#gemrb@irc.freenode.net logs for 11 Mar 2011 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[01:09:03] <MikeChelen> ok cool, let me try that
[01:15:49] <MikeChelen> fuzzie: still won't die
[01:16:21] <MikeChelen> code should be
[01:16:22] <MikeChelen> actor->WalkTo(p, 0, distance);
[01:16:24] <MikeChelen> right?
[01:25:01] <MikeChelen> they can be killed with ctrl+y though
[02:25:23] <-- edheldil_ has left IRC (Ping timeout: 255 seconds)
[04:44:43] <-- MikeChelen has left IRC (Ping timeout: 248 seconds)
[06:59:20] --> Bo_Thomsen has joined #GemRb
[07:07:56] --> edheldil_ has joined #GemRb
[07:08:12] --- Akileos-ZzZ is now known as Akileos
[07:08:14] <Akileos> Hi hi
[07:34:21] <tomprince> untested: https://github.com/tomprince/gemrb/commits/files
[07:35:36] <Akileos> How can test ?
[07:36:57] <tomprince> Probably nothing interesting for you to test. I just haven't gotten around to running the code yet.
[07:37:21] <tomprince> You could fetch the code and build+run it.
[07:37:45] <tomprince> Mostly I want the reaction of fuzzie's nose. ;)
[07:40:40] <Akileos> Is it supposed to decrease memory footprint ?
[07:42:36] <tomprince> No, it is moslty code reorgnaization. There should be little to no user visible differences.
[07:45:41] <Akileos> Last time I asked nobody answer, are there still some easy todos ?
[07:50:30] <tomprince> Most of the todos require knowledege of IE internals.
[07:51:01] <Akileos> Too bad
[07:51:16] <Akileos> Sor far the only thing I managed to understand is .CRE file format
[07:51:24] <tomprince> A while ago, I partially implemented a TIZ file importer, (which is used in some BG2 mods)
[07:51:37] <tomprince> I never did get around to getting to a mergeable state.
[07:52:05] <Akileos> Wish I could make music play on my mac, but don't even know where to start
[07:59:17] <-- edheldil_ has left IRC (Ping timeout: 255 seconds)
[08:01:15] <Akileos> [OpenAL]: Unable to create music source: 40963 [ERROR]
[08:01:15] <Akileos> Playing: THEME/THEMEA
[08:01:15] <Akileos> [OpenAL]: Failed to stop source: 40963 [WARNING]
[08:01:27] <-- Bo_Thomsen has left IRC (Quit: Leaving.)
[08:01:37] <Akileos> MUSImporter]: Loading /Volumes/BG/BG2/music/Cred.mus...[FOUND]
[08:01:38] <Akileos> [ResourceManager]: Searching for Combo1/Combo1B... Combo1/Combo1B.acm...[Music]
[08:01:38] <Akileos> [OpenAL]: Unable to create music buffers: 40963 [ERROR]
[08:01:38] <Akileos> Playing: Combo1/Combo1B
[08:01:38] <Akileos> [OpenAL]: Unable to query music source state: 40963 [ERROR]
[08:24:26] --> lynxlynxlynx has joined #GemRb
[08:24:26] <-- lynxlynxlynx has left IRC (Changing host)
[08:24:26] --> lynxlynxlynx has joined #GemRb
[08:24:26] --- ChanServ gives channel operator status to lynxlynxlynx
[08:29:29] --> lubos has joined #GemRb
[08:32:38] <-- Drakkar has left IRC (Read error: Connection reset by peer)
[08:33:06] --> Drakkar has joined #GemRb
[08:35:12] <tomprince> A *large*, but not requiring knowledge of IE internals project is designing and implementing a better system for logging.
[08:49:01] --> adominguez has joined #GemRb
[08:49:10] <Akileos> What about my openAL issue ? gemrb/plugins/OpenALAudio ?
[09:07:45] <tomprince> I don't know much about openal. Do you have any other program you can try with openal?
[09:08:26] <Akileos> Well everything works so far, except music
[09:08:35] <Akileos> foosteps, sfx, video's sound
[09:08:44] <Akileos> ambience snd
[09:09:54] <tomprince> I wonder if it is some sort of threading issue.
