#exult@irc.freenode.net logs for 13 Dec 2012 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage


[00:03:45] <sh4rm4> cmake is a pita
[02:05:33] <-- Dominus has left IRC (Ping timeout: 265 seconds)
[02:06:07] --> Dominus has joined #exult
[02:06:09] --- ChanServ gives channel operator status to Dominus
[02:49:32] <-- Dominus has left IRC (Ping timeout: 265 seconds)
[02:50:07] --> Dominus has joined #exult
[02:50:07] --- ChanServ gives channel operator status to Dominus
[04:12:47] --> i30817 has joined #exult
[04:14:27] <i30817> Hi. There is a problem with the configure.ac
[04:14:28] <i30817> It's the same as this: http://stackoverflow.com/questions/6128987/why-doesnt-autoconf-pass-the-ac-check-header-test-when-the-h-is-file-clearly-a
[04:14:30] <i30817> (check the end of the page for the solution).
[04:14:31] <i30817> The script is saying the language is 'c' so any header files of c++ libraries fail.
[04:15:41] <i30817> So my diff will remove that stray
[04:15:43] <i30817> "AC_LANG_CPLUSPLUS
[04:15:44] <i30817> AC_CHECK_HEADERS(hash_map hash_set ext/hash_map ext/hash_set tr1/unordered_map tr1/unordered_set sstream)
[04:15:46] <i30817> AC_LANG_C
[04:15:47] <i30817> ...(the rest)"
[04:15:49] <i30817>
[04:15:50] <i30817> Since it doesn't make sense anyway since exult is pretty much c++
[04:16:21] <i30817> I thought i'd mention it since it seems a pretty obvious bug.
[04:16:44] <i30817> Yes dominus, i will do a bug report
[04:56:51] --> DominusExult has joined #exult
[04:56:52] --- ChanServ gives channel operator status to DominusExult
[04:58:21] <-- Dominus has left IRC (Ping timeout: 276 seconds)
[04:58:21] --- DominusExult is now known as Dominus
[05:20:05] <i30817> What is this 'pentagram in exult' ifdef?
[08:17:38] <-- i30817 has left IRC (Ping timeout: 244 seconds)
[08:21:17] --> Rottingbeef has joined #exult
[08:27:50] --> i30817 has joined #exult
[08:32:15] <-- Dominus has left IRC (Ping timeout: 265 seconds)
[08:33:16] --> Dominus has joined #exult
[08:33:16] --- ChanServ gives channel operator status to Dominus
[09:47:15] <-- Kirben has left IRC ()
[09:57:04] <i30817> Made exult use the static library.
[09:57:05] <i30817> How, it's really very innacurate
[09:57:21] <i30817> Anyway; do you want the patch?
[09:58:17] <i30817> The problem with it is that the library is not available outside of my ppa ATM (though i had the idea that scummvm had used it as a library...)
[09:59:45] <i30817> Of course, i haven't touched the LowLevelMidiDriver... imagine that's where all the bugs are.
[10:00:57] <Dominus> yes, please your patch would be a good starting point on how to use the external lib
[10:02:06] <i30817> Let me upload it
[10:04:02] <i30817> http://bazaar.launchpad.net/~exult-team/exult-engine/packaging/view/head:/debian/patches/staticMt32Emu.diff
[10:07:02] <Dominus> looking good
[10:07:08] <i30817> Ah, you need to delete the audio/midi_drivers/mt32emu dir before compiling otherwise it gets confused with the
[10:07:10] <i30817> #include <mt32emu/mt32emu>
[10:07:27] <Dominus> aye, I see
[10:07:40] <i30817> Also need the library (it's not published to the ppa yet, but it's already building)
[10:08:02] <i30817> It's static so upgrades are less of a problem.
[10:08:40] <i30817> Pity the sound is so tinny.
[10:09:06] * Dominus will first have to successfully build munt lib for all three arches. struggled last night with it...
