IRC logs for #openttd on OFTC at 2008-06-24
⏴ go to previous day
00:05:12 <glx> compilable with mingw but not directly
00:05:38 <glx> and it should be compilable on linux too
00:08:08 <ccfreak2k> I can't get Kompare to make the changes.
00:10:08 <ccfreak2k> I dunno how to use it tho. :|
00:16:05 <ccfreak2k> Failed patching openttd_vs90.proj
00:16:19 <ccfreak2k> I don't have VS9 so I guess it's just as well.
00:26:24 <glx> TiberiusTeng: why did you remove directmusic support ?
00:27:50 <TiberiusTeng> I don't have DirectX SDK :Q
00:28:26 <ccfreak2k> opengl.hpp is expecting windows.h, so it looks like it won't compile on Linux.
00:28:31 <ccfreak2k> At least not without Windows SDK.
00:29:42 <TiberiusTeng> I'll clean it before release it, and those perhaps unwanted project file changes too.
00:30:09 <ccfreak2k> I'm gonna boot back into Windows.
00:30:14 <TiberiusTeng> actually you can remove that line, and it compiles ..
00:30:30 <TiberiusTeng> it's certainly something I used to experiment in early stage.
00:30:52 <glx> you should use #include "../glee.h" and put glee.h in src
00:31:05 <TiberiusTeng> anyway you can't use it outside Windows yet, unless you make a video driver that do proper initialization in your platform.
00:31:07 <ccfreak2k> Did I mention the win32 binary crashes on launch?
00:31:38 <TiberiusTeng> yes, but there's little useful information to trace it here ...
00:32:12 <ccfreak2k> I don't have any useful trace tools installed.
00:32:13 <TiberiusTeng> glx, I plan to treat GLee.c/h like libpng/zlib
00:32:51 <TiberiusTeng> nope, it must available
00:33:00 <ccfreak2k> #define WITH_GLEE /* :D */
00:33:15 <TiberiusTeng> but I'm reluctant to put that in OpenTTD's source directory ...
00:33:25 <TiberiusTeng> license-wise I'm allowed to do that though
00:33:39 <glx> it should be possible to do like we did for squirrel
00:34:14 *** Eddi|zuHause3 has joined #openttd
00:34:15 <glx> putting the files in 3rdparty and add external in src
00:34:56 <TiberiusTeng> I don't have these directory structure in trunk, what should I do?
00:35:18 <ccfreak2k> VisualStudio reads diffs, right
00:35:43 <TiberiusTeng> I'm using MSYS to hg/svn/patch/vim/whatever
00:36:30 <ccfreak2k> Well my laptop runs Slackware, so I can just do it here.
00:39:17 <glx> for now I applied the diff in an hg checkout ;)
00:40:02 <ccfreak2k> What about sticking the GLee files in the diff?
00:42:07 <TiberiusTeng> yeah, I think I'll do that
00:43:01 <TiberiusTeng> they're huge, but they're useful. manually linking to various OpenGL extensions on all platforms is not an easy task ...
00:51:01 <TiberiusTeng> I still can't connect to your host :~
00:52:05 <ccfreak2k> Now I just need to wait for VisualStudio to install.
00:54:18 <TiberiusTeng> glx, connection timed out, I think it's not DNS problem
00:57:22 <ccfreak2k> changeset: now compiles with mingw
00:57:24 <ccfreak2k> changeset 9584: 1433c2efe183
00:58:44 <TiberiusTeng> maybe my ISP did some strange blocking :p
00:59:24 <TiberiusTeng> ahh, my notebook at school connects. great :)
01:00:16 <TiberiusTeng> I'll do these modifications too, thanks
01:00:19 <glx> small changes but big effect ;)
01:01:22 <glx> (I have glee.c and glee.h in src)
01:14:51 <ccfreak2k> png and zlib are external dependencies, right?
01:24:28 <ccfreak2k> By "Platform SDK" it means "Windows SDK" right?
01:24:50 <ccfreak2k> Screenshots need updating then.
01:24:53 <glx> but it's included in 2008 express
01:29:44 <ccfreak2k> Oh god DirectX SDK is huge.
01:29:46 <ccfreak2k> I forgot about that.
01:30:10 <glx> you don't need directx for opengl patch
01:30:26 <glx> but you do for clean trunk
01:40:07 <ccfreak2k> dmksctrl.h is part of the DirectX SDK right?
01:40:28 <ccfreak2k> dmusic.cpp needs that.
01:41:08 <ccfreak2k> Also, it can't seem to find GLee.h.
01:41:49 <glx> of course you are using vs80.sln and opengl patch didn't change these files
01:42:31 <glx> edit openttd properties and remove DIRECT_MUSIC from preprocessor definitions
01:42:59 <ccfreak2k> Actuall it DID patch the vs80.proj.
01:43:26 <glx> it added files, but he disabled dmusic in vs90 only
01:43:46 <glx> for glee.h you need the changes I made for mingw
01:45:18 <ccfreak2k> Too bad I can't connect to your box it seems.
01:45:56 <glx> btw this hg server is usually not started ;)
01:46:42 <glx> I need to find a way to integrate it in my IIS
01:54:07 <ccfreak2k> dxguid.lib is also in DirectX SDK, right?
01:55:19 <ccfreak2k> Yeah the linker wants that.
02:07:57 <ccfreak2k> You mean -ldxguid right?
02:08:32 <glx> yes in linker properties for openttd
02:08:46 <glx> (like it's done for vs90)
02:09:47 <ccfreak2k> Oh, I missed that part.
02:12:53 <ccfreak2k> RandomizedBaseAddress and FixedBaseAddress are both set to 1.
02:23:17 <ccfreak2k> succeeded, that is.
02:37:08 <ccfreak2k> I built it and ran it.
02:37:13 <ccfreak2k> Now I get a black screen.
02:38:32 <ccfreak2k> Yeah, black screen.
02:59:51 <ccfreak2k> Someone beep me when TiberiusTeng gets back.
03:05:05 <ccfreak2k> I think I found the problem.
03:05:25 <ccfreak2k> glGenProgramsARB() expects a pointer as the second argument.
03:21:29 <ccfreak2k> According to the book, an array to recieve the program names.
03:21:37 <ccfreak2k> Although I don't think that's the problem actually.
03:21:48 <ccfreak2k> It's hard to pin down since I can't find the function being called where the error is happening.
03:27:12 <ccfreak2k> Actually I guess it's in a standard opengl library somewhere.
03:35:14 <ccfreak2k> I can't trace where it's defined in VS because glee.h defines it as one of its own functions...which calls the function in question.
04:09:53 <ccfreak2k> It would probably also benefit us ATi users if he supported EXT_texture_rectangle.
05:42:23 *** raimar3 has joined #openttd
05:48:18 *** XeryusTC is now known as Xeryus|bnc
06:47:53 <TiberiusTeng> ccfreak2k, it gave you a blackscreen? which ATI card are you using? I would bet the problem's at shader ...
07:01:33 *** einKarl has joined #openttd
07:21:48 *** Wezz6400 has joined #openttd
07:32:55 *** stillunknown has joined #openttd
07:35:17 *** GoneWacko has joined #openttd
07:39:30 *** flowOver has joined #openttd
08:17:28 *** stillunknown has joined #openttd
08:18:32 *** [com]buster has joined #openttd
08:25:19 *** thgergo has joined #openttd
08:34:51 *** Doorslammer has joined #openttd
08:40:42 *** dR3x4cK has joined #openttd
09:05:48 <CIA-6> OpenTTD: smatz * r13618 /trunk/ (9 files in 3 dirs):
09:05:48 <CIA-6> OpenTTD: -Codechange: move tar-specific declarations to separate file
09:05:48 <CIA-6> OpenTTD: Removes indirect dependency on <string> and <map> for over 40 files
09:15:58 <CIA-6> OpenTTD: smatz * r13619 /trunk/src/ (blitter/factory.hpp driver.cpp driver.h):
09:15:58 <CIA-6> OpenTTD: -Codechange: use 'const char *' instead of std::string for blitter and driver names
09:15:58 <CIA-6> OpenTTD: Removes indirect dependency on <string> for 20 files, reduces binary size by 16kB
09:24:30 *** Singaporekid has joined #openttd
09:36:40 *** [alt]buster has joined #openttd
09:36:42 *** [alt]buster is now known as [com]buster
09:38:51 * TiberiusTeng packing the OpenGL blitter alpha (?) release
09:49:56 *** GoneWacko has joined #openttd
10:20:39 *** Frostregen has joined #openttd
10:40:07 <SmatZ> Bjarni: if you will ever read that :) see commit r12930, NetBSD should be supported
10:40:19 <DorpsGek> SmatZ: Commit by rubidium :: r12930 /trunk (3 files in 3 dirs) (2008-04-30 07:39:46 UTC)
10:40:20 <DorpsGek> SmatZ: -Add: build support for NetBSD and HP-UX. Patch by ahoka.
10:43:39 <SmatZ> "Why not hardware-'accelerated' you say? Because it's not necessarily faster 8) " :-)
10:44:05 <TiberiusTeng> but it do 600fps on my machine 8)
10:44:46 <SmatZ> oh noes I did svn revert :-x
10:45:30 <Gekz> YOU SHOULDNT BE PLAYING AT WORK
10:45:48 <SmatZ> openttd can run as dedicated server :)
10:45:59 <SmatZ> TiberiusTeng: it doesn't even start compiling on linux :(
10:46:36 <TiberiusTeng> it'll need a linux (perhaps X) opengl driver to run on it ...
10:48:15 <TiberiusTeng> hey wait a sec, you'll need a videodriver to do OpenGL initialization under Linux!
10:48:22 <TiberiusTeng> or if you want to write one :D
10:49:47 <SmatZ> I have no chance doing that :)
10:50:35 <SmatZ> mmm I am still missing glu.h
10:51:13 <TiberiusTeng> I remember peter1138 gave me his OpenGL work for SDL earlier ... maybe he knows how to setup OpenGL environment using SDL. But I'm not familiar with SDL :(
10:51:42 <TiberiusTeng> Linux doesn't have glu.h ?
10:51:50 <SmatZ> TiberiusTeng: great, now it starts under wine :)
10:52:06 <SmatZ> TiberiusTeng: maybe, maybe in different directory, I don't know
10:52:18 <TiberiusTeng> wait, it works under WINE ?
10:52:24 <TiberiusTeng> with opengl blitter enabled ?!
10:52:33 <SmatZ> TiberiusTeng: ha, "GL/glu.h" and "GLee.h"
10:52:49 <TiberiusTeng> aha. I'll modify it then.
10:52:57 <SmatZ> dbg: [GL] Grow Spritebuffer capacity to 32768
10:53:30 <TiberiusTeng> could you gimme a screenshot? :D~
10:55:03 <SmatZ> hehe you don't make animated pixels transparent, so I think it is the opengl blitter
10:55:43 <TiberiusTeng> ahh, it's technically very difficult to do ...
10:55:55 <TiberiusTeng> I'm using the same method as 32bpp-anim blitter.
10:57:12 *** Brianetta has joined #openttd
10:58:29 *** [alt]buster has joined #openttd
10:58:38 <SmatZ> TiberiusTeng: don't have hard work with supporting animated pixels with transparency
10:58:55 <SmatZ> like, animated pixel behind transparent building or transparent animated pixels
10:59:20 <SmatZ> looks the same as on windows though :)
11:00:13 <TiberiusTeng> ahh, it works :D
11:00:33 <SmatZ> yeah, and yesterday it wasn't working :)
11:00:43 *** Xeryus|bnc is now known as XeryusTC
11:04:42 *** [alt]buster is now known as [com]buster
11:05:11 <SmatZ> TiberiusTeng: 64bit related problems:
11:05:13 <SmatZ> src/blitter/opengl.cpp:1598: error: cast from 'const byte*' to 'int' loses precision
11:05:59 <ln> does that kind of a cast make sense anyway?
