), nutracks (r117), ogfx-trains (r90), ogfx-trees (r41), opengfx (r554), openmsx (r97), opensfx (r97), smts (r19),
17:24:24  <Brot6> snowlinemod (r45), swedishrails (r188), swisstowns (r21), transrapidtrackset (r15), ttdviewer (r26), ttrs (r23), worldairlinersset (r667)
17:33:29  <Terkhen> <--- using that grf as a test, the only difference between trunk and patched is that when refitting to passengers, in the patched version the passenger value from CB15 takes precedence
17:34:05  <Terkhen> of course I had to correct when is mail capacity evaluated
17:35:30  <Terkhen> I still have to check if there are any differences in other newgrf sets, but IMO this behaviour for CB36 aircraft mail capacity is good
17:37:32  *** andythenorth has joined #openttdcoop.devzone
17:37:51  <Terkhen> hi andythenorth
17:37:58  <andythenorth> hi Terkhen
17:41:25  <planetmaker> ho andythenorth
17:46:38  <frosch123> Terkhen: looks fine :)
17:46:57  <Terkhen> frosch123: thanks :)
17:47:21  <Terkhen> I'm coding a patch to compare capacities more quickly, I want to see if it breaks something
17:50:25  *** andythenorth has quit IRC
17:51:13  <planetmaker> that was a silent and short intermezzo of his ;-)
17:51:58  <Terkhen> :P
17:52:03  <frosch123> "his andyness"?
18:18:32  <Brot6> 2cc train set - Feature #1840 (New): 6K (Voyager1) @
18:47:35  *** andythenorth has joined #openttdcoop.devzone
19:13:59  <Brot6> 32bpp-ez-patches: update from r21152 to r21176 done -
19:22:06  <Terkhen> hmmm... I found a difference; in the patched version, the chimaera from 2cc has 0 capacity for passengers instead of 400
19:23:28  <planetmaker> he
19:24:30  <Brot6> clientpatches: update from r21152 to r21177 done -
19:25:47  <Terkhen> wow, this thing is a monster :)
19:26:32  <Brot6> serverpatches: update from r21152 to r21178 done (19 errors) -
19:27:54  <planetmaker> the 2cc Chimera?
19:29:27  <Terkhen> yes, just running it for a year cost a king's ransom
19:29:42  <planetmaker> DJN likes it EXPENSIVE
19:30:22  <planetmaker> maybe it also has speed-dependent running costs. Who knows. 2cctrainset has it
19:31:32  <andythenorth> frosch123: you were working on something for shifting color selectively in grfcodec?
19:31:58  <frosch123> yes, but it needs polishing
19:32:27  <frosch123>
19:33:57  <Terkhen> hmm... even after coding these callbacks in nml, I can't understand their nfo version at all
19:34:11  * planetmaker hugs Terkhen
19:36:07  <frosch123> Terkhen: i do not understand nml generated nfo either :)
19:36:55  <frosch123> maybe one of the ahy magicians can :p
19:37:11  <planetmaker> that's not so bad actually...