[09:10:32] <Akileos> int OpenALAudioDriver::CreateStream(Holder<SoundMgr> newMusic)
[09:10:32] <Akileos> [....]
[09:10:38] <Akileos> if (MusicSource == 0) {
[09:10:38] <Akileos> alGenSources( 1, &MusicSource );
[09:10:38] <Akileos> if (checkALError("Unable to create music source", "ERROR")) {
[09:12:25] <tomprince> What happens if you remove the '#if 0's in OpenAlAudio/StackLock.cpp
[09:14:05] <tomprince> That is, what pattern is there to the locking messages.
[09:14:39] <Akileos> No error, no music
[09:14:57] <Akileos> " if (MusicSource == 0 && false) {
[09:15:43] <Akileos> Well. seems to be stange
[09:15:44] <Akileos> // Free old MusicReader
[09:15:44] <Akileos> MusicReader = newMusic;
[09:15:45] <Akileos> if (!MusicReader) {
[09:15:45] <Akileos> MusicPlaying = false;
[09:15:45] <Akileos> }
[09:15:55] <tomprince> No don't do that. remove the #if 0 and #endif in StackLock.cpp
[09:15:57] <Akileos> But never creating one if not existing
[09:16:14] <Akileos> Ho, scuse me, hold on
[09:19:55] <Akileos> Could it be that OpenAL can't decode ACM ?
[09:20:36] <tomprince> OpenAL shouldn't be seeing acm.
[09:20:53] <edheldil> we have to decode ACM ourselves, because it's a proprietary format
[09:21:23] <Akileos> MUSImporter]: Loading /Volumes/BG/BG2/music/THEME.MUS...[FOUND]
[09:21:23] <Akileos> [ResourceManager]: Searching for THEME/THEMEA... THEME/THEMEA.acm...[Music]
[09:21:24] <Akileos> Locking mutex musicMutex in CreateStream()
[09:21:24] <Akileos> [OpenAL]: Unable to create music source: 40963 [ERROR]
[09:21:33] <edheldil> and openal bothers with sound patches only anyway, I think
[09:21:35] <Akileos> ressourcemanager talking about ACM
[09:22:37] <edheldil> yes, but iirc fuzzie said that it's the music source number, which is suspicious
[09:23:58] <fuzzie> wjp asked some smart question, might be out of sources
[09:27:04] <fuzzie> tomprince's branch seems sensible at a glance
[09:27:18] <fuzzie> but let me wander off and get coffee first.
[09:28:06] <Akileos> Coffee ??
[09:28:16] <Akileos> Gimme some !!!! :D
[09:29:11] <tomprince> 40963 = 0xA003 = AL_INVALID_VALUE
[09:29:39] <wjp> that was not the first error in there, IIRC
[09:30:03] <wjp> wasn't there one about buffers earlier in the log?
[09:31:12] <tomprince> Somebody should change the %d to %x in checkALError
[09:31:52] <wjp> these particular ones are easy enough to read from the decimal too (due to the 0xA...) but, yes
[09:33:53] <wjp> now I can't test if it actually compiles here...
[09:34:29] <Akileos> Filename + line ?
[09:35:15] <wjp> ...but I'm just going to push it anyway
[09:35:23] <Akileos> I cant try it here
[09:35:25] <Akileos> can
[09:35:36] <wjp> that is, if I could actually push here
[09:35:39] <tomprince> I beat you to it.
[09:35:52] <wjp> ok :-)
[09:35:56] <tomprince> Although it didn't show up.
[09:36:13] <wjp> you missed one
[09:36:29] <wjp> 11 lines below
[09:36:56] <tomprince> I wonder if the post-commit hook should be run in the background.
[09:37:12] <tomprince> Would make commits faster, and make the script harder to interrupt.
[09:38:23] <CIA-27> GemRB: 03tom.prince * rf1f17ae93925 10gemrb/gemrb/plugins/OpenALAudio/OpenALAudio.cpp:
[09:38:23] <CIA-27> GemRB: OpenALAudio: Change other error code printing to hex.