11:06:51 <TiberiusTeng> ok, I'll check it
11:06:53 *** Yexo is now known as Guest1610
11:09:13 <yorick> didn't know stations would actually get the ratings before something arrived on them in TTO
11:11:35 *** Artamir has joined #openttd
11:12:02 <Artamir> I have problems with Tortoise SVN with .diff
11:12:14 <Artamir> I select "Apply Patch"
11:12:28 <Artamir> And there are error messages
11:13:29 <Artamir> "There are still unresolved conflicts at line 1691"
11:13:51 <yorick> have you applied any patches before?
11:14:09 <yorick> have you also reverted them when updating?
11:14:10 <Artamir> only patches without .diff and .patch
11:15:32 <yorick> possible the area YAPP changes got updated too
11:15:38 <TiberiusTeng> SmatZ, is there a good way to solve the 'format not a string literal' warning ?
11:15:51 <yorick> and tortoise can't find a way to make the changes there have been in trunk
11:16:01 <yorick> then it will say conflict
11:16:49 <Artamir> if the folder is patched yet tortoise can not?
11:17:06 <yorick> if there have been any updates on the server, tortoise will try to update it too
11:17:29 <yorick> look into the file it errors on
11:18:11 <SmatZ> TiberiusTeng: probably not
11:18:23 <yorick> could someone tell him how to resolve conflicts with SVN while I have lunch, thank you :)
11:19:10 <ln> look for a line containing ".mine", then resolve the conflict by editing the file near that line, removing unnecessary things.
11:19:32 <Artamir> I have not idea of C++ or this commands :S
11:20:40 <Artamir> there are lines with "????????????????????????????????"
11:23:42 <blathijs> Artamir: If you did not modify the source yourself, other than applying the YAPP patch, you should just revert all changes (or make a new checkout) and apply the patch on that
11:24:35 <SmatZ> and make sure you have downloaded correct SVN revision
11:24:45 <SmatZ> newest YAPP appliest to r13612
11:24:53 <blathijs> Artamir: And if applying the YAPP patch on a clean checkout does not work, your checkout is probably to new. You should then either get a newer patch or use the "svn update" command to get the revision the patch was made for
11:25:20 <Rubidium> that for sure isn't a clear checkout
11:26:07 <ln> i suppose Tortoise has an option to update a clean copy
11:26:07 <SmatZ> and really outdated, it uses old patch structure
11:26:11 <blathijs> Artamir: You can't just apply patches and expect them to work and keep working when you update
11:26:53 <blathijs> Artamir: You should probably start over (Using the "Revert" feature in tortoise) and then apply the patch(es) you want
11:27:46 <Artamir> "file list it's empty"
11:28:05 <blathijs> The easier way is just throw away your checkout and make a new one
11:28:12 <blathijs> (or just make a new one beside the old one)
11:31:11 <SmatZ> Artamir: using custom patches and combining them needs more or less experience and programming knowledge... when youtry several times, you will learn how to...
11:31:17 *** Artamir has joined #openttd
11:31:35 <SmatZ> Artamir: using custom patches and combining them needs more or less experience and programming knowledge... when you try several times, you will learn how to do anything you want :)
11:32:05 <Artamir> I have not experience
11:32:13 <SmatZ> but custom patches are not supported by developers, just because of problems you have...
11:32:27 <SmatZ> so I can tell you only one thing - if you don't want problems, use clean trunk :)
11:33:04 <SmatZ> there are several "patch packs" at tt-forums, but you will most likely have problems with them
11:33:18 <SmatZ> I don't know, it depends what you have in that folder
11:33:38 <SmatZ> o checkout to a new directory
11:34:00 <SmatZ> but nightly.openttd.org is safe place to download win binaries
11:34:33 <Artamir> I'm checking out in a new folder
11:34:49 <blathijs> That's a good start :-)
11:35:31 <SmatZ> try it, if it fails, then it wasn't the best thing to do :-P
11:36:21 <blathijs> Artamir: Yup, and if it fails you can try to change the revision of your clean trunk to match the revision of the patch
11:36:40 <SmatZ> the patch is outdated, you have to use older trunk
11:36:47 <SmatZ> but for YAPP, you need newer trunk
11:37:04 <SmatZ> you won't make those two patches work together without further coding
11:37:15 <SmatZ> ask patch developer to update it :)
11:37:45 <Artamir> I only wanted to apply the "speedlimit" patch
11:37:47 <CIA-6> OpenTTD: truebrain * r13620 /branches/noai/ (2 files in 2 dirs): [NoAI] -Fix: foreach() on empty lists gave weird error (bug by Yexo)
11:38:47 <SmatZ> Artamir: where did you download that patch?
11:39:26 <Artamir> I haven't got the link here
11:39:37 <Artamir> SpeedLimitsWIP_r10115v2
11:39:51 <SmatZ> so checkout revision r10115
11:40:31 *** Yexo is now known as Guest1615
11:40:32 <blathijs> Artamir: Use "update to revision" in Tortoise
11:41:06 <Yexo> something is wrong with my connection today
11:42:16 <CIA-6> OpenTTD: rubidium * r13621 /trunk/src/road_cmd.cpp:
11:42:16 <CIA-6> OpenTTD: -Fix: building roadbits in the wrong direction on bridges or building roadbits
11:42:16 <CIA-6> OpenTTD: from underneath the bridge to the bridgehead is impossible, so don't silently
11:42:16 <CIA-6> OpenTTD: ignore that error when building over houses and industries is not ignored.
11:44:17 *** dR3x4cK has joined #openttd
11:45:24 <ln> chaleco salvavidas debajo de su asiento
11:46:20 <ln> yorick: what if it doesn't sound as good in english?
11:48:00 <SmatZ> "life jacket is under his seat"?
11:49:43 <Artamir> (Why don't you shut up)
11:50:31 <ln> SmatZ: ever been in a spanish airplane?
11:52:17 <SmatZ> ah, isn't "su" more "official"?
11:52:32 <ln> SmatZ: tu/su is like german du/Sie, i've been told.
11:52:56 <Artamir> "tu chaqueta" = "your jacket"
11:53:04 <Artamir> "su chaqueta" = "his jacket"
11:53:28 <Artamir> "su" acts like "your"
11:53:40 <SmatZ> "usted" -> "su" ; "tu" -> "tu"
11:53:56 <ln> Artamir: what would be "here is Your jacket, sir"?
11:53:59 * SmatZ found his notebook from Spanish classes :)
11:54:55 <Artamir> "Aqui esta su chaqueta, caballero"
11:55:07 <Artamir> sir = "caballero", "seor"..
11:55:23 <Artamir> (caballero it's for men=
11:55:52 <yorick> how far goes the openttd support for neutral stations?
11:55:59 <yorick> I know oil rigs can have one
11:56:07 <SmatZ> yorick: from oil rigs to buoys :-P
11:56:30 <yorick> only ships and aircraft?
11:56:42 <SmatZ> I don't know about any newgrf support
11:56:49 <ln> airports should definitely be "neutral".
11:57:32 * yorick hacks airport toolbar into scenario editor
12:00:17 *** [com]buster has joined #openttd
12:01:26 <ln> yorick: does it work that easily?
12:04:05 <yorick> ok, it's most likely a twoliner
12:04:58 *** GoneWacko has joined #openttd
12:07:01 <yorick> I got the toolbar+gui, but it doesn't really like neutral airports so far
12:10:37 <yorick> adding a town might help tho
12:14:19 <Noldo> yorick: what are you doing?
12:14:49 <yorick> hacking airports into the scenario editor
12:15:01 <yorick> just to prove ln it is possible
12:17:01 <yorick> just I don't think that it'll work without tracksharing
12:17:26 <yorick> but with tracksharing I just ran a plane into that airport
12:17:46 <yorick> but maybe...the oilrig support and so
12:18:57 <yorick> you can't build anything without the tracksharing patch, I think
12:26:56 *** TiberiusTeng has joined #openttd
12:35:59 *** Dred_furst has joined #openttd
12:41:00 *** dragonhorseboy has joined #openttd
12:43:25 <dragonhorseboy> doing okay yorick
12:43:33 <dragonhorseboy> just trying to figure out car tires (for real) .. go figure :p
12:43:56 <dragonhorseboy> otherwise not doing much
12:44:30 <fjb> Tires are usually round und you usually need at least four for a car.
12:44:42 <yorick> and trucks usually have about 8
12:45:01 <dragonhorseboy> what you trying to figure out out of curiousity?
12:45:14 <dragonhorseboy> fjb....I meant a lot more than that :p
12:45:44 <yorick> if players can use neutral airports without the tracksharing patch
12:46:27 <dragonhorseboy> fjb I'm curious about if its possible to find something that can handle slight muddy access roads while still being fair enough for typical 110km/h highways altogether
12:47:24 <dragonhorseboy> yorick ah hmm good luck I'll like to hear how it goes ^-^
12:48:25 <dragonhorseboy> btw yorick interesting waterlevel island there either way :p
12:48:41 *** dR3x4cK has joined #openttd
12:48:58 <Gekz> that's something I've considered
12:49:11 <Gekz> you could literally thunderpwn it with planes
12:49:57 <SmatZ> stop one airport at runway and block the airport forever
12:50:11 <Gekz> this game could do with a cleanup script
12:50:28 <dragonhorseboy> smatz...thats breaking the rule on many servers ;)
12:50:30 <Gekz> "Truck crushed due to inactivity"
12:50:31 <dragonhorseboy> unless its one of these deathmatch one (which I've never seen for a long time oddly enough)
12:50:36 <yorick> someone suggested a talking paperclip called Tee Tee
12:50:46 <yorick> scanning the game for such things
12:51:00 <SmatZ> dragonhorseboy: hehe, you are right :)
12:51:36 <dragonhorseboy> and which of I'm just curious about it anyhow but....
12:51:56 <dragonhorseboy> is there actually anything happening toward more custom airports or its still dead for the time being yet?
12:52:05 <yorick> I don't like talking paperclips watching me or whatever I'm doing
12:52:22 <yorick> and even falling asleep if you don't talk to it
12:52:23 <fjb> I don't like talking paperclips at all.
12:52:55 <SmatZ> I don't like talking abything at my computer
12:53:09 <SmatZ> paperclip, any stupid asistant
12:53:27 <SmatZ> that "What do you want to search today?" doggie in Win XP
12:53:40 <yorick> but I noticed that TTO has a tutorial mode
12:53:52 <yorick> showing you how to build a railway
12:54:09 <SmatZ> yorick: wiki.openttd.org
12:54:24 <TiberiusTeng> ahh ... scaling problem. this will make me headache ... but not now ;)
12:54:34 <yorick> that's no ingame tutorial mode with self-moving-mouse
12:54:41 <fjb> Hm, why are road vehicles not refitable to any cargo when the first part can not carry any cargo? Is that because the test in line 2056 of roadveh_cmd.cpp?
12:55:22 <fjb> SmatZ: Thank you. I will look at that.
12:55:39 <dragonhorseboy> I guess its dead then? (re airports)
12:56:19 <glx> dragonhorseboy: what do you mean with custom airports?