[10:10:07] <Dominus> I noticed something similar when I compared the fluidsynth soundfont use vs the direct coreaudio soundfont use. Same soundfont but very tiny with coreaudio. probably needs some more oomph somewhere...
[10:10:57] <i30817> What's the requirements to enable fluidsynth?
[10:11:10] <i30817> Have to mess with timidity.cfg?
[10:11:35] <i30817> Would be nice to support many soundfonts at once.
[10:11:44] <i30817> If that's happening.
[10:11:54] <Dominus> have fludisynth libs and headers
[10:12:09] <Dominus> then exult will use them automatically
[10:12:39] <Dominus> (or needs to be enabled in configure --enable-fluidsynth - but I think it's automatic)
[10:12:54] <Dominus> fluidsynth requires glib2, though
[10:14:15] <i30817> For building only? Because i just installed it and:
[10:14:16] <i30817> The following extra packages will be installed:
[10:14:18] <i30817> qsynth
[10:14:19] <i30817> Suggested packages:
[10:14:21] <i30817> jackd
[10:14:22] <i30817> The following NEW packages will be installed:
[10:14:24] <i30817> fluidsynth qsynth
[10:14:40] <Dominus> qsynth? no idea what that is
[10:15:18] <Dominus> fluidsynth will link against glib2 so it is needed for more than building
[10:15:26] <Dominus> *unfortunately*
[10:28:37] <i30817> libfluidsynth is a static library too?
[10:29:34] <i30817> Or normal? ( .a file == static, la or so == dynamic)?
[10:30:37] <i30817> Thinking of enabling it for the ppa.... Of course, i read the soundfonts are 128 mb so maybe not.
[10:31:26] <i30817> Still, there is one 'optional' which doesn't download by default.
[10:33:48] <i30817> BTW, in that patch, you'll not i disabled a
[10:33:50] <i30817> "AC_LANG_C" in the configure.ac
[10:33:51] <i30817>
[10:33:53] <i30817> That's because that was telling automake to consider the source as c instead of c++; so it wouldn't link against c++ libs. Considering exult is c++ , i made executive decision.
[10:34:07] <i30817> not/note/s
[10:37:06] <i30817> FluidSynth requires a 'fluidsynth_soundfont' setting
[10:37:07] <i30817> Ohohoh.
[10:42:40] <i30817> Still, i can work with this, put it in the default place for ubuntu on install.
[10:43:40] <i30817> what's the better soundfond GS or GM?
[11:07:37] <i30817> Think i can make a (ubuntu only) patch that will set those config settings with defaults so the user doesn't have to be bothered.
[11:18:17] --> Kirben has joined #exult
[11:18:18] --- ChanServ gives channel operator status to Kirben
[11:24:59] <Dominus> do you have to provide *everything* for this? I wouldn't bother with adding a soundfont and setting the path to it. When people use soundfonts they tend to use big, really good ones, that you probably wouldn't be allowed to add anyway
[11:26:29] <Dominus> how do I make cmake install stuff to a custom prefix. I got as far that it actually honors my custom prefix for building but not yet for installing the stuff
[11:29:26] <Dominus> ah, I see
[11:29:42] <Dominus> CMAKE_INSTALL_PREFIX
[11:32:54] <-- sh4rm4 has left IRC (Ping timeout: 276 seconds)
[11:40:58] <i30817> Just thought of something... lol i guess only i will be able to use the mt32emu library. I don't think it will build on the launchpad buildbots, because the lib is not on the repositories.
[11:41:18] <i30817> Doh.
[11:43:46] <Dominus> LOL
[11:43:56] <Dominus> sorry :)
[11:44:17] <Dominus> didn't you notice that with dosbox before?
[11:47:51] <Dominus> still useful that you de-hardcoded munt from exult...
[11:49:09] <Dominus> but sucks that you spent so much time on it...
[11:50:10] * Dominus has cmake built of munt ready, now only need to make a big script for building Exult with it (while updating everything...)