[09:38:23] <CIA-27> GemRB: Signed-off-by: Tom Prince <tom.prince@ualberta.net>
[09:38:28] <Akileos> [OpenAL]: Unable to create music source: a003 [ERROR]
[09:38:29] <Akileos> Playing: THEME/THEMEA
[09:38:29] <Akileos> [OpenAL]: Failed to stop source: a003 [WARNING]
[09:38:36] <Akileos> changed the two lines
[09:40:52] <wjp> but wasn't there an error about buffers earlier in the logs?
[09:41:02] <wjp> I seem to remember one last time in your pastebinned log
[09:43:21] <tomprince> Well, I am going to bed. I'll probably get a chance to test out my branch tomorrow night, after fuzzie gets a chance to tear it to shreads.
[09:43:29] <tomprince> ...... which is a good thing. :)
[09:43:30] <fuzzie> the branch doesn't seem to touch anything functional
[09:44:02] <fuzzie> i mean, anything which could be a problem for situations/issues we haven't yet encountered
[09:44:20] <fuzzie> and that code was a mess, so i imagine you will return to cheering, but will be sure to have a look :)
[09:44:35] <tomprince> It shouldn't, except maybe in some corner cases. But I haven't actually run any of that code. :)
[09:44:43] * tomprince sleeps
[09:46:46] <Akileos> http://pastebin.com/wcXBGhPU
[09:48:12] <wjp> #
[09:48:14] <wjp> [OpenAL]: Unable to create music buffers: a003 [ERROR]
[09:48:17] <Akileos> 226
[09:48:20] <wjp> is the first openal error in there
[09:48:20] <fuzzie> 30 streams sounds ok
[09:48:29] <wjp> yes, that's the requested number
[09:48:44] <wjp> but the error seems to occur with buffers
[09:50:01] <fuzzie> 8192 samples is a fairly tiny buffer, too
[09:52:31] <fuzzie> so, threading issue?
[09:52:48] <wjp> we do ask for 10 buffers, but that is still tiny
[09:52:52] <wjp> (IIRC)
[09:52:59] <wjp> so, yes, maybe
[09:53:19] <wjp> although this is the same error as last time
[09:54:04] <-- |Cable| has left IRC (Remote host closed the connection)
[09:56:43] <fuzzie> almost everyone is using openal soft, though
[09:57:07] <fuzzie> and it has been pretty reliable for a while if you avoid the pulse backend
[09:57:51] <wjp> is this openal soft too?
[09:58:15] <wjp> or does macports/os x/...? still use the creative one?
[09:58:19] <fuzzie> oh, i suppose this is OS X
[09:58:22] <fuzzie> umm
[09:58:25] <fuzzie> good question
[09:58:47] <fuzzie> Akileos is using the macports one, we discussed this before
[09:59:19] <fuzzie> but they are pulling the MacOSX1-3_Spec1-1 from Creative's svn
[10:00:48] <fuzzie> and not building the soft version, i guess
[10:01:12] <Akileos> > port list openal
[10:01:12] <Akileos> Warning: port definitions are more than two weeks old, consider using selfupdate
[10:01:12] <Akileos> openal @1.3 audio/openal
[10:02:56] <fuzzie> no recent work on Creative's trunk for OS X though.
[10:03:14] --> edheldil_ has joined #GemRb
[10:09:12] <Akileos> Any SOAP Webservice guru in there ?
[10:11:17] <wjp> only slightly off-topic :-)
[10:13:30] <lynxlynxlynx> reading the logs, there's plenty of work to be done and simple todos can be split out of that too
[10:13:59] <lynxlynxlynx> you could also just grep for TODO/FIXME, but a more concentrated approach would be better
[10:15:38] <Akileos> I'm sometime supposed to be doing actual work you know ;)
[10:19:45] <edheldil> also peruse the bugtracker. And there's still plenty of work in Guiscripts, and it's easily aproachable
[10:21:32] <lynxlynxlynx> yep
[10:21:54] <lynxlynxlynx> a lot of boring stuff could be done to pst
[10:22:09] <lynxlynxlynx> since that code can't be merged with the rest so easily
[10:22:59] <edheldil> is it so bad? The guiscripts once used to work, but I guess the API changed a lot since then
[10:23:44] <edheldil> Akileos: and there is webdesigner job open ;-)
[10:23:58] <fuzzie> the guiscripts for pst were always buggy
[10:24:07] <fuzzie> and there is the mess of levelup for pst too, i assume
[10:27:19] <edheldil> also, afaik iwd2 is even more work, is not it?