12:56:31 <SmatZ> dragonhorseboy: for me, yes
12:56:57 <SmatZ> maybe if ever infrastructure sharing will be implemented
12:57:45 <yorick> the neutral airports work without tracksharing
12:57:51 <yorick> just you can't use their depots
12:57:56 <yorick> and you can with tracksharing
12:58:05 *** dR3x4cK2313 has joined #openttd
12:58:16 <dragonhorseboy> glx hmm let me find one thread to show you one moment
12:58:37 <yorick> I think he means newgrfports
13:00:47 <fjb> SmatZ: So it is possible to refit the vehicle when the grf states the possible cargos even if the amount is 0?
13:01:20 <SmatZ> fjb: I don't know, it shouldn't be
13:02:15 <SmatZ> maybe Rubidium, he replied in that FS, probably he knows the most :)
13:03:03 *** Osai^zZz`off is now known as Osai
13:03:18 <dragonhorseboy> hope that helped for now glx
13:20:01 *** Progman has joined #openttd
13:22:26 *** [com]buster has joined #openttd
13:22:35 *** dR3x4cK has joined #openttd
13:26:00 <dragonhorseboy> hm I'll just assume dead then and troubleshoot the diff errors myself :/
13:28:58 <flowOver> so hello. i haven't coded c++ for the better part of a decade and all i've done was a very succesful tetris clone with oop as a HS final proj. i've peddled with php since then. any room for somone of my experience to help?
13:41:23 <glx> dragonhorseboy: it's still WIP
13:42:27 <dragonhorseboy> guess I'll just ignore the minor passable bugs then in case it already gets fixed later on
13:42:28 <dragonhorseboy> thanks anyhow glx
13:45:28 <dragonhorseboy> glx if you don't mind me asking you anyhow...
13:46:06 <dragonhorseboy> what did you think of the diagonal road crossing concept?
13:47:41 <CIA-6> OpenTTD: rubidium * r13622 /trunk/src/vehicle_gui.cpp: -Fix [FS#2109]: only the front of a RV would be considered when determining to what cargos a vehicle can be refitted instead of all cargos.
13:49:25 <CIA-6> OpenTTD: rubidium * r13623 /trunk/src/ (roadveh_gui.cpp vehicle_gui.cpp): -Codechange: only show the 'load' of articulated RV parts that have a capacity instead of showing 'empty' parts when a vehicle is completely full.
13:55:07 <ln> A fix is not a codechange?
13:55:35 <ln> What is a commit by default?
13:56:01 <SmatZ> ln: you have too many questions
13:56:30 <dragonhorseboy> glx one of the airport I kinda like for some reason is that thin&long one (two runaways almost lined up with each others) ... kinda fun to place that on a wide cliff or so where it would had otherwise required way too much trees destorying to ever build even a commuter/city airport next to the very town
13:56:34 <TiberiusTeng> SmatZ, which display chard are you using ?
13:56:53 <SmatZ> TiberiusTeng: GF8600GT or so
13:57:04 <TiberiusTeng> ok, so it's also an NV card ...
13:57:08 <dragonhorseboy> but of course if its still a current WIP I'll just stay quiet and wait for till non-diff players finally can play it as well
13:57:43 <TiberiusTeng> bokkie posted in the forum said X1950PRO only gives a black screen ...
13:58:09 * dragonhorseboy has one nvidia and one 3dfx .. not counting onboard gpu of laptop & other desktop yet
13:59:07 <dragonhorseboy> (bonus if you even actually knew what 3dfx was heh)
13:59:23 <TiberiusTeng> don't even think about running it with Voodoo 1. :p
13:59:34 <TiberiusTeng> Glide is ... well, GL lite
14:02:30 <dragonhorseboy> gekz...yeah if you can find some actual screenshots you'll be surprised at how many games actually looked better on glide than at-the-time directx ;)
14:02:43 <dragonhorseboy> descent2 and tomb raider especially
14:03:13 <dragonhorseboy> and tiberius... voodoo1's are difficult to find in non-shelf-displaying condition .. seem there's something with their age that is making many of them not work well anymore
14:03:19 *** ChanServ sets mode: +o orudge
14:03:32 <dragonhorseboy> but voodoo2's still works well if you want to use them tho for some reason
14:09:16 <dragonhorseboy> gekz and tiberius if you two don't mind me asking -- what graphic card(s) you have now?
14:09:47 <dragonhorseboy> gekz no agp slots?
14:09:54 <Gekz> dragonhorseboy: none, in any computer
14:10:33 <dragonhorseboy> gekz..wait...are you sure that voodoo3 is onboard?
14:10:45 <Gekz> dragonhorseboy: google "Bora Pro"
14:12:56 <dragonhorseboy> damn...where you located?
14:14:08 <dragonhorseboy> hrm gekz not to sound crazy but do you still use that particular board at time?
14:15:19 <Gekz> you can buy it should you want it
14:17:01 *** Lakie` is now known as Lakie
14:18:09 <dragonhorseboy> gekz hmm well does it have any cards installed or all pci/isa slots are empty?
14:18:45 <Gekz> it has a Geforce 64mb PCI
14:19:10 <Gekz> processor can run with no fan
14:20:53 <dragonhorseboy> sounds like you have a 233-400mhz one then
14:21:05 <dragonhorseboy> for these are easily fanless (even although they keep coming with fans attached for some reason)
14:21:32 <dragonhorseboy> either way gekz would you be willing to ship it like motherboard+ram only if I asked? (if you wanted the ram, keep them then)
14:24:20 <Gekz> dragonhorseboy: the ram is shit, so ok
14:24:26 <Gekz> how much are you talking here?
14:26:09 <dragonhorseboy> well I thought that the board alone couldn't be too big an inside-wrapped box anyhow (and the ram can be left in the dimm slots since they're too small to cause much inertia) ;)
14:26:18 <dragonhorseboy> how much? hmmm good question .. do you know how much ram its got?
14:26:35 <Gekz> a 128MB stick and a 64MB stick
14:27:03 <dragonhorseboy> hm is that at least pc100/133?
14:27:16 <dragonhorseboy> ah hm 66mhz .. not much worth
14:27:33 <dragonhorseboy> do you perhaps know which one was the pc66 one?
14:27:37 <Eddi|zuHause3> man... i have to improve my mirror-reading skills
14:27:47 <Gekz> excuse my spelling, my r key is broken
14:28:18 <dragonhorseboy> so....which one? :P
14:28:22 <Eddi|zuHause3> i just drove on the road, behind me was a truck, and the country code looked like "JP", and i was like "wtf? a truck from japan?"
14:28:37 <Eddi|zuHause3> then i realized that i was looking through a mirror, and it actually said "PL"
14:29:40 <dragonhorseboy> gekz...hmm tell you what how about discard that 64mb one .. I could take the board + 128mb pc100 for hmm perhaps $20aud+shipping?
14:29:44 <dragonhorseboy> just a first offering
14:30:00 <Gekz> that offer makes me want to cry
14:30:23 <Gekz> I reckon shipping would be more expensive than anything else
14:30:47 <Gekz> shipping is expensive -_-
14:31:19 <dragonhorseboy> whats your postal code?
14:32:05 <Eddi|zuHause3> you should know that trucks from poland are not that uncommon in germany
14:34:59 <glx> like SP and NL in france ;)
14:35:11 <dragonhorseboy> gekz well I'm sure that its not expensive to ocean-ship it considering australia is straight across ocean from west canada :p
14:35:30 <Gekz> dragonhorseboy: you'd be surprised lol
14:35:42 <TiberiusTeng> those good old days ... TTO on a 486 + 8MB RAM
14:36:13 <dragonhorseboy> gekz...what? :p
14:36:26 <dragonhorseboy> gekz mind pm for a moment?
14:36:56 <dragonhorseboy> tiberius..well ttdx could run on that too and I think you could get away with using the patch as well (just might have to add more ram and/or use -lowmemory flag I think)
14:37:07 <Rubidium> TiberiusTeng: OTTD runs on a Nintendo DS, which has a whopping 4 MB of RAM (which includes the binary)
14:37:42 <Rubidium> okay, it is dual processor
14:38:04 <Rubidium> arm7 @ 33 MHz and arm9 @ 67 MHz
14:38:36 <Gekz> Rubidium: does it have a kernel?
14:38:54 <Gekz> so how much workable ram do you have that you know of
14:39:05 *** ChanServ sets mode: +v tokai
14:39:20 <glx> memory access in DS are weird
14:39:56 <glx> IIRC some part of memory can be accessed 16bits by 16bits
14:40:39 <glx> but ask dominik if you have more questions about the DS
14:48:05 <Eddi|zuHause3> <SmatZ> "usted" -> "su" ; "tu" -> "tu" <- i learned that in spain practically nobody uses the "usted" form, only in really formal meetings, opposed to german, where you would say "Sie" to any stranger >14 years
14:48:19 <yorick> why does GetValidAirports have a 32 bit mask where only 9 bits are used?
14:49:10 <glx> yorick: think about future
14:49:24 <yorick> think about expanding it later then
14:49:37 <Rubidium> because C doesn't know a 9 bits datastructure?
14:50:04 <Eddi|zuHause3> which it silently extends to 32bit for alignment purposes anyway
14:50:23 <ln> what modern architecture supports 16-bit integers? ...
14:50:27 <yorick> I don't care about that one
14:50:32 <Eddi|zuHause3> or 64bit, depending on architecture
14:51:21 <Eddi|zuHause3> which are 9x32bit?
14:51:37 <yorick> afaik, a boolean is 1 bit?
14:51:37 <Rubidium> and uint32 takes less finger movement than uint16
14:51:47 <SmatZ> Rubidium: bitfield? int a:9; ?
14:52:00 <Eddi|zuHause3> yorick: in your dreams
14:52:30 <Rubidium> and int a:9 is still 32 bits
14:52:47 <Rubidium> just bits 9-31 can contain garbage
14:53:02 <yorick> at least it can be useful garbage
14:53:08 <ln> yorick: you know, boolean algebras exist for domains greather than just {0,1}. (which has nothing to do with this subject.)
14:53:27 <yorick> I don't care about that
14:53:46 <yorick> I don't care about booleans that can be both 1 and 0 at the time too
14:53:52 <Eddi|zuHause3> yorick: ALL variables are aligned to 32/64 bit
14:54:01 <Eddi|zuHause3> even 1 bit "booleans"
14:54:17 <Eddi|zuHause3> only exception are packed structs
14:54:30 <yorick> 9 booleans in a packed struct?
14:54:55 <Eddi|zuHause3> yorick: it's not stupid, accessing an unaligned variables take several magnitudes longer to access
14:55:15 <TiberiusTeng> SGIS_generate_mipmap ... zooming out case closed heh
14:55:16 *** EightBot has joined #openttd
14:55:58 <EightBot> is there a way of preventing industries fomr closing down? e.g. is there a setting for that?
14:56:00 <SmatZ> Eddi|zuHause3: not true
14:56:20 <Eddi|zuHause3> like "get 32-bit word from ram to register 1; generate imm32 mask in register 2; AND register 1 and 2 into register 1; shift register 1"
14:56:35 <SmatZ> variables are usually aligned in such a way that the alignment as the same as variable lenght
14:56:46 <EightBot> i find it annoying that so many industries close down
14:56:46 <SmatZ> uint16 is aligned at 2B
14:57:10 <Eddi|zuHause3> that's 4 instructions where 1 would suffice for 32-bit alignment
14:57:12 <TiberiusTeng> well not that easy >_<
14:57:16 <yorick> EightBot: probably you can patch it up
14:57:27 <Eddi|zuHause3> SmatZ: you mean left alignment vs. right alignment?
