IRC logs for #openttd on OFTC at 2008-11-13
⏴ go to previous day
00:20:34 *** gauthier__ has joined #openttd
00:24:47 *** gauthier__ is now known as gauthier
00:28:30 *** SmoovTruck has joined #openttd
00:29:20 <SmoovTruck> minor website bug... frontpage shows a screenshot thumbnail... clicking on it, takes me to the screenshots page... the screenshot shown in the thumbnail, isn't found on the page
00:33:09 *** Eddi|zuHause has joined #openttd
00:54:23 <DASPRiD> Hey, probably someone of you can help: When have a byte (hex: 0D), and interpret it as signed int, what should I get?
00:56:06 <DASPRiD> means the grf format definition can't be right
00:56:20 <DASPRiD> it's about the compression
00:56:57 <DASPRiD> when $byte & 1, it follows a repeated chunk, but 13 is the very first byte of the data stream, which means there can't be bytes before to repeat
01:00:58 <xerxesdaphat> what compression format do they use?
01:01:43 <SmatZ> sorry, I don't really understand your problem - but yes, it is possible to force the decoder to "seek" before the beginning of decoded stream
01:01:51 <DASPRiD> some variation of LZ77
01:02:55 <DASPRiD> SmatZ, a) it defiens an offset of 1200, but we are only at byte position 200 yet, b) the grccodec doesn't "support" seeking to a previous sprite stream
01:03:44 <xerxesdaphat> heh my saturday morning exam is on stuff similar to this
01:03:58 <xerxesdaphat> i think the case study is LZW though :P
01:04:40 <SmatZ> DASPRiD: neither OTTD would decode such sprite
01:05:41 <SmatZ> is it possible you mistaked the byte order in word? (big / little endian)
01:06:59 <DASPRiD> is big endian default machine order on pc?
01:07:25 <DASPRiD> that could then actually lead to some problems ;)
01:08:07 <DASPRiD> why does GRF use big endian anyway? :X
01:08:55 <SmatZ> I don't know, and I don't know who is responsible for it :)
01:09:04 *** SmoovTruck has joined #openttd
01:09:49 <SmatZ> xerxesdaphat: you are swapping nibbles, not bytes :)
01:10:31 <xerxesdaphat> you're right -- i think i'm correct if i'm using an alpha processor though
01:10:31 <SmatZ> 00 0D = 13 in BE, but 3328 in BE
01:10:53 <SmatZ> xerxesdaphat: I don't know :-x probably you are, when you say so :)
01:11:17 <DASPRiD> are you actually sure about the big endian thingy?
01:11:28 <glx> <SmatZ> 00 0D = 13 in BE, but 3328 in BE <-- good way to confuse people :)
01:11:30 <SmatZ> 00 0D = 13 in BE, but 3328 in LE
01:11:52 <DASPRiD> here is nothing written about the byte order
01:12:08 <vraa> does anyone know if openttd is multi threaded?
01:12:25 <glx> well it is but only for saving
01:12:39 <vraa> is there any work being done to make it so? because it seems like it could be parallelized
01:13:05 <glx> everything rely on a precise order
01:13:44 <DASPRiD> SmatZ, also i get quite strange results in big endian, which don't fit to the result grfcodec gives
01:13:46 <xerxesdaphat> static void GRFInfo(byte *buf, int len)
01:13:46 <xerxesdaphat> /* <08> <version> <grf-id> <name> <info>
01:13:46 <xerxesdaphat> * B version newgrf version, currently 06
01:13:48 <xerxesdaphat> * 4*B grf-id globally unique ID of this .grf file
01:13:51 <xerxesdaphat> * S name name of this .grf set
01:13:53 <xerxesdaphat> * S info string describing the set, and e.g. author and copyright */
01:13:56 <xerxesdaphat> /* TODO: Check version. (We should have own versioning done somehow.) */
01:14:03 <xerxesdaphat> const char *name;
01:14:06 <xerxesdaphat> const char *info;
01:14:08 <xerxesdaphat> check_length(len, 9, "GRFInfo");
01:14:11 <xerxesdaphat> version = buf[1];
01:14:13 <xerxesdaphat> /* this is de facto big endian - grf_load_dword() unsuitable */
01:14:16 <xerxesdaphat> grfid = buf[2] << 24 | buf[3] << 16 | buf[4] << 8 | buf[5];
01:14:18 <xerxesdaphat> name = (const char*)(buf + 6);
01:14:21 <xerxesdaphat> info = name + strlen(name) + 1;
01:14:23 <xerxesdaphat> _cur_grffile->grfid = grfid;
01:14:26 <xerxesdaphat> _cur_grffile->flags |= 0x0001; /* set active flag */
01:14:28 <DASPRiD> glx, he can't he pasted yet
01:14:28 <xerxesdaphat> DEBUG(grf, 1) ("[%s] Loaded GRFv%d set %08lx - %s:\n%s\n",
01:14:28 *** xerxesdaphat was kicked by DorpsGek (glx)
01:15:00 <SmatZ> DASPRiD: Bits 0 to 2 are the high bits of an offset, with the low bits being in the next byte.