[11:58:31] <i30817> In dosbox case it was not a library, but plopping the source code into the build and adjusting the make file 'on the fly'
[12:00:56] <Dominus> well, you could do that for exult, too, couldn't you?
[12:00:57] <i30817> Still, if it helps i will benifit on the end.
[12:01:44] <i30817> It's harder because exult makefiles are more complex.
[12:02:47] <i30817> But still possible.
[12:04:58] <Dominus> I'd say wait for what wjp and marzo say about your current de-hardcoding and how we should go about it
[12:05:58] <Dominus> when that gets accepted you may not have to change too much and you safe some work not having to do it for current Exult and then again for a changed one
[12:06:21] * Dominus is not going to be able to test this patch for some more hours
[12:07:54] <i30817> BTW, you don't need to build all of munt, just the mt32emu folder.
[12:08:05] <i30817> That's the library
[12:15:39] <i30817> Also the library is only a build time dependency.
[12:15:41] <i30817> Seeing as it as static library.
[12:34:23] <-- Kirben has left IRC ()
[12:48:53] --> TheCycoONE has joined #exult
[12:56:07] <-- Colourless has left IRC (Ping timeout: 244 seconds)
[14:08:13] <Dominus> cmake is driving me nuts :)
[14:13:40] <TheCycoONE> but... cmake is excellent
[14:24:34] --> Matt_O has joined #exult
[14:29:04] --> Marzo_android has joined #exult
[14:31:45] <Dominus> hmm, Marzo, Marzo_android: it seems when one starts Exult with mt32 as midi driver it segfaults here. Not sure if that is something you did... will investigate
[14:32:33] <Dominus> the exult menu doesn't like the mt32emu it seems
[14:34:30] <Dominus> hmm, seems to be that way for a long time already..
[14:34:44] <Eviltar> it was working a while ago
[14:35:02] <Eviltar> around the 1.5 release im sure
[14:35:26] <Dominus> yeah, but all my snapshots seem to fail
[14:36:12] <-- Marzo_android has left IRC (Ping timeout: 265 seconds)
[14:36:47] <Eviltar> when you guys released 1.5 i used that to build it for the xbox360 and use the mt32emu, that much i remember
[14:37:24] <Dominus> I'm thinking it might be limited to OS X 10.8
[14:37:30] <Eviltar> because i started on an older 1.42 branch that had sdl_mixer removed
[14:38:22] <Eviltar> i noticed that pentagram's makefile seems a lot more simple than exults
[14:38:44] <Eviltar> unlrelated
[14:39:07] <Eviltar> do you have the mt32emu roms in place?
[14:40:17] --> Marzo_android has joined #exult
[14:41:37] <Dominus> yes and it works when I switch in game
[14:42:41] <Marzo_android> Dominus: can you try it in gdb and post a backtrace?
[14:43:32] <Dominus> yes, will do
[14:51:48] <-- Marzo_android has left IRC (Ping timeout: 264 seconds)
[14:52:09] --> Marzo_android has joined #exult
[14:58:04] <Dominus> odd, there is something fishy but I don't know what
[15:02:29] <Dominus> ok, sorry, found half of the culprit
[15:02:54] <Dominus> some process locked the roms apparently, when I change the data folder to something else it works again
[15:03:23] <Dominus> of course i30817's patch is the main culprit I think
[15:04:01] <Dominus> I guess that locked the files somewhere and now I can't unlcok it
[15:06:39] <-- Marzo_android has left IRC (Ping timeout: 260 seconds)
[15:07:02] --> Marzo_android has joined #exult
[15:10:20] <-- Marzo_android has left IRC (Read error: Connection reset by peer)
[15:13:38] --> Marzo_android has joined #exult
[15:14:09] <-- Marzo_android has left IRC (Read error: Connection reset by peer)
[15:15:04] --> Marzo_android has joined #exult
[15:20:07] <i30817> You've got a point. I don't think there is any reason to hold the FP open after getting the data from the roms although the interface has a close method.