14:57:38 <Eddi|zuHause3> SmatZ: which would differ at big/little endian?
14:57:59 <SmatZ> I doubt it would differ, do you have any example?
14:58:40 <SmatZ> what does endianness to do with alignment?
14:59:29 <Eddi|zuHause3> on little-endian, you can interpret a 32-bit register as 16-bit by ignoring the right 16 bits
14:59:45 *** ChanServ sets mode: +v tokai
14:59:49 <Eddi|zuHause3> SmatZ: possibly i just misunderstood what you meant
15:00:04 *** dragonhorseboy has left #openttd
15:00:20 <SmatZ> Eddi|zuHause3: probably, my English isn't that good :-)
15:01:08 <Eddi|zuHause3> mainly i meant that big and little endians must ignore different sets of bits
15:01:49 <SmatZ> still, you can take "lower" 2B from 4B variable by reading bytes at address +2
15:02:36 <EightBot> is there a way of preventing industries fomr closing down?
15:02:44 <Eddi|zuHause3> that looks awfully platform specific to me
15:03:29 <glx> I remember nice bugs due to boolean being 'different' on BE and LE
15:03:36 <SmatZ> will be places at offsets 0,2
15:03:39 <Eddi|zuHause3> not only big/little endian problems
15:03:43 <SmatZ> on both BE and LE machines
15:03:52 <Eddi|zuHause3> but also from which side the addresses "grow"
15:04:59 <Eddi|zuHause3> SmatZ: i would never trust this kind of pointer arithmetics
15:05:14 <Eddi|zuHause3> but i'm a too high level computer scientist for that :p
15:05:43 <SmatZ> you have given ABI, and all compilers have to keep the same alignment rules in structs
15:06:12 <SmatZ> but sure, if ABI says that all variables have to be 64bit aligned, then everything will be aligned in that way
15:06:52 <SmatZ> glx: savegame problems :)
15:06:57 <Eddi|zuHause3> yeah, of course you can rely on the ABI, but it'll be a porting nightmare once the ABI changes
15:07:37 <SmatZ> I am not saying you should rely on these rules
15:07:55 <SmatZ> [16:54:02] <Eddi|zuHause3> yorick: ALL variables are aligned to 32/64 bit <== but this is not true
15:08:31 <Eddi|zuHause3> and i am saying i _would_ not rely on it
15:22:51 *** Kommer- is now known as Kommer
15:31:21 <dih> # na na na na na na na hey hey hey
15:32:39 <SmatZ> I don't watch Euro anymore :-/
15:39:20 <Doorslammer> *cheers for the Netherlands, believing an upset is in order :P
15:39:26 <Doorslammer> Stranger things have happened
15:40:13 <Doorslammer> If Max Mosley ever run football...
15:40:33 <Doorslammer> Everyone would get a 10 second penalty and the Italians would always win
15:58:25 <CIA-6> OpenTTD: rubidium * r13624 /trunk/src/terraform_gui.cpp: -Fix [FS#2106]: crash when one tries to raise the nothern corner of MP_VOID tiles (i.e. the southern corner of the tiles on the southern map edge) in the scenario editor.
16:02:49 *** ChanServ sets mode: +o Bjarni
16:04:10 <yorick> Bjarni: I have some troubles with autoreplace trying to replace an aircraft that's on a neutral hangar
16:04:24 <yorick> lets say it gives you the money, but never takes it back
16:04:43 <yorick> while you can't build anything there
16:05:36 <yorick> problem is that the bug is'nt technically in trunk
16:05:48 <yorick> because one can't build neutral hangars
16:06:12 <Bjarni> I was thinking "did anybody commit that without me noticing it" :)
16:06:34 <Bjarni> you shouldn't order me around
16:07:23 *** GoneWacko has joined #openttd
16:12:13 *** einKarl has joined #openttd
16:17:28 <flowOver> am i insane to be crawling through the code with eclipse?
16:17:48 <TiberiusTeng> try doing this with grep/find/vim
16:22:56 <Eddi|zuHause3> <yorick> Bjarni: I have some troubles with autoreplace trying to replace an aircraft that's on a neutral hangar <-- add an assert ;)
16:25:06 <Bjarni> Eddi|zuHause3: he used a modified source. It will not happen in trunk
16:25:28 <Eddi|zuHause3> Bjarni: yes, which is why you should add an assert instead of a "fix"
16:25:44 <Eddi|zuHause3> in case it is at any future time possible to have neutral hangars ;)
16:26:18 <ccfreak2k> TiberiusTeng, around?
16:26:20 <Bjarni> if that happens odds are that I already finished my current work on autoreplace
16:26:29 <Bjarni> in which case the assert would no longer be needed
16:26:52 <Eddi|zuHause3> i mean that is exactly what asserts are useful for
16:27:00 <Eddi|zuHause3> alerting people of preconditions
16:30:30 <TiberiusTeng> I don't have ATI cards around so probably can't help
16:30:52 <TiberiusTeng> hmm ... I think the ATI driver doesn't like my shader source :p
16:30:58 <ccfreak2k> Also, I noticed you use EXT_texture_non_power_of_two when available.
16:31:16 <TiberiusTeng> it makes almost no difference for now
16:32:15 <TiberiusTeng> but lacking the shader could result in black screen, yes
16:32:43 <TiberiusTeng> which card are you using?
16:32:54 <ccfreak2k> ATi Radeon X850 XT PE.
16:39:00 <TiberiusTeng> ccfreak2k, can you compile the patch yourself ?
16:39:10 <ccfreak2k> That's how I got it to work.
16:39:35 <ccfreak2k> Well, "work" as in "didn't crash on startup".
16:40:08 <TiberiusTeng> I'm editing opengl.cpp to make it display fragment program error string, hold on ...
16:45:41 <TiberiusTeng> replace blitter/opengl.cpp with this one, compile, and see what message it displays after startup
16:48:03 <ccfreak2k> Compiling threw a bunch of errors.
16:51:48 <TiberiusTeng> are you using my latest patch ?
16:52:01 <ccfreak2k> I used the one you linked to last night, plus some of glx's changes.
16:52:04 <TiberiusTeng> the .hpp seems mismatch
16:52:32 <TiberiusTeng> please use this one instead
16:54:07 <TiberiusTeng> no, not that one, check the thread
16:54:21 <TiberiusTeng> opengl-080624_r13612.patch
16:57:04 *** frosch123 has joined #openttd
17:05:50 *** Swallow has joined #openttd
17:08:35 <ccfreak2k> TiberiusTeng, you have the DirectX SDK installed, right?
17:09:14 <Eddi|zuHause3> no, he removed that part
17:09:33 <ccfreak2k> Well I get dmksctrl.h missing from dmusic.cpp.
17:10:16 <Eddi|zuHause3> i can't say anything about that...
17:12:16 <glx> ccfreak2k: did you remove DIRECT_MUSIC define in project?
17:13:16 <ccfreak2k> Wait no it's still there.
17:15:28 <TiberiusTeng> I didn't touch that part in the new patch, but I don't have DirectX SDK installed either ...
17:17:45 <TiberiusTeng> ahh change the header prototype
17:17:57 <TiberiusTeng> I was experimenting some mipmap work
17:18:50 <ccfreak2k> Ok where am I changing it.
17:19:11 <ccfreak2k> It appears once in the header: GLuint CreateTexture2D();
17:19:22 <TiberiusTeng> GLuint CreateTexture2D(bool interpolate = false);
17:19:40 <TiberiusTeng> find CreateTexture2D in the .hpp and replace with this one
17:20:30 <ccfreak2k> I have to add the dependencies to the vcproj AGAIN.
17:27:37 <glx> will be easier to install DX ;)
17:28:27 *** Scaevolus has joined #openttd
17:29:38 <Scaevolus> Is there a more extensive "Cargo Payment Rates" chart? Or does it just scale linearly up with distance?
17:30:12 <Swallow> it scales linearly with distance, yes
17:30:39 <Scaevolus> when it measure distance, is that station distance, or distance between the resources?
17:30:52 <ccfreak2k> TiberiusTeng, do I still need to pass the -v switch?
17:31:01 <Swallow> I think manhattan distance between statinos
17:31:34 <Swallow> you may be interested in this patch... *searches*
17:32:20 <TiberiusTeng> glx, what's the advantage of dmusic driver anyway ?
17:32:30 <TiberiusTeng> the roland wavetable of dmusic ?
17:33:32 <TiberiusTeng> cargo payment formulas here
17:33:57 <Scaevolus> Swallow: I guess I'll just dig around in the source to find the algorithms
17:35:23 <TiberiusTeng> Error string: line 2: invalid option name
17:35:52 <TiberiusTeng> remove the printf("(%3d, %3d)... in DoEncode
17:36:08 <TiberiusTeng> and comment out the line "OPTION ARB_precision_hint_fastest;"
17:36:27 <TiberiusTeng> there are four such OPTION's
17:36:35 <TiberiusTeng> goddamn it's in specifications
17:40:01 <ccfreak2k> I'm not an OpenGL programmer. D:
17:40:12 <Swallow> I have a coding question regarding a patch I'm working on...
17:40:22 <Swallow> I have some arrays that have to be accessible from DrawOrderString() and the OrdersWindow struct...
17:40:29 <Swallow> where should I define them?
17:43:06 <Scaevolus> what revision is svn at?
17:46:11 <dih> Swallow: what files are those functions in?
17:46:35 <Swallow> both in order_gui.cpp
17:46:58 *** daspork has joined #openttd
17:47:34 <dih> why dont you use another variable as an example?
17:48:17 <TiberiusTeng> ahh ... the new graph window ...
17:48:21 <TiberiusTeng> perhaps I should clean up
17:48:40 <TiberiusTeng> but I have a report to do 12 hours later >_<
17:49:27 <dih> you only need the variable in that file?
17:50:58 <Swallow> Should I just define it there?
17:51:19 <dih> have a look at network/network.cpp
17:51:30 <dih> (first file that comes to mind)
17:52:30 <Swallow> k thanks, i'll look at it
17:52:54 <TiberiusTeng> what, you commented out OPTION ARB_precision_hint_fastest and it still complains for OPTION ARB_draw_buffers ?!
17:53:25 <yorick> Bjarni: I only used a modified source to generate the map, but it's normally playable on trunk, EXCEPT FOR THAT AUTOREPLACE OF YOURS
17:53:32 <ccfreak2k> / "OPTION ARB_precision_hint_fastest;"
17:53:36 <ccfreak2k> // "OPTION ARB_precision_hint_fastest;"
17:53:46 <ccfreak2k> And I commented out the printf() line.
17:54:45 <ccfreak2k> So yes my guess is that it claims ARB_draw_buffers is an invalid option.
17:54:50 *** Pikka is now known as Pikka|afk
17:55:15 <TiberiusTeng> ARB_draw_buffers was originally ATI_draw_buffers! and the card claims supporting it in PostDriverInit() ...
17:55:37 <ccfreak2k> I'm gonna run ext-checker just in case.
17:55:48 <TiberiusTeng> how about changing ARB_draw_buffers to ATI_draw_buffers ?
17:55:58 <Scaevolus> is there a commandline utility to extract metadata from grf files?
17:56:17 <TiberiusTeng> if this still fails ... I would suggest you go get an GeForce 6100 or such, I believe it's only ~usd $30 for now
17:56:27 <TiberiusTeng> and ditch that ATI one
17:57:45 <ccfreak2k> Plus I don't get a black screen anymore, although some pixels are a bit messed up.