01:15:17 <DASPRiD> SmatZ, i know that, cause i read that ;)
01:15:24 <DASPRiD> i even use the shifting described there
01:15:26 *** HerzogDeXtEr has joined #openttd
01:16:21 <SmatZ> DASPRiD: (But "code" must be a signed char, of course.) is it signed in your code?
01:16:40 <DASPRiD> SmatZ, it always results in 13, independent of signed or unsigned
01:17:02 <SmatZ> DASPRiD: why do you think it is unsigned data?
01:17:09 <SmatZ> errrrrrrr compressed data?
01:17:23 <DASPRiD> which means the first bit is set
01:17:33 <SmatZ> The high bit of the code shows whether this is a verbatim chunk (not set) or a repetition of earlier data (set).
01:17:41 <SmatZ> first bit = code & 0x80
01:18:18 <DASPRiD> that could actually be my bug :)
01:20:27 <DASPRiD> SmatZ, ok probably you can help me a bit further
01:20:43 <DASPRiD> unsigned long length = -(code >> 3);
01:21:00 <DASPRiD> i don't have unsigned integers in php, so length results in a negativ number
01:21:25 <SmatZ> but code should be negative
01:21:42 <SmatZ> if it is compressed stream
01:21:47 <DASPRiD> you mean, as it is signed?
01:21:59 <DASPRiD> you mean, as it is signed?
01:22:03 <glx> high bit set means negative for signed
01:22:37 <DASPRiD> oh yeah, now it is :)
01:23:43 <SmatZ> .... I typed twice the same sentence ... and I didn't know about it :-x I should really sleep
01:29:43 <Kisfvo> How is everyone doing?
02:01:26 *** DASPRiD is now known as DASPRiD|off
02:27:50 *** benjamingoodger has quit IRC
02:46:23 *** Zealotus has joined #openttd
03:56:01 *** roboboy has joined #openttd
04:03:00 *** AgentLeMan has joined #openttd
04:05:48 <AgentLeMan> hello everybody :o) can someone help me with an error i get while compiling? the error is "C:\Program Files\Microsoft SDKs\Windows\v6.0A\include\winnt.h(7818) : error C4430: missing type specifier - int assumed." even though i redirected msvc8 to the downloaded SDK.
04:31:43 *** roboboy has joined #openttd
05:51:08 *** roboboy has joined #openttd
07:43:21 *** FloSoft has joined #openttd
07:49:30 *** Doorslammer has joined #openttd
08:00:48 *** roboman has joined #openttd
08:05:47 *** [com]buster has joined #openttd
08:10:30 *** TrogDoor has joined #openttd
08:13:56 *** ProfFrink has joined #openttd
08:14:18 *** DASPRiD|off is now known as DASPRiD
08:18:13 *** TrogDoor is now known as Doorslammer
08:20:16 *** ProfFrink is now known as Prof_Frink
09:03:20 *** Brianetta has joined #openttd
10:12:58 *** Vikthor has joined #openttd
10:22:13 <planetmaker> question: if I want to use the OpenGFX instead of the proprietary files. Can they be in a subfolder of ./openttd/data or must they be directly in that?
10:24:06 <planetmaker> well... but I cannot right now :)
10:24:19 <planetmaker> and there might be people who know :)
10:37:42 <Ammler> DASPRiD: OpenGFX is a base grf
10:37:52 <Ammler> not like other NewGRFs
10:39:36 <DASPRiD> Ammler, well, i have it in a subfolder as well?
10:40:39 <DASPRiD> dasprid@dasprid-desktop:~/openttdcoop/data$ find -name "OpenGFX*"
10:49:16 *** Aperculum has left #openttd
10:50:16 *** Seberoth has joined #openttd
10:50:56 <Ammler> DASPRiD: that is the newgrf version
11:03:53 <Rubidium> planetmaker: why wouldn't it be working?
11:24:26 <planetmaker> Rubidium: I have no idea and my guess is it will. :) I'll test it tonight :)
11:24:50 <planetmaker> The question arose because we might want to include it into our grf pack - and that usually has it's own subdirectory.
11:25:58 <planetmaker> that way it subsequently might be possible to answer the question "what files do I need" with "get that grf pack" instead of "get your original CD".
11:26:16 <planetmaker> and it'd be 100% legal.
11:30:34 <Ammler> planetmaker: it does work
11:30:56 <Ammler> now, we can discuss it further :P
11:35:27 <Ammler> Rubidium: great fix with sample.cat
11:36:20 <FauxFaux> I wonder how hard it'd be to play with just the minimap (hardcore mode).
11:36:56 <Ammler> but why does it need a file at all?
11:51:34 <china> being low on cash and building railway trough crops.