[15:21:45] <i30817> Just close the file where is is opened and delete the FILE *fp; and the body of close()
[15:22:14] <i30817> (close the file before both returns of couse)
[15:22:21] <-- Marzo_android has left IRC (Ping timeout: 276 seconds)
[15:23:37] --> Marzo_android has joined #exult
[15:25:50] <-- Marzo_android has left IRC (Read error: Connection reset by peer)
[15:26:05] --> Marzo_android has joined #exult
[15:30:36] <-- Marzo_android has left IRC (Ping timeout: 252 seconds)
[15:31:20] --> Marzo_android has joined #exult
[15:35:53] <-- Marzo_android has left IRC (Ping timeout: 255 seconds)
[15:38:12] --> Marzo_android has joined #exult
[16:02:12] <-- Marzo_android has left IRC (Ping timeout: 244 seconds)
[16:30:06] --> Colourless has joined #exult
[16:30:06] <-- Colourless has left IRC (Changing host)
[16:30:06] --> Colourless has joined #exult
[16:30:06] --- ChanServ gives channel operator status to Colourless
[16:31:37] <-- i30817 has left IRC (Quit: ChatZilla 0.9.88.2 [Firefox 17.0.1/20121129151842])
[17:33:07] --> i30817 has joined #exult
[17:34:34] <i30817> Here is a version of the patch that should not leave your Roms locked:
[17:34:36] <i30817> http://bazaar.launchpad.net/~exult-team/exult-engine/packaging/view/head:/debian/patches/staticMt32Emu.diff
[17:35:20] <i30817> Haven't test though (the svn is uncompilable with my debian dir now, because exult-studio is broken).
[17:35:43] --> sh4rm4 has joined #exult
[17:54:03] <i30817> Or actually, just fails with extreme optimization options.
[18:01:41] <-- i30817 has left IRC (Ping timeout: 255 seconds)
[18:11:13] --> i30817 has joined #exult
[18:12:17] <Dominus> thamks I'll give it a try later, when I have time.
[18:12:36] <Dominus> just don't build studio by default :)
[18:16:40] <Dominus> also I think the built in munt was louder than the external one. could be that newer munt honors the volume set by the sysex (more) and we used to set a hgher base volume somewhere
[18:49:21] <Dominus> still need to find the process locking the files...
[19:06:15] <-- Dominus has left IRC (Quit: good night)
[19:07:35] --> Dominus has joined #exult
[19:07:35] --- ChanServ gives channel operator status to Dominus
[19:09:17] <Dominus> god dammn it, what does munt do with the files? where does it cache it... I CANNOT anymore run exult with the default data folder....
[19:09:27] <Dominus> even after a restart
[19:12:19] <i30817> munt does not cache files anymore.... unless you build with the old dir still there.
[19:12:34] <i30817> It's all in memory now
[19:18:44] <Dominus> it borked *something*
[19:26:46] <-- Dominus has left IRC (Quit: good night)
[19:28:03] --> Dominus has joined #exult
[19:28:03] --- ChanServ gives channel operator status to Dominus
[19:42:54] <-- i30817 has left IRC (Remote host closed the connection)
[20:25:50] --> TheCycoTWO has joined #exult
[20:34:43] <-- TheCycoONE has left IRC (*.net *.split)
[20:51:48] <Dominus> other than the file lock issue I had, the mt32emu patch is working nicely here
[20:53:03] <Dominus> wjp, maybe you can review it and then we need to decide how to handle this. Colourless' input would be appreciated as well. After all he is our audio engine master :)
[22:05:03] <-- Rottingbeef has left IRC ()
[22:07:54] <-- TheCycoTWO has left IRC (Quit: And then there were n-1)
[22:45:32] --> Dominus_ has joined #exult
[22:45:54] <-- Dominus_ has left IRC (Remote host closed the connection)
[23:11:58] --> Kirben has joined #exult
[23:11:58] --- ChanServ gives channel operator status to Kirben
[23:49:04] <-- nutron has left IRC (Quit: I must go eat my cheese!)