17:57:49 <yorick> Bjarni: I think CmdBuildAircraft doesn't technically fail without a valid tile
17:57:56 <TiberiusTeng> ahh. passed. great. !@#$%^&*()) ATI
17:58:16 <TiberiusTeng> messed pixels ? how about a screenshot ?
17:58:18 <ccfreak2k> Gotta love those vendor-specific extensions.
17:58:44 <TiberiusTeng> I wonder if you're using a very-old driver?
17:59:00 <ccfreak2k> If by "very old" you mean "two months old max", no.
17:59:00 <TiberiusTeng> ATI_draw_buffers got promoted to OpenGL ARB extensions I believe ...
18:00:50 <ccfreak2k> EVEREST shows ARB_draw_buffers as supported.
18:01:06 *** Osai is now known as Osai`off
18:01:18 <TiberiusTeng> yes, the card claims it does support, but it don't like OPTION ARB_draw_buffers; in the shader
18:01:35 <TiberiusTeng> which is defined in ARB_draw_buffers standard ....
18:04:07 <TiberiusTeng> so ... ccfreak2k, you're playing it right now? :p
18:04:15 <ccfreak2k> I'm uploading the screenshot.
18:05:06 <Scaevolus> opengl rendering openttd?
18:05:24 <ccfreak2k> The pixels that are "out of place" will change, as if they were in some animated palette or if they were transparent.
18:05:51 <TiberiusTeng> ahh. I think ATI have unique ways of interpreting texture coordinates ...
18:06:07 *** gregor_ has joined #openttd
18:06:12 <Scaevolus> are you using the 2d acceleration bits of opengl?
18:06:37 <TiberiusTeng> drawing quads on a flat surface ...
18:06:44 <Scaevolus> oh, well there's your problem
18:06:48 <gregor_> is there any roadmap, what you want to include to openttd?
18:06:54 <TiberiusTeng> orthgonally projected on screen
18:07:25 <Scaevolus> should really be using 2d, not quads
18:07:40 <TiberiusTeng> ccfreak2k, you can try to mess with PARAM constants, and TexCoordScale
18:07:56 <ccfreak2k> I can, but I'd probably break something in the process.
18:07:59 <TiberiusTeng> the 2d pipeline is actually MUCH slower on my card
18:08:18 <Scaevolus> ...but isn't this just a replacement for the blitter?
18:08:52 <TiberiusTeng> the commented (val + 0.5f) / total ... and changing PARAM constants = { 1, 0.00195, 0, 0 }; ... something like that
18:09:04 <TiberiusTeng> it is a blitter ...
18:10:14 <TiberiusTeng> you can try yourself, glDrawPixels - glReadPixels - glDrawPixels again is much more slower than drawing a texture-mapped quad, glTexSubImage2D, then draw another quad ...
18:11:01 <Scaevolus> sure, the 3d pipeline is more optimized
18:11:13 <Scaevolus> but I doubt you would be pushing the 2d pipeline anyways
18:11:50 <TiberiusTeng> it was something like 15fps vs 60fps in my experiment (I forgot to turn off VSync then) ...
18:12:05 <TiberiusTeng> so I sticked to 3D pipeline and did some batching
18:12:37 <TiberiusTeng> anyway now I know it's always a bad idea to support ATI cards ...
18:13:10 *** stillunknown has joined #openttd
18:13:20 <TiberiusTeng> ah. ccfreak2k, check CreateTexture2D() in opengl.cpp
18:13:43 <TiberiusTeng> change that GL_LINEAR to GL_NEAREST (disable interpolation) and see how it goes)
18:13:57 <Scaevolus> if that was the problem ;)
18:14:15 <TiberiusTeng> if you did changed TexCoordScale() or PARAM constants, change them back ...
18:14:20 <TiberiusTeng> I don't think so, actually ...
18:14:24 <ccfreak2k> I didn't change them.
18:14:35 <TiberiusTeng> it won't do harm on 7600GT nor i855GME ...
18:15:18 <ccfreak2k> Colors issue still there.
18:15:29 <Scaevolus> can't you just set sdl to use opengl?
18:15:46 <TiberiusTeng> but I don't know SDL ...
18:15:51 <TiberiusTeng> so I modified win32v instead
18:16:56 <ccfreak2k> The color issue only afflicts specific colors.
18:17:01 <Scaevolus> because if you just changed sdl, it should magically work on all platforms
18:17:08 <ccfreak2k> For example, roadway dividers scroll red and orange.
18:17:20 <TiberiusTeng> peter1138 did changed sdl_v to support OpenGL before, he even gave me the diff
18:17:37 <TiberiusTeng> but I really don't know how to modify/test it to fit my current situation
18:18:04 <TiberiusTeng> ccfreak2k, yes, it's some kind of texture-coordinate rounding
18:21:46 <TiberiusTeng> the texture coordinate just 'touched' the place where the card thinks it's boundary between two colors
18:21:53 <TiberiusTeng> hence the color flickering
18:22:35 <TiberiusTeng> check if you forced texture interpolation in the driver
18:22:58 <ccfreak2k> This happens when panning.
18:25:12 <ccfreak2k> I turned off all of the optimizations I could find, and I still get the color thingy.
18:25:46 <TiberiusTeng> ahh I give up, it don't even do glReadBuffer/glDrawBuffer/glTexSubImage2D well ......
18:25:58 *** [1]BigBoss has joined #openttd
18:27:14 <ccfreak2k> I changed it to GL_LINEAR and it's all trippy.
18:27:38 <TiberiusTeng> ccfreak2k, go to ScrollBuffer() and try replace all GL_RGBA in it to GL_RGBA8 and see how it works?
18:27:53 <TiberiusTeng> you can't interpolate between palette indexes!
18:29:50 <ccfreak2k> What change am I looking for with GL_RGBA8?
18:30:06 <TiberiusTeng> only in ScrollBuffer()
18:30:22 <ccfreak2k> I aleady made the changes.
18:30:22 <TiberiusTeng> correct panning of the map
18:31:16 *** [1]BigBoss is now known as BigBoss
18:31:57 <ccfreak2k> Also, it doesn't like being resized.
18:33:13 <TiberiusTeng> resizing isn't a problem on i855/7600gt ...
18:33:47 <TiberiusTeng> perhaps I should go purchase some ATI card and spend time debugging, or just rename it to nvogl.cpp/non-ati-opengl.cpp
18:33:49 <ccfreak2k> I can resize the window fine, but if I choose a pre-baked resolution in options, the screen goes black.
18:33:51 *** stillunknown has joined #openttd
18:34:10 <TiberiusTeng> ahh. that could be a bug. I'll check it later
18:34:25 <ccfreak2k> Alternatively, you can just code for "everyone else" and someone could maintain an ATi patch.
18:35:05 <ccfreak2k> glGenProgramsARB (sprite) still emits "invalid operation".
18:35:42 *** dR3x4cK has joined #openttd
18:36:18 <TiberiusTeng> I think the error flag is left by some earlier OpenGL calls, I didn't add CheckGLError() for every functions ...
18:38:41 <ccfreak2k> pngcrushing screenshot.
18:41:57 <ccfreak2k> All instances of GL_RGBA replaced with GL_RGBA8 in ScrollBuffer().
18:43:00 <TiberiusTeng> hmm ... so the buffer copying completely fails ...
18:43:28 <TiberiusTeng> the white region should be 'copied' from previous frame, and the blitter would only draw 'new' parts
18:43:54 <TiberiusTeng> i.e. those vertical bands
18:47:40 *** Dred_furst has joined #openttd
18:48:49 <Bjarni> now this looks interesting... are you guys trying to get a full openGL implementation working?
18:49:10 <ccfreak2k> TiberiusTeng is trying to make an opengl blitter.
18:49:15 <TiberiusTeng> it depends on what you mean by 'full OpenGL implementation' :p
18:49:24 <ccfreak2k> And, well, it works, so long as you don't have ATi hardware.
18:49:32 <TiberiusTeng> I just want ATI card owners can see what NV card owners see now
18:49:58 <TiberiusTeng> Bjarni, I believe you already see that thread ... :p
18:50:18 <Bjarni> because I have been quite busy lately
18:51:28 <Bjarni> <ccfreak2k> And, well, it works, so long as you don't have ATi hardware. <-- it will not be accepted if you leave some poor souls behind without suppport
18:51:42 <Bjarni> specially if one of those poor souls is me ;)
18:52:00 *** ben_goodger has joined #openttd
18:52:42 <TiberiusTeng> it really hurts when ATI didn't support the spec they originally proposed :Q
18:53:29 <TiberiusTeng> now I wonder if it runs on M$ Windows software OpenGL 1.1 ...
18:53:44 <TiberiusTeng> (of course not, it used 1.2 features :p)
18:54:50 <TiberiusTeng> but honestly, it still depend on some hardware features and drivers
18:55:17 *** ladisCZ has joined #openttd
18:55:19 <TiberiusTeng> I still want it work correctly with Intel/NV hardwares if I do include some hack for ATI ...
18:55:42 <Bjarni> I was about to ask the same thing
18:55:47 <ccfreak2k> Does Bjarni use OS X?
18:56:04 <glx> read the credits in openttd
18:56:14 <ccfreak2k> I don't read credits.
18:56:16 <TiberiusTeng> now if someone make a SDL OpenGL driver ... :p
18:56:18 <ccfreak2k> I rarely even read documentation.
18:56:32 <TiberiusTeng> Bjarni, which graphics card are you using ?
18:57:55 <ccfreak2k> Also, if you intend to use EXT_texture_non_power_of_two later on, ATi doesn't support that.
18:58:02 <ccfreak2k> They use EXT_texture_rectangle instead.
18:58:12 <TiberiusTeng> I already supported the situation without that extension
18:58:24 <ccfreak2k> Welcome to the world of graphics. :)
18:58:28 <TiberiusTeng> it's just not that memory efficient
18:58:42 <TiberiusTeng> (I'm playing it without remapped colors on my good old i855GME)
18:58:49 <Bjarni> I prefer working over memory efficiency ;)
18:59:04 <glx> and I though opengl (or directx) was there to ease development
18:59:27 <Bjarni> glx: that's totally wrong
18:59:34 <TiberiusTeng> EXT_texture_rectangle is not that useful
18:59:40 <Bjarni> directx is intended to prevent porting windows games to other platforms
18:59:57 <TiberiusTeng> since it uses another coordinates system, and lacks support of many useful functions (like mipmapping)
19:00:03 <ben_goodger> "those who would trade a little memory-efficiency for a little functionality deserve neither and will lose both"
19:01:02 <yorick> that's exactly why directx 10 doesn't run on anything else than vista
19:01:02 <TiberiusTeng> in contrast ARB_texture_non_power_of_two is not a 'special' texture target, it supports all functions ...
19:01:49 <Mchl> ben_goodger: Was Ben Franklin a software developer as well?
19:01:51 <TiberiusTeng> and speaking of memory efficiency, texture atlases are primary memory consumers
19:02:11 <TiberiusTeng> something like two 1024x4096x4 textures ...
19:02:17 <ben_goodger> Mchl: not quite...
19:02:48 <Mchl> a man of many virtues nonetheless
19:02:59 <TiberiusTeng> actually I really hope ccfreak2k can solve ATI problems for me ;)
19:03:21 <ccfreak2k> I know next to nothing about OpenGL.