12:14:11 <china> im having major issues with this map
12:15:03 <Doorslammer> Bah, sodding bloody soddery
12:15:24 <Doorslammer> One day I'll have TRS2004 run something until it jolly well enjoys it
12:16:06 <Doorslammer> Trainz Rail Simulator 2004
12:27:17 <welshdragon> my content dispatcher plus can't find it's installation
12:27:53 <welshdragon> Doorslammer, #trainz !
12:27:59 *** stillunknown has joined #openttd
12:30:26 <Sacro> So... who here does tram development, we have a bug!
12:31:19 <Sacro> Aye, on standard server
12:31:31 <Sacro> all other trams navigate same tile just fine
12:32:32 <Doorslammer> Maybe hes lost? ;)
12:36:38 *** xerxesdaphat has joined #openttd
12:37:45 <Brianetta> It's the Teleporting Tram Bug
12:38:05 <Brianetta> It's caused when a tram is turned around on a tile boundary, and is supposed to have been fixed for 0.6.3
12:38:25 <Brianetta> Apparently it isn't quite as fixed as hoped
12:40:56 *** [alt]buster has joined #openttd
12:40:59 *** [alt]buster is now known as [com]buster
12:54:07 <china> i guess this channel is more suited for you programers behind the game, i guess.
12:54:14 * china is only init for the money
12:56:05 * Doorslammer declares himself a nonentity also
13:01:12 *** Antdovu has joined #openttd
13:40:09 *** Doorslammer has left #openttd
13:57:14 <DASPRiD> How to determine the palette which is used by a GRF? according to grfcodec, it's just defined by the beginning of the filename?
14:34:51 *** HerzogDeXtEr has joined #openttd
14:41:37 <Rubidium> DASPRiD: there's no reliable (for a computer) way to determine the palette
14:42:00 <DASPRiD> Rubidium, so how does openttd know it?
14:42:26 <DASPRiD> so which palette does openttd use then? :x
14:42:46 <glx> it just knows the palette to use based on original files names
14:42:59 <Rubidium> we know the md5 of the dos/windows original graphics and assume the rest has the same palette
14:43:45 <DASPRiD> there should really be a standard in newgrf defining the palette ;)
14:44:23 <glx> there's a "standard', windows palette grf have a 'w' in the name
14:44:38 <glx> but we don't rely on it :)
14:47:00 <glx> so you want each grf to be 768 bytes bigger?
14:47:12 <DASPRiD> it should just define the palette name
14:47:24 *** FauxFaux has joined #openttd
14:54:23 <Eddi|zuHause> some grfs do not even have graphics that need specifying a palette
14:58:21 *** Dred_furst has joined #openttd
14:59:46 <Eddi|zuHause> it should theoretically be possible to create dual dos/win grfs, that have both graphics in them, and chose the right ones based on context
14:59:55 <Eddi|zuHause> but no grf author does that
15:00:01 <glx> btw in nightlies you can switch palette in newgrf config
15:01:48 <Eddi|zuHause> has anyone considered, for NFO version 8, to introduce a flag in the defining action 8 that the grf specifies dos/win/don't care?
15:05:10 *** davis is now known as davis-
15:12:35 *** Progman has joined #openttd
15:22:51 <DASPRiD> glx, Eddi|zuHause: long offset = BE_SWAP16(ibuffer[4+y]) + 8; <-- what does this line do?
15:22:51 <Priski> How do I can get a sertain nightly build (not the latest), is it even possible without compiling from source?
15:23:34 <Eddi|zuHause> DASPRiD: big endian conversion?
15:24:03 <DASPRiD> so uhm, what exacly? :)
15:24:25 <Eddi|zuHause> Priski: there should be an archive somewhere
15:24:35 <Eddi|zuHause> try binaries.openttd.org
15:25:20 <Priski> ah thanks, I remembered that it should be somewhere but couldn't find a link anywhere
15:25:21 <Eddi|zuHause> DASPRiD: conversion between little endian and big endian means swapping the high and low byte
15:25:56 <DASPRiD> (i'm no c-coder, sorry :))
15:25:59 <Eddi|zuHause> an array access?
15:26:28 <DASPRiD> ah so it get's the 4+y byte of iBuffer?
15:26:52 <Eddi|zuHause> two bytes probably
15:27:07 <Eddi|zuHause> as the 16 indicates 16 bit
15:27:23 <DASPRiD> oh okay, and the + 8 at the end then?
15:28:13 <Eddi|zuHause> it's an offset correction... it really depends on context what that +8 means
15:28:29 <DASPRiD> U16 *ibuffer = (U16*) buffer;
15:28:35 <Eddi|zuHause> could be related to the 4 in the array
15:28:55 <Eddi|zuHause> 4th index, 2 bytes each, makes 8 bytes
15:29:54 <DASPRiD> after look again now at the definition it's clear
15:30:06 <Eddi|zuHause> but it could be totally unrelated also
15:30:15 <DASPRiD> no it looks correct :)
15:32:55 <Eddi|zuHause> i'm still not sure what you are actually trying to do ;)
15:34:46 <DASPRiD> Eddi|zuHause, a small library for reading grf files in a php application ;)
15:34:56 <DASPRiD> without depending on external programs like grfcodec
15:44:52 <Belugas> DASPRiD, something like grf2html?