[10:27:42] <fuzzie> iwd2 needs core work first, i think?
[10:47:30] <Akileos> edheldil: I'm more into VMWare, Cisco and so on; some PHP/Perl/C/Java for admin tasks
[10:49:48] <lynxlynxlynx> edheldil: work and work perfectly are two things ;)
[10:51:47] <Akileos> So need to crawl megaupload for iwd2 isos
[10:52:37] <wjp> IWD2 is easily available for sale
[10:52:46] <fuzzie> more to the point, cheap :P
[10:52:57] <wjp> and that :-)
[10:53:15] <Akileos> Well, got to get out, that's hard
[10:53:16] <Akileos> :D
[10:53:27] <wjp> www.gog.com sells a drm-free download
[10:54:51] <fuzzie> which is somewhat less cheap, but i guess not so much they can do about that
[11:44:06] --> Maighstir has joined #GemRb
[13:23:07] <-- lynxlynxlynx has left IRC (Ping timeout: 248 seconds)
[13:23:43] <edheldil> btw, what is the goal of tomprince's patch? I saw it's replacing the CachedFileStream with a SlicedFileStream or st. like that, but the diff is not really helping
[13:24:41] <fuzzie> de-duplicating code
[13:24:55] <fuzzie> i think
[13:25:52] <fuzzie> it makes more sense if you start at the first patch :)
[13:34:00] <edheldil> oh, ty
[13:40:57] --> lynxlynxlynx has joined #GemRb
[13:40:57] <-- lynxlynxlynx has left IRC (Changing host)
[13:40:57] --> lynxlynxlynx has joined #GemRb
[13:40:57] --- ChanServ gives channel operator status to lynxlynxlynx
[14:07:40] --> PixelScum has joined #GemRb
[14:08:48] <-- adominguez has left IRC (Quit: Saliendo)
[14:09:28] <-- Drakkar has left IRC (Ping timeout: 250 seconds)
[14:47:31] --> MikeChelen has joined #GemRb
[16:06:13] --> pupnik has joined #GemRb
[16:06:13] <-- pupnik has left IRC (Changing host)
[16:06:13] --> pupnik has joined #GemRb
[16:10:01] <-- pupnik_ has left IRC (Ping timeout: 260 seconds)
[16:17:11] <-- Demitar has left IRC (Quit: Burn the land and boil the sea, you can't take the sky from me!)
[16:30:25] --> Avenger has joined #GemRb
[16:30:26] --- ChanServ gives channel operator status to Avenger
[16:30:35] <Akileos> Hi Avenger
[16:33:15] <Avenger> hello
[16:33:22] <pupnik> hola
[16:45:02] <edheldil> Hi, Avenger, I have a question: did you have any reason to use REALMS2 for movie subtitles font in iwd/how? at least here it looks like NORMAL
[16:48:29] <Avenger> i don' know, change it, check it, if it is more like the original commit it
[16:48:58] --> |Cable| has joined #GemRb
[16:49:00] <Avenger> just don't commit it without checking :)
[16:49:44] <Avenger> maybe it was ugly because of lack of palette, or maybe it was too large to fit
[16:49:46] <-- lubos has left IRC (Quit: Leaving.)
[16:52:21] <Akileos> Got IWD2 ;) Now have to igure out how to "install" and use a separte config file
[16:56:25] <edheldil> Avenger: well, REALMS2 is way too big and without accented chars, NORMAL looks just like in the orig game I have. But I only have a translated version ...