19:03:22 <ben_goodger> it's a shame nobody pays any attention to him any more
19:03:39 <yorick> let me rephrase that bug: autoreplace has unexpected results if Sell works correctly but build doesn't
19:04:32 <TiberiusTeng> right now I'm guessing, and ccfreak2k is helping me proving if it works (or, mostly, not) on an ATI hardware ...
19:05:08 <TiberiusTeng> so ... map panning don't work; is the water animating ? how about world map ?
19:05:10 <ccfreak2k> It'd be nice if I could profile it.
19:05:31 <TiberiusTeng> and check cargo payment rates graph too
19:05:53 <TiberiusTeng> the one I did yesterday ...
19:06:19 <TiberiusTeng> ofcourse you can use gprof ...
19:06:24 <ccfreak2k> I'm trying to use VisualStudio's, if it has one.
19:06:32 <TiberiusTeng> I just wanna play the evaluation version
19:06:51 <ccfreak2k> I saw a Second Life architech demo Second Life in OSX's profiler, which got me interested.
19:06:59 <TiberiusTeng> Visual Studio Performance Tool ... is only available in Team something version
19:07:42 <ben_goodger> the 2.4 gigabyte text editor
19:07:50 <yorick> why can I not replace helicopters by small airplanes
19:07:59 <TiberiusTeng> anyway we have progress today ... at least ATI cards can display something
19:08:00 <ben_goodger> it takes about fifteen minutes to load on my college computer
19:08:02 <yorick> while the small airplanes ARE in the autoreplace list?
19:08:27 <TiberiusTeng> wait a sec, ccfreak2k, are you still here ?
19:08:28 <ben_goodger> Prof_Frink: no, visual studio. emacs remains steady at about ten megs, I believe
19:08:40 <Yexo> yorick: I guess you can't replace them because that would fail on helidepots
19:09:06 <Prof_Frink> ben_goodger: Steady? That's gone up 2 meg since the acronym.
19:09:20 <TiberiusTeng> replace all glDrawBuffersARB with glDrawBuffersATI and see if you can pan the map correctly?
19:09:21 <ben_goodger> I think the acronym originates in the late eighties
19:09:39 <ccfreak2k> VisualStudio 2005 loads in about 5-10 seconds on my desktop.
19:09:48 <ccfreak2k> Less if I launched it at least once in the past few hours.
19:09:50 <yorick> Yexo: but why are they displayed?
19:10:18 <ccfreak2k> TiberiusTeng, is the water supposed to animate?
19:10:19 <yorick> I know why you can't autoreplace them
19:10:44 <ccfreak2k> Because it's not for me.
19:10:45 <yorick> because: if selling a plane works, and buying another one doesn't, it has some ugly effects
19:10:54 <Yexo> problems with orders and such, for example order to heliports or oil platforms
19:10:54 <TiberiusTeng> so the DrawBuffers isn't working ...
19:11:23 <ccfreak2k> The newspaper black & white "photo" also experiences incorrect colors.
19:11:32 <peter1138> ben_goodger: eight megs and constantly swapping
19:11:41 <peter1138> later revised to eighteen, and then eighty...
19:11:50 <ccfreak2k> Cargo payment rates graph looks ok.
19:11:55 <ben_goodger> oh, it can't be eighty, surely
19:12:05 <yorick> Yexo: I patched a fix for that orders
19:12:18 <peter1138> vim is 32MB for me ;)
19:12:21 <Prof_Frink> peter1138: exabytes and constantly swapping
19:12:21 <ben_goodger> obviously I've not been loading enough extensions into it
19:12:33 <peter1138> gedit is... er... 280MB
19:12:41 <peter1138> although that's virtual
19:12:45 <yorick> and, if selling a plane works, but buying another one doesn't, you get the money, but keep the old plane
19:12:59 <peter1138> vim is a mere 3MB resident
19:13:59 <Yexo> <yorick> and, if selling a plane works, but buying another one doesn't, you get the money, but keep the old plane <- that's clearly a bug, but another issue as why replacing planes <> helicopters doesn't work
19:14:25 <Yexo> see players.cpp line 746
19:14:31 <yorick> I know why it doesn't work
19:15:04 <blathijs> peter1138: And even vim is completely bloated with functionality already :-)
19:15:14 <yorick> because you get the bug when trying to buy an airplane on a helistation
19:15:20 *** Boyinblue0 has joined #openttd
19:15:31 <TiberiusTeng> looks it can't get glCopyTexImage2D working
19:15:40 <yorick> but they don't bother at fixing the bug because they filter out all cases in which it could happen
19:15:58 <TiberiusTeng> try add a CheckGLError("glCopyTexImage2D") below that function in DrawColorMappingRect() and see how it says?
19:16:43 *** grumbel has joined #openttd
19:17:18 <TiberiusTeng> invalid_enum is impossible, invalid_value is not very distinguishing, and invalid_operation is next to impossible ... holy.
19:17:36 *** grumbel_ has joined #openttd
19:17:50 <yorick> FOUNT UNLIMITED MONEY EXPLOIT
19:18:05 <TiberiusTeng> ahh. ok. it shouldn't work. I forgot to modify here to support power-of-two textures ...
19:18:09 <ccfreak2k> Which function is "that function"?
19:18:10 <yorick> fortunatly, related to the bug I want fixed
19:18:22 <TiberiusTeng> ccfreak2k, nope, my fault
19:18:24 <peter1138> blathijs: "already" does not fit there ;)
19:18:34 <peter1138> i think the word you mean is "now"
19:19:22 <blathijs> I meant that it is bloated, even in 3MB :-)
19:19:52 *** sato_san has joined #openttd
19:20:04 <SmatZ> yorick: are you sure about that?
19:20:19 <TiberiusTeng> modify the ScrollBuffer() function, add a CheckGLError("glTexImage2D"); below glTexImage2D(), and another CheckGLError("glCopyTexSubImage2D"); below glCopyTexSubImage2D() ...
19:20:39 <yorick> build airport, cheapest airplane, set autoreplace, care NOT to have enough money to autoreplace
19:20:39 <TiberiusTeng> (is anyone suggesting a better way to trace opengl error messages under this situation? ~_~)
19:20:44 <yorick> and make it autoreplace
19:20:56 <yorick> and money got raised by 30,000
19:21:11 <Prof_Frink> blathijs: Yeah, you shouldn't use an editor. Just echo and sed.
19:21:12 <yorick> which you could 'donate' to another company and press the autoreplace button again
19:21:29 <yorick> and there you got your unlimited money
19:21:52 <Prof_Frink> yorick: Or invest in buying land.
19:21:57 <yorick> most likely it's in 0.6 too
19:22:05 <SmatZ> yorick: doesn't work for me
19:22:45 <Rubidium> Prof_Frink: why do you need echo?
19:23:37 <ccfreak2k> The water animates during map generation, but not while playing.
19:23:45 <SmatZ> yorick: in clean trunk?
19:23:53 <Prof_Frink> Rubidium: So you can pretend your computer is talking to you.
19:24:06 <yorick> SmatZ: don't use the autoreplace button ;)
19:24:51 <Rubidium> Prof_Frink: you can use sed for that too
19:25:18 <SmatZ> yorick: you are right, sir!
19:26:26 <yorick> SmatZ: the more expensive the plane, the more effective this works
19:27:11 <SmatZ> yorick: hopefully doesn't work in 0.6
19:27:34 <yorick> it's possibly the vehicle economy backup
19:27:35 <ccfreak2k> Also, two more screenshots:
19:27:41 <SmatZ> yorick: most likely :-P
19:27:53 <yorick> and if so, Bjarni caused a critical bug
19:28:35 <TiberiusTeng> hmm. possibly unpaired glBegin/glEnd (very unlikely) or the texture didn't correctly created ... hmm ...
19:29:05 <TiberiusTeng> and the cursor sprite have a strange tint :p
19:29:19 <TiberiusTeng> ahh, the cursor won't leave any tail, will it ?
19:29:27 <ccfreak2k> The cursor doesn't have a trail, no.
19:29:43 <ccfreak2k> It's just like the regular cursor, except for the odd coloring at the tip.
19:29:48 <TiberiusTeng> hmm so CopyToBuffer is really working.
19:29:54 <yorick> yeah, it doesn't happen in 0.6 :)
19:30:04 <TiberiusTeng> water still lack animation ?
19:30:10 <ccfreak2k> Yep, no water animation.
19:30:12 <yorick> peter1138, I've been trying that for like the half day
19:30:29 <ccfreak2k> I get the white still when scrolling, but if I "jerk" the screen and make it all redraw, it doesn't happen.
19:30:55 <yorick> oh, it could help if I tested correctly
19:31:28 <ccfreak2k> glx, the water is "stuck" like that, sort of.
19:32:13 <TiberiusTeng> the palette keeps updated, but the screen isn't constantly redrawn with updated palette
19:32:26 <yorick> ah, it still doesn't happen
19:32:41 <TiberiusTeng> so it'll look like 'partially animated' when the gfx decide to redraw some water sprites
19:33:13 <glx> without any animation all water should look the same
19:33:38 <glx> so animation works, but you don't see it in realtime
19:33:43 <TiberiusTeng> the palette's changing, but the part redrawing the screen with new palette is broken
19:33:51 <SmatZ> glx: when animation doesn't work, the water is drawn with palette that was actual while that part was redrawn
19:33:59 <SmatZ> but the colours don't cycle
19:34:09 *** Brianetta has joined #openttd
19:34:42 <TiberiusTeng> ccfreak2k, did you change GL_RGBA8 back to GL_RGBA ?
19:34:49 <TiberiusTeng> (you should do this)
19:35:08 <ccfreak2k> one of them was already GL_RGBA8 before I changed it.
19:35:52 <TiberiusTeng> your card shouldn't take that path (easy_texture_size)
19:36:52 <TiberiusTeng> I wonder why cursor works, but scrolling don't
19:36:56 <TiberiusTeng> they're basically the same ...
19:37:15 <ccfreak2k> Is the cursor supposed to do something special?
19:37:17 <ccfreak2k> I thought it was static.
19:37:44 <TiberiusTeng> but the cursor background is constantly copied to a buffer, and restored when you moved the cursor
19:38:01 <TiberiusTeng> the area below cursor, to be precise
19:38:01 <ccfreak2k> Yeah, moving the cursor looks fine.
19:38:27 <TiberiusTeng> and that's also what ScrollBuffer() does, only it operates on a way more bigger region
19:38:41 <ccfreak2k> Changing the things back to GL_RGBA makes one error go away, but I still get this when I scroll: OpenGL Error at glTexImage2D: invalid operation
19:39:38 <ccfreak2k> As someone else said, the water still "animates", but it doesn't get drawn.
19:39:52 <ccfreak2k> If I pan across the water, the animation paints itself in the redrawn sections.
19:40:27 <TiberiusTeng> does reducing the window size work?
19:40:32 <glx> maybe the card does too much optimisation, thinking the screen didn't changed
19:40:46 <ccfreak2k> Window resizing works.
19:40:47 <TiberiusTeng> I start wondering if it exceeds 2048x2048 size limit ... or some memory caps
19:40:56 <ccfreak2k> The water updates...just when it resizes, though.
19:41:12 <TiberiusTeng> make the window smaller then try panning the map?
19:41:15 <ccfreak2k> Moving an in-game window around over the water causes it to animate, again, I assume when it redraws the water.
19:41:28 <ccfreak2k> Resizing and panning causes the same problem.