15:45:01 <Belugas> granted, not in php, in DElphi, but still...
15:45:20 <DASPRiD> Belugas, not for such a purpose, but yeah something like that
15:45:31 <DASPRiD> wonderful, tile decoding works :)
15:45:57 <DASPRiD> and the entire grf decoding process only takes like < a second :)
15:51:17 *** Sacro|Uni has joined #openttd
16:02:42 *** benjamingoodger has joined #openttd
16:06:42 <benjamingoodger> still campaigning against channel noise, then, ln?
16:10:00 <ln> that was only my second hello today.
16:12:05 *** [alt]buster has joined #openttd
16:15:06 *** [alt]buster is now known as [com]buster
16:22:22 *** NukeBuster has joined #openttd
16:24:06 *** Seberoth was kicked by DorpsGek (Wrong channel. Retry in #openttdcoop.)
16:24:06 *** Seberoth has joined #openttd
16:27:14 <benjamingoodger> gets the job done without being rude
16:27:17 <benjamingoodger> ...for a script
16:32:31 *** Singaporekid has joined #openttd
16:39:37 <DASPRiD> Eddi|zuHause, are you still there? :x
16:40:17 <DASPRiD> i should read the correct documentations *G*
16:41:21 *** frosch123 has joined #openttd
16:51:49 <Rubidium> Eddi|zuHause: you should tell your GRF suggestion to frosch123
16:52:42 <Eddi|zuHause> has anyone considered, for NFO version 8, to introduce a flag in the defining action 8 that the grf specifies dos/win/don't care?
16:52:58 <frosch123> does that need a version bump?
16:53:26 <frosch123> or do you want to enforce specifying the palette?
16:53:36 <Eddi|zuHause> not necessarily, but it could just be fit in there, if it's going to be bumped anyway :)
16:53:48 <china> dihedral: (im playing him)
16:54:01 <china> dihedral: (and he's leading)
16:54:08 <Eddi|zuHause> i don't want to enforce anything, it was just a thought
16:55:23 <Rubidium> frosch123: it's useful for OpenTTD to know what palette the developer made the set with (so we don't have to guess the palette)
16:55:27 <frosch123> Eddi|zuHause: I meant, you do not need a version bump, when you want to add something.
16:55:32 <dihedral> china, then you two should getyourselves a channel for your game
16:56:09 <china> dihedral: does my one-liner offend you?
16:56:35 <dihedral> i was just thingking of future chatter ;-)
16:56:39 <frosch123> so, we want to make specifiying a palette (win/dos/don't care) mandatory for version 8?
16:56:51 <china> dihedral: i was just kicking his but in another medium. :)
16:57:22 <Eddi|zuHause> well, grfcodec could automatically fill this flag from the -p option
16:57:28 <Rubidium> dihedral: just /ignore people that annoy you; they're not worth to get an answer anyways ;)
16:57:46 <dihedral> yeah - i should do more often
16:57:59 <dihedral> but then i always forget to add the ignore to the other clients i use :-P
16:58:17 <Rubidium> then use only one client
16:58:22 <Eddi|zuHause> and it struck me that the action 8 would be an appropriate place
16:58:26 <china> dihedral: well, he and me, for the important part.
16:59:52 <Ammler> Rubidium: We tested your sample.cat "fix", works really nice, would it be complicated to allow no file instead of needing a empty file?
17:07:33 <frosch123> the sprite starts with FF
17:07:44 <frosch123> the follows the raw data
17:07:54 <frosch123> somewhere around the FF is the length as a word
17:08:06 <DASPRiD> little or big endian?
17:08:23 <frosch123> take a look at grf2html, it does not use those magic signed/unsigned/shifting stuff to decode it
17:08:42 <DASPRiD> he okay, probably a good idea
17:09:01 <Eddi|zuHause> i'd say with grfs it's generally a safe bet that everything is little endian
17:09:18 <DASPRiD> Eddi|zuHause, that's true, it is ;)
17:10:42 *** rubyruy has joined #openttd
17:13:17 <DASPRiD> frosch123, damn, the sourcecode of grf2html is even less readable to me than grfcodec ;)
17:13:28 <DASPRiD> well i just try a bit around, maybe it works ;)
17:13:42 <Eddi|zuHause> it's pascal, how unreadable can it be?
17:13:57 <frosch123> it is pascal, it uses english words to describe things instead of & | ~ !