[16:59:44] <Avenger> then it is fine
[17:10:21] <pupnik> game controller for Nokia N8 http://twitpic.com/3bewqb
[17:11:04] <-- lynxlynxlynx has left IRC (Ping timeout: 250 seconds)
[17:25:27] --> lynxlynxlynx has joined #GemRb
[17:25:27] <-- lynxlynxlynx has left IRC (Changing host)
[17:25:27] --> lynxlynxlynx has joined #GemRb
[17:25:27] --- ChanServ gives channel operator status to lynxlynxlynx
[17:32:43] <-- edheldil_ has left IRC (Ping timeout: 248 seconds)
[18:04:14] <-- |Cable| has left IRC (Read error: Connection reset by peer)
[18:06:02] --> kettuz has joined #GemRb
[18:06:14] --> |Cable| has joined #GemRb
[18:07:21] <tomprince> patch refreshed with a couple of fixes.
[18:08:12] <tomprince> I think it is almost possible to remove the FILE emulation from VFS.
[18:08:43] <tomprince> Just need to change LoadConfig/INI and CacheFile to using FileStreams.
[18:09:19] <tomprince> Then the emulation can just be folded into FileStream directly.
[18:11:05] <lynxlynxlynx> btw, did anything else bother you about the mac patch?
[18:11:35] <lynxlynxlynx> i'm about to merge it without the visibility flag
[18:15:00] <tomprince> No, other than I would split it in two.
[18:15:55] <lynxlynxlynx> yeah and trailling whitespace
[18:16:35] <tomprince> That is what 'git config apply.whitespace fix' and 'git add -p' are for. :)
[18:17:34] <fuzzie> I wonder what 'fix' *does* there.
[18:17:56] <tomprince> I think I've fixed all the facepalm bugs with my code. All that is left, I think is to try to test all the corner cases.
[18:19:07] <tomprince> Look at core.whitespace in git-config(1) for what it considers error.
[18:19:29] <CIA-27> GemRB: 03lynxlupodian * r10884371c221 10gemrb/ (CMakeLists.txt gemrb/CMakeLists.txt): support static linking on mac, patch by Hana Dus??kov??
[18:19:33] <CIA-27> GemRB: 03lynxlupodian * r4c146188a246 10gemrb/gemrb/plugins/SDLAudio/CMakeLists.txt:
[18:19:33] <CIA-27> GemRB: SDLAudio: also search for includes in the sdlmixer include dir
[18:19:33] <CIA-27> GemRB: patch by Hana Dus??kov??
[18:19:43] <fuzzie> The manpage here doesn't specify what it fixes, though.
[18:21:00] <tomprince> Well, it fixes everything it detects as errors.
[18:21:12] <fuzzie> well, mostly i am curious if it's smart enough not to fix space-before-tab
[18:21:40] <tomprince> not fix it?
[18:21:47] <fuzzie> it breaks patch files
[18:21:54] <fuzzie> but i am running a relatively old git here, 1.7.1
[18:23:04] <fuzzie> i shall go forth and try it
[18:24:59] <tomprince> Doesn't seem like it.
[18:25:55] <tomprince> But, you can configure it if you are dealing with a bunch of patches in your tree.
[18:26:21] <fuzzie> well, at some point i was experimenting with the idea of gemrb just patching bits of original data
[18:26:29] <fuzzie> which is why i'm thinking about it
[18:28:07] <tomprince> It would be nice if gemrb didn't have to deal with all the broken bits of the original data.
[18:28:32] <fuzzie> well, to some extent we are stuck with it, if we want mod compatibility
[18:29:34] <tomprince> Well, how many mods actually have broken files.
[18:30:09] <fuzzie> a good question
[18:31:01] <tomprince> If we are able to patch all the original data, I don't think it would be unreasonable to just say we don't support broken data.
[18:31:33] <tomprince> I am thinking of incomplete files, or scripts that don't parse properly or the like.
[18:31:48] <-- |Cable| has left IRC (Remote host closed the connection)
[18:32:24] <fuzzie> the fixpacks usually do a rather wonderful job of patching up the script parsing
[18:33:11] --> |Cable| has joined #GemRb
[18:33:23] <-- kettuz has left IRC (Quit: Leaving)
[18:41:23] <lynxlynxlynx> oh
[18:43:15] <lynxlynxlynx> do you have the bg1 fixpack unpacked too?