19:42:43 <yorick> the 4 critical bugs all have to do with autoreplace currently ;(
19:43:06 <ccfreak2k> The sides of windows, the status bar on the bottom and tool bar on the top cause a kind of line of demarcation.
19:43:23 *** Progman has joined #openttd
19:45:45 <TiberiusTeng> ccfreak2k, you said it's glTexImage2D failing ... could you watch or printf() texw and texh ?
19:46:04 <TiberiusTeng> yeah, something like that
19:46:37 <TiberiusTeng> hmmmmmm ............
19:46:56 <ccfreak2k> Notice the sort-of invisible boundaries.
19:47:23 *** [1]BigBoss has joined #openttd
19:47:32 <ccfreak2k> texh is 1024, texw is 256.
19:47:40 <TiberiusTeng> ahh, not bigger than 2048 ...
19:48:41 <Bjarni> <ben_goodger> "those who would trade a little memory-efficiency for a little functionality deserve neither and will lose both" <-- I say that getting my own hardware to be supported isn't considered "a little functionality"
19:49:05 <ccfreak2k> glTexImage2D only seems to throw an error every OTHER time it's called.
19:49:30 <yorick> finally found a way to phrase my bug :)
19:49:30 <Bjarni> I just love errors that happens at random times
19:49:34 <TiberiusTeng> now that's strange ~_~
19:49:51 <peter1138> Bjarni needs supports for his 'hardware'
19:49:53 <yorick> it's called "free money on autoreplace failure" now :)
19:50:05 <peter1138> do they ladies fear you?
19:50:09 <TiberiusTeng> I wonder if OSX machines can be equipped with NV cards
19:50:11 <ccfreak2k> texw is sometimes 512, texh remains unchanged at 1024.
19:50:20 <Bjarni> TiberiusTeng: they can
19:50:27 <ccfreak2k> TiberiusTeng, that's the great thing about standards like AGP. :)
19:50:30 <Bjarni> but I just don't have such a thing here
19:50:34 <TiberiusTeng> yep, that's normal, notice there are more than one regions to be scrolled
19:51:03 <TiberiusTeng> gfx will cut the screen into few rectangles, then move them one by one
19:51:09 <TiberiusTeng> and then redraw the new gaps
19:51:20 * peter1138 pokes ccfreak2k with a driver
19:52:02 <peter1138> physical bus compatibilty is no use if there are no drivers
19:52:14 <TiberiusTeng> actually at first I think I would met most problems on i855GME, not something like X850XT :p
19:52:56 <ccfreak2k> I wasn't aware the i800 series even supported fragment programs.
19:53:03 *** [1]BigBoss is now known as BigBoss
19:53:32 <TiberiusTeng> yes it doesn't ... but like I said, it still runs and displays something, just without 'remapped colors' and 'palette animation'.
19:53:59 <TiberiusTeng> did you replaced glDrawBuffersARB with glDrawBuffersATI ?
19:54:37 <TiberiusTeng> and expect to see me online 24 hours (or even 48 hours) later ... whatever it works or not. :~
19:54:40 <TiberiusTeng> yes, all instances.
19:55:40 <TiberiusTeng> ha, it's cached to reduce state changes.
19:56:02 <ccfreak2k> Scroll problem is still there.
20:00:29 *** Progman has joined #openttd
20:02:03 <TiberiusTeng> sorry, gotta take a rest ...
20:02:09 <TiberiusTeng> will back on this later
20:02:59 <TiberiusTeng> I think I'll find some computer with capable ATI card to debug it myself, it's too tired for both me an ccfreak2k ...
20:08:53 <ccfreak2k> From what I can see, VS8 Profession (what I'm using) should have everything that the Team editions do.
20:09:04 <ccfreak2k> Professional, I mean.
20:22:03 <TiberiusTeng> try Catalyst 8.6 ?
20:22:28 <TiberiusTeng> just couldn't fall asleep ... 04:25 and I have to do a presentation at 13:00
20:23:29 <ccfreak2k> I didn't get 8.6 because 8.5 was the newest when I was looking for drivers.
20:24:07 <TiberiusTeng> from what I read, ATI drivers does support OPTION ARB_precision_hint_fastest ...
20:25:42 <ccfreak2k> I can't find it in EVEREST's list.
20:26:20 <TiberiusTeng> no, it's a fragment program option defined by ARB_fragment_program ...
20:26:41 <TiberiusTeng> catalyst 8.5 isn't that old either, I wonder ...
20:27:08 <ccfreak2k> Nope. Its version indicates May, 2008.
20:40:46 <TiberiusTeng> I'm just considering to purchase an HD2600 myself ...
20:41:30 *** stillunknown has joined #openttd
20:41:32 <ccfreak2k> Just find a Radeon 9600 Pro or something from eBay or craigslist.
20:42:09 <TiberiusTeng> it'll be funny if HD2600 runs great while X850XT doesn't ...
20:42:27 <TiberiusTeng> but I do suspect I did something wrong before compiling those shaders
20:43:26 <TiberiusTeng> ccfreak2k, if you have some free time, please get gDEBugger trial, use it to run openttd.exe and see what you got in GL_AUX0 buffer
20:43:50 <TiberiusTeng> it should contain color indexes from sprites ...
20:53:44 *** stillunknown has joined #openttd
21:01:31 <ccfreak2k> Hmm, vs8 seems to have locked up
21:01:46 <ccfreak2k> Oh wait, nevermind.
21:03:29 <ccfreak2k> What frame terminators do I want to uh...check?
21:04:27 <Scaevolus> how's it going, ccfreak2k?
21:04:57 <ccfreak2k> I'm trying to find where the watch is on this thing.
21:05:19 <Dred_furst> TiberiusTeng, out of interest why have you made it win32 only, and secondly why shaders?
21:05:20 <TiberiusTeng> or wglSwapBuffers(), both are fine
21:05:40 <TiberiusTeng> because I only have win32 computers :P
21:06:04 <TiberiusTeng> it will only need a thin layer of OpenGL initialization to work on other platforms.
21:06:21 <Dred_furst> Just out of interest, openTTD uses SDL already right?
21:06:31 <TiberiusTeng> and about shaders ... it's the most flexible way to support palette-based sprites & palette animations now
21:06:37 <Scaevolus> Dred_furst: yeah, I tried proposing the idea of just setting sdl to opengl render mode
21:06:50 <TiberiusTeng> but I don't know how
21:06:53 <Dred_furst> plus it makes it portable
21:06:53 <ln> Dred_furst: not on every platform.
21:07:04 <TiberiusTeng> so please do it ;)
21:07:07 <Scaevolus> ln: win/lin/mac all have opengl ;)
21:07:07 <Dred_furst> whatever SDL and opengl is supported on
21:07:24 <Scaevolus> which is what, 99+% of the PC market?
21:07:32 <TiberiusTeng> just diff win32_v.cpp and win32gl_v.cpp and you'll know what should be prepared for opengl blitter ...
21:07:50 <Scaevolus> 12:17:19 < TiberiusTeng> peter1138 did changed sdl_v to support OpenGL before, he even gave me the diff
21:07:56 <Prof_Frink> Scaevolus: Remember that openttd also supports OS/2
21:07:58 <ln> Scaevolus: yeah, but OpenTTD doesn't use SDL on every platform is what I was saying.
21:07:59 <Scaevolus> TiberiusTeng: do you still have that diff?
21:08:18 <Dred_furst> and if you want opengl, I would reccommend it
21:08:34 <TiberiusTeng> yes, but not at my hand
21:08:34 <ln> Dred_furst: "can", but shouldn't.
21:08:46 <Scaevolus> ln: but it's just another display mode, so it's okay if it's not completely portable across all the platforms openttd runs on
21:08:59 <ccfreak2k> TiberiusTeng, where IS the GL_AUX0 buffer? :|
21:09:24 <TiberiusTeng> pause the execution in gDEBugger
21:09:37 <TiberiusTeng> view -> viewers -> textures and buffers viewer
21:09:38 <Dred_furst> in the case of opengl, it would be preferred because all SDL handles is the creation of the window
21:09:58 <ccfreak2k> Yeah there's nothing named GL_AUX0 in that window.
21:10:02 <Scaevolus> ...how many OS/2 systems have 3D acceleration?
21:10:15 <TiberiusTeng> is there a AUX0 in buffers list ?
21:10:17 <ln> Dred_furst: adding a new dependency to a new library is to be avoided.
21:10:33 <Dred_furst> but ln its already a partial dependancy
21:10:37 <TiberiusTeng> but you have textures, front/back buffers ?
21:11:09 <ccfreak2k> Static buffers, depth buffer, stencil buffer, back buffer, front buffer.
21:11:47 <ln> Dred_furst: it's not a "partial dependency" as SDL is not required for OTTD on e.g. Mac and Windows.
21:12:14 <TiberiusTeng> so the card claims it support draw buffers? hmm ......
21:14:28 <Dred_furst> but it is on linux
21:14:32 <Dred_furst> and SDL is cross platform
21:14:41 <Dred_furst> and is known to compile with TTD on those platforms
21:15:37 <Bjarni> SDL+OSX caused weird bugs in OpenTTD
21:15:46 <Bjarni> like you couldn't quit the game without crashing it
21:15:59 <Bjarni> and it crashed inside the SDL library o_O
21:15:59 <Dred_furst> then something is up with the coding *shrug*
21:16:17 <TiberiusTeng> the file from peter1138 ... hope it's OK to redistribute it
21:17:35 <DorpsGek> SmatZ: ludde was last seen in #openttd 24 weeks, 0 days, 0 hours, 53 minutes, and 50 seconds ago: <ludde> :)
21:18:28 <bowman> saw him recently in a dreambox-related setting
21:18:39 <bowman> hmm maybe not so recently
21:19:56 <bowman> yep, still on efnet :)
21:21:11 <Dred_furst> TiberiusTeng, anything submitted to openttd must be GPL :)
21:21:50 <SpComb> Dred_furst: there are GPL-compatible licenses
21:22:30 <SpComb> afaik GPL code can use e.g. BSD-licensed code, and even re-license it and any modifications under theGPL
21:23:10 <SmatZ> when you code something based on GPL'ed code, nobody can force you to make it "public available"
21:23:18 <SmatZ> when you don't release binaries or so
21:23:59 <SpComb> indeed, the only obligation you have under the GPL is to provide the sources along with the binaries
21:24:09 <ccfreak2k> make the source available somehow.
21:24:16 <SpComb> although if you only provide a written notice of how the sources are available with the binary, then you have to give the source to anyone who asks
21:24:21 <Eddi|zuHause3> well, he "distributed" it to TiberiusTeng, he must have done this under the GPL
21:24:34 <Scaevolus> more likely he didn't mention license at all
21:24:35 <SpComb> but if you distribute the source along with the binary, then you're under no obligation to hand out the source to anyone else
21:24:42 <Eddi|zuHause3> yes. because it is a derived work
21:24:42 <Scaevolus> just said "hey, I did this hack, have a look"
21:25:00 <Scaevolus> oh right, forgot it was derived
21:25:20 <SpComb> code without a license follows the plain 'ol copyright laws
21:26:03 <SpComb> hmm, perhaps if you distribute modifications to GPL code without licensing it under the GPL, then the copyright holders of the GPL software could sue you for copyright infringement
21:26:14 <Eddi|zuHause3> well, like i said, linking to his file will not get you in license/copyright trouble ;)
21:26:49 <SmatZ> hmm can you distribute diffs without mentioning that it is under GPL and where the licence is available for download?