17:14:04 <DASPRiD> i knew turbo pascal... like 15 years ago ;)
17:14:22 <frosch123> however you have to look into grfbase.pas IIRC
17:14:24 <Eddi|zuHause> well, pascal is significantly older than that :)
17:15:02 <frosch123> turbo vision 2 is about that age :p
17:15:33 <Eddi|zuHause> and reading a language you don't know is usually very easy, unless it's php :p
17:26:10 <DASPRiD> frosch123, may i ask another question? :)
17:26:28 <DASPRiD> is the first sprite, defining the total number of sprites, required in every grf?
17:27:01 <frosch123> i.e. the trg* start with a real sprite
17:27:27 <frosch123> btw. the number is not necessarily correct :p
17:28:38 <DASPRiD> frosch123, oh so i should better ignore it?
17:28:49 <DASPRiD> instead of throwing errors :)
17:29:15 <frosch123> yes, "ignoring" would be a good idea :)
17:46:03 <Belugas> hooo.. the Precipice!
17:59:29 *** Antdovu has joined #openttd
18:02:47 * Belugas guess it means good morning, so...
18:27:21 *** |Jeroen| has joined #openttd
18:32:32 <Priski> Damn, adding scenario sure takes lots of edits on wiki :/
18:37:02 <Antdovu> is there an easy way of creating a new type of window in ottd?
18:39:02 <DorpsGek> frosch123: Vikthor was last seen in #openttd 20 hours, 38 minutes, and 40 seconds ago: <Vikthor> night
18:40:27 *** ChanServ sets mode: +v tokai
18:47:40 <Belugas> Antdovu: it's not that hard. it's just a bit tedious
18:49:48 <Belugas> roughly: need to create the widgets array (boring part) ad the window's class, who is responsible for handling creation, painting and events
18:53:26 *** Killer11 has joined #openttd
18:57:10 <Prof_Frink> The ability to bridge over more stuff has certainly affected the AI's love of the bridge
18:58:09 <Killer11> dang i need to test that
18:58:30 <Prof_Frink> It's built a track, doubled back on itself and bridged directly over itself
18:58:30 <Killer11> got any retarded screenshots?
18:58:37 <Prof_Frink> then doubled back again
18:58:46 <Killer11> that's KINDA awesome
19:00:09 <Killer11> i have almost rebuilt my lithuanian set
19:00:21 <Killer11> one engine left to reinclude
19:00:39 <Killer11> God i LOVE my D1 DMU
19:00:57 <Killer11> was certainly one of my better works
19:01:06 <Killer11> and it was my 2nd ever train graphic
19:01:22 <Antdovu> that doesn't stop it from building 5 lines between 2 points and using sharp corners everywhere so all the trains get stuck
19:02:17 <DASPRiD> Eddi|zuHause, frosch123: thanks for your help, I've got the php grf decoder up and running :)
19:03:32 <Antdovu> does noai support multiple ai-s in the same company?
19:05:13 <Antdovu> it could be a great way to add garbage collection
19:06:58 <Prof_Frink> "Prenway Station #1336"
19:09:44 <Yexo> <Antdovu> does noai support multiple ai-s in the same company? <- no, only one AI per company, and no human players in the same company as an AI
19:10:00 <Eddi|zuHause> Prof_Frink: better luck next time :p
19:11:48 <Antdovu> I guess that means no FrenchMaidAI
19:12:29 <Aali> multiple AI's in the same company would be a great way to add insanity
19:13:03 <Antdovu> if they don't cooperate then obviously
19:13:04 <Aali> not only do AI's have to keep track of everything happening in the ottd world, they have to keep each other in check too
19:13:23 <Antdovu> it is the same with humans
19:13:35 <Antdovu> but openttdcoop isn't too bad, is it?
19:13:54 <Aali> last i checked, openttdcoop wasn't a bunch of AI's
19:14:26 <Eddi|zuHause> you can try to import multiple AIs into a controller AI
19:14:55 <Eddi|zuHause> or simply program one AI with a split personaltiy
19:15:16 <Antdovu> two human players in a single company might have problems as well
19:15:45 <Prof_Frink> manicdepressiveai: Builds loads of tracks really fast... then dynamites the lot.
19:15:45 <Aali> two human players in a company can understand each other
19:16:25 <Antdovu> Aali: just join a popular game and don't use a password
19:16:53 <Aali> and what does that prove?
19:17:04 <Aali> that there are idiots on this earth?
19:17:07 <Eddi|zuHause> i don't see a point in putting two AIs together that are not specifically prepared for it
19:17:14 <Aali> i knew that already, thank you very much
19:17:35 <Belugas> idiots on this earth? I'd be lloking over forums for those...
19:18:10 <Antdovu> Eddi|zuHause: my idea was that the "garbage collector" AI would be kind of like garbage collection in Java
19:18:53 <Aali> so just implement a garbage collector in your AI
19:19:09 <Antdovu> yeah, that's like C/C++
19:19:39 <Antdovu> I am not really fighting for or against either of them
19:19:50 <Antdovu> it is just a mind game
19:19:58 <Aali> you're not really going anywhere with any of this, i can see that
19:20:00 <Eddi|zuHause> if you try to mimic programming languages, how do you determine apart abandoned tracks from unfinished tracks?