19:37:24  <planetmaker> but generally these switch statements in nfo are not much fun
19:38:16  <frosch123> planetmaker: well, i fear somewhen we need a newgrf debugger during the grf loading stage
19:39:06  <planetmaker> hm... yes, maybe
19:39:13  <planetmaker> ingame it works already, though :-)
19:39:37  <frosch123> yes, ingame is fine, but i mean for all the action679d stuff :)
19:40:10  <planetmaker> :-) yeah... that's a mystery
19:40:12  <frosch123> during coding mi2 i was totally lost in between as i had no idea how to debug it
19:40:47  <frosch123> for hand-written nfo grf2hml was quite fine, but for nml-generated it is useless :)
19:42:07  <planetmaker> hm... yeah... it's too easy to write too complicated if then else there :-)
19:43:31  <frosch123> hmm, maybe nml could create actionCs with line number information
19:52:50  <Terkhen> <--- if any nfo magician can help me... I want to understand what are CB15 and CB36(capacity) doing, and if it is possible that any of them can return 0 under some case
19:53:13  <frosch123> +		for (char **s = this->strings; s < lastof(this->strings); s++) free(*s); <- <= or endof()
19:53:29  <frosch123> wrong channel :)
20:00:06  <Terkhen> meh
20:00:16  <Terkhen> there are many differences with other train sets too
20:00:42  <Terkhen> and I'm only checking the engine callback (with the default cargo)
20:03:05  <frosch123> the ingame one won't differ much
20:14:27  <frosch123> Terkhen: that engine does not distinguish cb15 between purchase list and ingame
20:16:08  <Terkhen> I still can't see why it returns 0 if cb15 is given priority... I can't make much sense of the nfo code, but given that there are defines for the capacity amounts, I doubt that it is intented to return 0
20:16:25  <frosch123> cb15 returns 0 if the cargoclass is not mail, armored or paaenger+express
20:16:58  <frosch123> line 120 - 125
20:17:47  <Terkhen> oh
20:18:29  <Terkhen> so the value for normal passengers is actually missing?
20:18:33  <Terkhen> (from cb15)
20:18:36  <frosch123> yes
20:18:51  <frosch123> quite weird usage of cargo classes, if you ask me :)
20:19:39  <Terkhen> the other sets with problems are NARS (cattle car having capacity of 8 instead of 9) and UKRS (a lot of different vehicles with different capacities, some of them even halved)
20:19:48  <frosch123> but considering lines 113-118, maybe the tourist line shall just get duplicated
20:20:07  * Terkhen tests
20:21:10  <frosch123> Terkhen: maybe add some printf or whatever in VehicleGetVariable to test whether the grfs access variables unavilable in purchase list
20:24:05  <Terkhen> okay
20:26:04  <Terkhen> yes, the problem with the chimaera is solved just by copying the extra pax line in 113-118 so the problem with this set is something minor
20:26:25  <V453000> frosch123: hi :) would it be possible to add an option to manual industries to totally disable oil rigs in temperate?
20:26:49  <V453000> so that oil behaves like in tropic/arctic with default industries
20:27:20  <frosch123> possible? yes. but then i would have to add 20 (?) switches for all industries...
20:27:26  <V453000> :|
20:28:13  <frosch123> why do you hate oilrigs so much? :)
20:29:28  <V453000> they arent random enough :)
20:29:35  <V453000> oil wells make much better balance over the whole map
20:31:44  <V453000> oh no, there is only fixed date of 1960 :(
20:32:22  <V453000> wouldnt it be possible to make it variable ? setting that to like 10 000 would be quite useful :)
20:34:34  <frosch123> hmm, you mean extending the bool switches to a year...
20:34:37  <frosch123> sounds better :)
20:38:31  <frosch123> would that make mi2 interesting for a #ccop server? :p
20:39:08  <planetmaker> frosch123: at some stage I'll canibalize parts of mi2 and name that OpenGFX+ Industries ;-)
20:39:19  <frosch123> haha
20:51:42  *** andythenorth has quit IRC
21:02:19  *** Alberth has left #openttdcoop.devzone
22:23:05  *** avdg has quit IRC
22:24:40  *** avdg has joined #openttdcoop.devzone
22:37:02  *** andythenorth has joined #openttdcoop.devzone
22:42:46  *** Lakie` has joined #openttdcoop.devzone
22:49:53  *** Lakie has quit IRC
22:56:48  *** frosch123 has quit IRC
23:09:58  *** Lakie` has quit IRC
23:20:17  *** andythenorth has quit IRC
23:32:21  *** ODM has quit IRC
23:34:33  *** Yexo has quit IRC
23:34:44  *** Yexo has joined #openttdcoop.devzone
23:50:18  *** thgergo has quit IRC

Powered by YARRSTE version: svn-trunk