21:27:54 <Eddi|zuHause3> that's a very theoretical question, as "everybody knows" that the GPL is found at the openttd repo
21:28:16 <Scaevolus> THIS DIFF VIOLATES COPYRIGHT
21:28:17 <Eddi|zuHause3> which is the same repo as the diff is made to apply with
21:28:43 <ccfreak2k> I put my license (BSD) in every one of my source files. :)
21:29:08 <Eddi|zuHause3> ccfreak2k: but that does not usually make it appear in the .diff
21:29:21 <ln> the situation is even more complex, because the copyright holder of OTTD source code is not known.
21:29:33 <ccfreak2k> If the .diff modifies one of my files, it too falls under the BSD license (since they're contributing to my work).
21:29:41 <ccfreak2k> Not that I, myself, would care.
21:30:25 <Eddi|zuHause3> ccfreak2k: yes, but you must (theoretically) state that license for every single diff you give out
21:30:52 <ccfreak2k> Well, if the diff created a NEW file, then possibly.
21:31:15 <ccfreak2k> If it only modified my existing files, then I would argue that it's not necessary, since the diff is made to only patch against my files.
21:31:23 <Eddi|zuHause3> ccfreak2k: there are no implicit licenses
21:31:47 <Eddi|zuHause3> only copyright is implicit (and not transferable, in germany)
21:31:59 <ccfreak2k> Then again, if someone is going to pedant over a diff file for openttd, do you really want his contribution? :)
21:32:28 <Eddi|zuHause3> well, there are always two choices
21:33:22 <Eddi|zuHause3> a) he released it under GPL and didn't mention it, or b) that he ignored the GPL
21:34:00 <Eddi|zuHause3> the question is whether you can safely assume a) is the case
21:34:18 <Eddi|zuHause3> and the question has been asked multiple times already
21:35:05 <ccfreak2k> IIRC TiberiusTeng didn't include a license in his diff/patch files. What does THAT mean?
21:35:35 <Eddi|zuHause3> i mean this discussion is coming up every half year or so
21:37:49 *** ben_goodger has joined #openttd
21:38:32 <planetmaker> may "static void somefunction()" be in a header file?
21:38:43 <ccfreak2k> You mean prototypes?
21:38:44 <planetmaker> or is static not allowed there?
21:38:55 <SmatZ> everything is allowed there
21:39:13 <SmatZ> there are many "static inline" functions in heard files
21:39:37 <planetmaker> I'm thinking about putting static void SubtractMoneyFromAnyPlayer in functions.h
21:40:01 <planetmaker> it's so far only accessible in player.cpp, but I need it in sharedinfrastructure.cpp
21:40:57 <SmatZ> planetmaker: put only declaration there (prototype)
21:41:03 <SmatZ> and remove static from definition
21:41:27 <planetmaker> what's the purpose of defining something static?
21:42:02 <SmatZ> that its symbol isn't exported probably
21:43:13 <planetmaker> hm... the wiki references objects. this procedure is not part of a class. So I trust SmatZ :)
21:43:21 <SmatZ> static member function is something different
21:43:50 <ccfreak2k> There's several types of "static" depending on what you're talking about. :)
21:44:42 <ln> planetmaker: yes, SmatZ is right, a static function is only visible within the same compilation unit.
21:44:43 <planetmaker> I guess then it relates to linking here, yes
21:45:10 <planetmaker> k, so it has to go definitely :)
21:46:52 <CIA-6> OpenTTD: truebrain * r13625 /branches/noai/src/toolbar_gui.cpp: [NoAI] -Fix: re-introduce the About button in the last menu, to make Yexo really really happy again :)
22:02:47 *** Osai`off is now known as Osai
22:10:47 <Eddi|zuHause3> planetmaker: "static" means that the function is only valid in the file you define it (or files that include this file, then in each file separately), so you cannot link to/call that function from outside of the file
22:11:48 <Eddi|zuHause3> completely different from the java-meaning of "static", where that refers to a class-function (instead of an object-function)
22:12:12 <Bjarni> if you make an inline function in a header without using the static keyword the compiler will presume that you will link to this function and your binary will grow as it has functions you will never call
22:13:46 <Eddi|zuHause3> but he doesn't want to inline, he wants to export the function, which means unstatic-ing is the right way ;)
22:31:23 *** Osai is now known as Osai^zZz
22:32:50 <Scaevolus> Is the distance a good travels calculated as manhattan distance between the origin and receiving stations?
22:39:20 <Scaevolus> I'm writing a bit on cargo revenues for the wiki, should I add a section to the Cargo page, or should I make a separate Cargo_revenues page?
22:40:19 <SmatZ> I don't know, I don't use wiki :-x
22:40:34 <ccfreak2k> Stick it on the cargo page.
22:40:39 <SmatZ> that often to know its structure :)
22:40:45 <ccfreak2k> If it's so necessary to put it on its own page, someone will move it.
22:46:25 *** Lakie` is now known as Lakie
23:00:20 *** Osai^zZz is now known as Osai^zZz`off
23:18:57 <Tim__> Hm... Any developer here?
23:19:19 <Eddi|zuHause3> developers are never going to answer such a question
23:19:40 <Belugas> indeed, Eddi|zuHause3, indeed
23:20:00 <Tim__> picture says more than thousands word, i guess
23:20:15 <Tim__> if removing signals one for one, the signals turn red like they should
23:20:25 <Belugas> i found a bugeuhh, i found a bugeuhh...
23:20:54 <Lakie> Being a developer is so much more fun when no-one knows...
23:21:28 <Eddi|zuHause3> Tim__: bugs go to bugs.openttd.org
23:21:32 <Lakie> Belugas: I get that constantly with aRV's and thats not even my project!
23:22:02 <Belugas> i can imagine, Lakie :)
23:22:27 <Eddi|zuHause3> Tim__: preferably with savegame and description how to reproduce
23:22:35 <Belugas> Tim__, Eddi|zuHause3 is right. Go to bugs.openttd.org, enter your bug there, with as much info as possible
23:24:43 <Belugas> boring... oinly one video of ratts in capital
23:25:03 <Belugas> how culd I figure out what they are playing then???
23:27:42 <Belugas> ho... of course... the master of the signals!
23:28:17 <Eddi|zuHause3> i assume it's a missing call to the "update signals on track" function
23:28:46 <glx> or incorrect direction to update
23:29:56 <SmatZ> but it won't be anymore possible to do what Tim__ did
23:30:40 <Eddi|zuHause3> that option also crossed my mind when viewing the picture
23:30:58 <Eddi|zuHause3> "that should probably not be possible at all"
23:31:39 <Scaevolus> who's responsible for the OpenTTD wiki?
23:31:53 <Scaevolus> no, I mean the admin
23:31:58 <Scaevolus> it doesn't seem to have LaTeX :(
23:32:09 <Scaevolus> so I can't do nice-looking mathematical formulas
23:33:22 <DorpsGek> SmatZ: Commit by smatz :: r11871 /trunk/src (10 files in 2 dirs) (2008-01-16 01:18:15 UTC)
23:33:23 <DorpsGek> SmatZ: -Fix [FS#1074]: do not update signals after each tile when building/removing a large block of track/signals/station
23:34:14 <Eddi|zuHause3> so it's even your own fault? :p
23:35:56 <Scaevolus> who's the webmaster of openttd.org?
23:36:34 <ccfreak2k> Tord Viktor it looks like.
23:36:45 <Scaevolus> no, that's the registrar
23:36:49 <Tim__> narf, i just spend a lot of time producing a savegame to show the bug and now you just fixed it *grrr*
23:37:21 <Tim__> but thank you nevertheless :)
23:37:22 <Eddi|zuHause3> happens when you don't follow the discussion ;)
23:37:35 <Tim__> hey, you said report it...
23:38:31 <SmatZ> that was hard to write that commit message :-x
23:38:32 <CIA-6> OpenTTD: smatz * r13626 /trunk/src/rail_cmd.cpp:
23:38:32 <CIA-6> OpenTTD: -Fix (r11871): signals were not updated when player removed signals from nonexistent track sharing bits with existing track
23:38:32 <CIA-6> OpenTTD: So disallow removing signals from nonexistent track
23:39:16 <Tim__> ehm... this bug also happens if you remove signals on the same track by dragging
23:39:27 <SmatZ> r11871 is mentioned for easier backporting :)
23:39:47 <SmatZ> Tim__: wait for next nightly, should be fixed
23:40:01 <Tim__> it only happens when you remove signals on adjacent tracks...
23:40:32 <SmatZ> Tim__: there are more bugs than you posted?
23:40:48 <Tim__> No, just the one in the image :)
23:41:01 <SmatZ> it is not there anymore :-P
23:41:21 <Tim__> Since when are you a dev? Did i miss something? :D
23:42:53 <Scaevolus> gah, the bugtracker requires authentication by an admin :(
23:42:56 <Tim__> As long as he still has some time to play on openttdcoop...
23:42:56 <ln> virtually everyone is a dev nowadays.
23:45:00 <Tim__> By the way :D - Any chance for that Powerplant-behaviour to be changed?
23:45:09 <Eddi|zuHause3> Tim__: yes, newgrf
23:45:29 <ln> notgoingtohappennotfunnyendofdiscussion
23:45:48 <Tim__> I have 70% on my side xD
23:46:16 <Eddi|zuHause3> then make the newgrf, and everyone will be happy ;)
23:46:39 <Eddi|zuHause3> and powerplants already behaved like this in original TT
23:46:39 <Scaevolus> should I add a bug to the bugtracker, or should I email mihamix, in order to request LaTeX on the wiki?
23:47:12 <Tim__> Yes, but only because they behaved like that in the original it does not mean we can't do it better, does it?
23:49:01 <SmatZ> but there is general rule to touch these things as little as possible
23:50:10 <Eddi|zuHause3> i assume this'd be an action 0 in newgrf
23:50:24 <Tim__> Hm... Just wondering - Would that be hard to change, or just 2 lines of code?
23:50:40 <Eddi|zuHause3> see my previous line ;)
23:52:52 <SmatZ> I think there is a flag stating that industry can never close down
23:52:59 <SmatZ> do you could either remove that flag
23:53:05 <SmatZ> that would be easy change
23:53:25 <SmatZ> but then power plants would die as often as other industries
23:53:33 <SmatZ> another solution would be to create another flag
23:53:42 <SmatZ> to "don't die that often"
23:53:58 <SmatZ> it is quite possible I am wrong, Belugas is the industry guru ;)
23:54:21 <Tim__> That would be enough, as i don't see a reason why they should die less often than normal industries...
23:55:52 <SmatZ> so maybe change INDUSTRYLIFE_BLACK_HOLE to INDUSTRYLIFE_PROCESSING at build_industry.h : 1217
23:56:00 <SmatZ> maybe it would work, maybe not
23:56:21 <Tim__> Come on Smatz, noone will ever notice if you do it :D
23:56:32 <Tim__> Well, except for the changelog
23:56:46 <SmatZ> maybe if I changed coding style
23:56:53 <Scaevolus> subtropical wood makes more money than valuables :O
23:56:57 <Tim__> But people only read the "Feature"-ones anyway
23:57:10 <Tim__> Or just make 2 changes at once and only document one of them *hehe*
23:57:30 <Eddi|zuHause3> for the record, i vote against doing this in the code
23:57:44 <SmatZ> other devs read commit changes :-P
23:58:09 <Tim__> Yeyey, i'm already searching for that Newgrf-tutorial...
23:59:33 <Tim__> Anyone has a link to one? Can't really find it -.-
continue to next day ⏵