19:20:23 <Eddi|zuHause> as opposed to objects that have no references anymore
19:20:48 <Eddi|zuHause> Antdovu: what if the track takes 5 years to finish?
19:21:17 <Antdovu> you could have a concept of a side track
19:21:28 <Aali> you run out of money, and the "garbage collector" destroys everything you've built
19:21:46 <Eddi|zuHause> you're making less and less sense with every statement you make...
19:21:57 <Killer11> ok seriously i need to get that last train coded
19:23:05 <Eddi|zuHause> no, your alter ego, that is talking garbage (collection)
19:23:35 <Antdovu> I am talking about AI, not AS :P
19:24:07 <Antdovu> there is no shortage of artificial stupidity :P
19:25:26 <Aali> who said anything about artificial stupidity?
19:26:06 <Antdovu> if done incorrectly then yes, it might destroy non-completed track, it might not understand what is being used etc.
19:26:32 <Antdovu> but there is no fundamental reason for it to actually do anything like that
19:26:50 <Aali> which is why this is a bad idea, if coupled to the "real" AI, it wouldn't have to guess
19:27:56 <Antdovu> it depends on the level you want to achieve
19:28:46 <Antdovu> one approach would require garbage collection from all AIs, the other would not
19:29:21 <Antdovu> obviously both ways have their pros and cons.
19:30:12 <Aali> if you really want to do this, a high-level resource manager library (with garbage collection) would probably be more useful
19:30:32 <Aali> that way its not AI-dependant and not as clueless
19:31:09 *** Yeggstry has joined #openttd
19:35:11 *** lobster_MB has joined #openttd
19:46:09 *** Vikthor has joined #openttd
19:48:03 <Killer11> ok Set is reimplemented
19:48:21 <Killer11> would there be anyone willing to test?
19:52:21 *** Patrick has joined #openttd
19:53:49 <Patrick> I have a conundrum for you concerning transfer orders
19:53:56 <Patrick> I have the order - transfer and wait for full load
19:54:12 <Patrick> and my trains just reload the stuff they dropped off if there isn't someone else waitloading on that cargo
19:55:39 <Aali> and why wouldn't they? you told them to full load
19:57:28 <Patrick> becuase I don't see how I can actually transfer any cargo then
19:57:34 <Patrick> or the purpose of that order
19:57:49 <Patrick> I want to relay passengers in both directions through an exchange
19:58:03 <Aali> indeed, that combination of order flags is fairly useless
19:58:08 <Antdovu> pretty sure that that is impossible
19:58:09 <Aali> you want cargodest, then
19:58:19 <Yexo> that's only possible if you split the transfer station in two
19:58:23 <Patrick> my pax have an "en-route" flag on it
19:58:28 <Patrick> this is actually an accident
19:58:41 <Patrick> I built my goods dropoff shuttles from engines that could hold pax
19:58:57 <Patrick> and now they're all full of people going round and round and round in a big circle
19:59:03 <Patrick> my pickup doesn't accept pax, so ...
20:00:03 <Aali> with cargodest you dont have to worry about transfers, the cargo will route itself through your network
20:00:43 <Patrick> last I heard this was speculative
20:00:47 *** Nite_Owl has joined #openttd
20:01:01 <Belugas> [14:50] <Patrick> long time no see <-- heheh
20:01:33 <Yexo> it's not yet in trunk, but it is (almost) done
20:01:39 <Aali> Patrick: well, cargodest isn't "done", but its certainly functional
20:01:44 <Yexo> see openttd.org/download-cargodest
20:01:48 <Patrick> oh, not this knob jockey again
20:02:45 <Antdovu> anyone know any XML editors/viewers for windows that don't die on a file just a few MB large?
20:04:24 <Patrick> Aali: well, the flags aren't totally useless, I just need a waitload train and it Just Works
20:04:47 <Aali> still, fairly useless :)
20:05:35 <Patrick> now I can have trains that carry two types of cargo and deliver one at the same station they route the other one to
20:05:41 <Patrick> without any extra orders or linework
20:05:43 <Vikthor> frosch123: Your patch seems to be working, thank you very much.
20:05:50 <Patrick> man, so many new faces
20:05:58 <Patrick> and some very old ones as well
20:06:10 <Aali> Patrick: with cargodest all you need is regular goto orders ;)
20:07:20 <Patrick> I prefer it that my cattle can't read maps
20:07:32 <Patrick> next they'll operate automatic weapons and it'll be a tom lehrer song
20:07:40 *** Dred_furst has joined #openttd
20:11:36 <frosch123> Vikthor: thanks for testing :)
20:13:03 <Vikthor> though I tested it only for RVs
20:14:14 <Patrick> who wrote timetabling and shared order management
20:14:30 <planetmaker> exist at least for a year.