[18:43:31] <lynxlynxlynx> i wonder if they fix landri.dlg
[18:44:25] <lynxlynxlynx> http://forums.gibberlings3.net/index.php?showtopic=21418&st=30&#entry179720
[18:51:09] <fuzzie> no idea
[18:51:55] <fuzzie> hate those
[19:29:58] <-- MikeChelen has left IRC (Remote host closed the connection)
[19:54:17] --> edheldil_ has joined #GemRb
[20:48:30] <-- Avenger has left IRC (Ping timeout: 246 seconds)
[20:49:12] <tomprince> win32 q: What type is HANDLE, and can I treat it as a void*, and is NULL an empty value?
[20:51:17] <fuzzie> i think it's 32/64 bit according to platform, so i guess you can treat it as void*
[20:51:22] <tomprince> re: files-win32 branch on github
[20:52:09] <fuzzie> i don't have the includes handy to look at whether it actually is a void*
[20:52:24] <fuzzie> but i think it's been a long time since any handles were actually pointers, rather than just indexes..
[20:53:16] <fuzzie> i think to be safe you should use -1 for empty but looking at the documentation for the functions in question is no doubt a far better idea than asking me
[20:54:22] <tomprince> I am just wondering if I can use void* in the FileStream header, rather than having a fake _FILE struct.
[20:55:44] <fuzzie> and there's certainly no lack of code assuming that NULL is an invalid handle so i suppose it can't be that unsaf :)
[20:55:47] <fuzzie> e
[21:02:05] --> Avenger has joined #GemRb
[21:02:05] --- ChanServ gives channel operator status to Avenger
[21:37:18] <edheldil_> tomprince: actually msdn documentation is not so bad
[21:51:45] <Avenger> japan is quite quake proof, a 8.8 quake caused only about 200 casualties. That's impressive building.
[21:54:01] <lynxlynxlynx> when you have big earthquakes so often, you prepare well
[21:57:04] <-- Avenger has left IRC (Quit: bye!)
[21:57:57] <pupnik> amazing footage on youtube
[22:14:56] <edheldil_> is the deathtoll final yet? I suspect that there might be many more dead in the tsunami
[22:41:32] <tomprince> Can someone with win32 try my files-win32 branch?
[22:41:38] <tomprince> I like this solution much better.
[22:47:08] <fuzzie> hm
[22:48:34] <fuzzie> do we use the result of the seek stuff anywhere?
[22:49:26] <fuzzie> i was going to say that your error checking there is bad, but since nothing seems to care, i guess it's meaningless anyway
[22:51:01] <fuzzie> (I would also very much prefer FileStream::File to be in seperate .cpp files rather than #ifdeffed, but I realise that is pretty picky.)
[22:51:12] <fuzzie> (And I have not even x86 here to test, sorry.)
[22:52:34] <tomprince> We could certainly do that, although that would require adding cmake code to condionally compile code. Also, everything should get inlined this way.
[22:58:55] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[22:58:58] <fuzzie> i'm not sure what the fscanf you removed in config was meant to do
[22:59:16] <tomprince> It was to read to the end of the line, I think.
[22:59:54] <tomprince> To skip over comments. Which I don't need to do, since I read the line in first.
[23:00:45] <fuzzie> ah. so this is much nicer your way.
[23:02:28] <fuzzie> i am tired, so pondering out loud; ReadLine swallows \r and \n, but LoadINI() still looks for them in the return result
[23:03:28] <fuzzie> and it does not look for nulls, but this is not interesting anyway since the code just ignores lines which don't match the sscanf there
[23:05:16] <fuzzie> also the printf in CacheFile still says 'CachedFileStream' unless you changed it in a later commit, in the 'irrelevant picky things' column
[23:06:03] <tomprince> That I did fix later.
[23:06:17] <tomprince> I think you are right about the ini.
[23:09:47] <fuzzie> you can perhaps tell i am reading this from the web ui :) with that in mind, your patches look great
[23:56:52] --> Bo_Thomsen has joined #GemRb