20:20:29 <Antdovu> Vikthor, frosch123: what does the patch do?
20:20:57 <Vikthor> Allows for refiting to zero capacity
20:21:14 <Patrick> big planes crash and die horribly at commuter airports?
20:26:21 <Nite_Owl> with alarming frequency - it is considered a small airport
20:26:29 <CIA-5> OpenTTD: frosch * r14578 /trunk/src/ (roadveh_cmd.cpp train_cmd.cpp): -Fix (r2389, r10811)[FS#Vikthor]: Allow capacity callbacks (15, 36) to return zero capacity.
20:27:07 <Vikthor> hey, I do not feel like being a FS bug number :p
20:27:51 <Vikthor> but thank you for the fix
20:49:58 <Killer11> anyone wants to try a new set?
20:50:11 <Killer11> 6 trains and 1 truck
20:50:17 <Killer11> not big but could be fun
20:56:55 <Killer11> you obviously don't know how horrible can mailmen be
20:57:16 <Killer11> and ponies are overused
20:58:40 <Antdovu> I don't want to know why you like goats, everyone likes unicorns
20:59:01 <Killer11> becouse goats are constantly high
20:59:10 <Killer11> they usualy eat anythign you give them
20:59:16 <Killer11> not caring what it is
20:59:31 <Killer11> or how dangerous it is to eat it
20:59:36 <benjamingoodger> my system of RFID-tagged vacuum tubes is far superior to the british postal system; unfortunately, it will never be implemented, because the british postal system is an effective state monopoly
20:59:52 <benjamingoodger> also, I cba to chuck the hundreds of millions of pounds at it that it would require to get set up
21:00:07 <Killer11> oh that is nothing my friend
21:00:18 *** AgentLeMan has joined #openttd
21:00:31 <AgentLeMan> hello everybody :o) can someone help me with an error i get while compiling? the error is "C:\Program Files\Microsoft SDKs\Windows\v6.0A\include\winnt.h(7818) : error C4430: missing type specifier - int assumed." even though i redirected msvc8 to the downloaded SDK.
21:00:31 <Killer11> lets not forget american mail system run my GREASY REDNECK Mailmen
21:00:53 <Vikthor> (21:58:40) Antdovu: I don't want to know why you like goats, everyone likes unicorns -- hmmm unicorns ...
21:02:03 <AgentLeMan> or was that not enough info?
21:04:17 <Antdovu> I use v6.0A without problems
21:04:29 *** mortal` has joined #openttd
21:04:36 *** Mortal is now known as Guest3117
21:04:36 *** mortal` is now known as mortal
21:04:46 <glx> but I had no problem with v6.0A (except it doesn't support vista icons)
21:05:02 <Antdovu> except that building a release takes about 3 * forever
21:05:28 <Antdovu> what do you mean it doesn't support vista icons?
21:06:38 <glx> rc.exe doesn't understand the format used for big icons in vista
21:06:54 <AgentLeMan> glx, hm, i set up the system with turtoisesvn and all the downlaoded things as described in uhm... "readme_windows_msvc.txt" of the source of 0.6.3 . ( using win xp here ) and i thought, as i told MSVC that is hall use the downloaded libraries.. now it still searches in the other folder :o( im lost
21:07:17 <AgentLeMan> is hall = it shall
21:07:32 <Antdovu> glx: so I am just missing big icons?
21:07:33 <AgentLeMan> sadly buildopenttd wont work for me either
21:08:07 <glx> there are no big icons for openttd for now
21:08:34 <glx> but maybe one day we'll use it (and most MSVC 2005 users will have compile failure)
21:08:57 <Antdovu> I still have the orange square + $ sign icon...
21:08:58 <AgentLeMan> oh, wait, i think i found the error growls
21:09:22 <AgentLeMan> the readme says "NOTE: make sure that the directory for the DirectX SDK is the first one in the
21:09:22 <AgentLeMan> list, above all others, otherwise compilation will most likely fail!!"
21:09:41 <AgentLeMan> but... i na forumpost i read, it has actually to be at the end of the list
21:11:34 <glx> Antdovu: vista needs a 256*256 icon
21:11:59 <glx> hmm wait 256*256 is very big
21:12:13 <benjamingoodger> ha, imagine if any OS needed a 512x512 one...that'd be stupid...
21:12:37 <AgentLeMan> glx, what do you use to compile it?
21:15:28 <glx> 2005 express and 2008 express
21:15:44 <AgentLeMan> hm, same as i ( 2008 express )
21:15:45 <glx> SDK v6.1 is installed with 2008 express
21:16:00 <AgentLeMan> okay, it works now, just getting various other errors
21:16:38 <AgentLeMan> so it really has to be at the lower end of the list, not the top, as "first one" implies
21:16:41 <glx> you don't need to install platform sdk yourself for 2008 express
21:17:06 <AgentLeMan> hm, i did already, as said, like written in the compile.readme
21:18:05 <glx> $(WindowsSdkDir)\include <-- this one is the sdk
21:18:47 <glx> wiki is more accurate I think
21:18:50 <AgentLeMan> "Compiling OpenTTD using Microsoft Visual C++
21:19:12 <glx> btw it's valid for 2005 express I think :)
21:19:31 <AgentLeMan> well, i dont get the sdk errors anymore. thats solved
21:19:53 <AgentLeMan> but as it seems, i now need and older DX SDK, as the newer one doesnt support direcvt music (?)
21:20:59 <AgentLeMan> i found a forujmpost with a link to the wiki
21:21:12 <Antdovu> the music isn't any good anyway...
21:21:55 <benjamingoodger> you take that back!
21:21:57 <Antdovu> do you have the sound effects if compiled without music?
21:22:24 <glx> and you also have music without dmusic
21:22:26 *** sulai4511 has joined #openttd
21:23:06 <Antdovu> so why would you need the old directx music stuff?
21:23:42 <glx> because win32 music doesn't work for some people while dmusic works
21:24:39 <AgentLeMan> glx, i assume, i have to deinstall the new SDK? or can i just extract it to somewhere and move the files over as an added SDK and then reference it ( somehow?)
21:24:57 <AgentLeMan> oh hell... another 450 MB
21:25:58 <glx> the wiki says you can extract the required files from the older DX SDK
21:26:35 <glx> but as I don't do any DX programming I just have the old SDK :)
21:28:20 * Belugas leaves in a hurry and waves good night to those who care
21:28:35 <AgentLeMan> good night, Belugas
21:29:16 *** sulai4511 has joined #openttd
21:30:18 *** sulai4511 has joined #openttd
21:51:58 <ln> so.. err.. in CUDA, besides the actual __global__ function that does things on GPU, one is supposed to have a host function. but since the host function is in the .cu, is it like compiled to object code by nvcc, or what?
22:06:07 <AgentLeMan> glx, your right yawns it takes aaaages
22:08:30 *** Nite_Owl has joined #openttd
22:09:04 *** DeGhost has joined #openttd
22:21:12 <dihedral> \o/ found a bug in orders :-P
22:21:40 <dihedral> 2 vehicles with no orders, let them share the (empty) orders
22:22:00 <dihedral> delete the second vehicle
22:22:05 <dihedral> create a new vehicle
22:22:14 <dihedral> orders are still shared
22:23:47 <Ammler> if you delete a vehicle and create one, it does use the just deleted orders.
22:24:15 <Ammler> iirc, that was in TTO already.
22:24:29 <AgentLeMan> must be the same driver then, who just not wants to learn a new route, as hes old
22:25:16 <Antdovu> did you know that pink, sparkling unicorns drive vehicles in ottd?
22:25:55 <Ammler> dunno, how long the "TTL" is.
22:26:53 <Antdovu> except when run over by a train
22:27:19 <benjamingoodger> that explains why the drivers never need to rest
22:29:19 <benjamingoodger> may god have mercy on our souls
22:31:24 <AgentLeMan> the drivers have the easiest job in ottd.... imagone how those masses must suffer, who, if you change your companycolor, run around like crazy in the land and repaint railfences, cars, trains and all that. WHILE they move those trains....
22:32:28 <Antdovu> what about repainting the flying planes? ;)
22:32:46 <Antdovu> that should be fun :)
22:33:04 <Prof_Frink> Antdovu: Paintball anti-aircraft guns.
22:34:21 <Antdovu> anyone have any good unicorn pictures?
22:34:23 <AgentLeMan> dang, where the hell is my trunk?
22:35:15 <AgentLeMan> Antdovu.. actually, i have a photo of YOU
22:36:01 <Antdovu> how did you manage to take that photo?
22:36:04 <AgentLeMan> its 1280x1225 pixels, should be enough as wallpaper
22:37:31 <AgentLeMan> im resourceful ;o)
22:37:57 <Antdovu> great, then I can casually use it in a lab
22:38:58 <AgentLeMan> wooooooohooooo, a clean build! dances around happyly
22:39:13 <AgentLeMan> now ican look for the stuff im intersted to change
22:39:47 <Antdovu> yeah, and you can make it crash more often :)
22:39:56 <Antdovu> I'm an expert at that
22:42:55 <AgentLeMan> hmmm, Antdovu, you just injected anidea into me... crash more often?...
22:43:08 <AgentLeMan> make cars crash trains too!
22:45:27 <AgentLeMan> waves thanks for help and a good day/night to you all :o)
23:02:18 *** HerzogDeXtEr has joined #openttd
23:24:49 *** Yeggstry is now known as Yeggzzz
23:50:27 *** Doorslammer has joined #openttd
continue to next day ⏵