#uwadv@irc.freenode.net logs for 19 Nov 2002 (GMT)

Archive Today Yesterday Tomorrow
Underworld Adventures homepage


[01:46:30] --> Coren_ has joined #uwadv
[01:46:34] * Coren_ waves.
[01:57:53] <-- Eld^zz has left IRC (Read error: 113 (No route to host))
[02:06:05] --> QQtis has joined #uwadv
[02:06:05] --- ChanServ gives channel operator status to QQtis
[02:26:57] <Coren_> Hey QQtis.
[02:27:09] <Coren_> You wanted to speak to me last night?
[04:14:23] --> QQtis_ has joined #uwadv
[04:15:15] * Coren_ pokes QQtis_
[04:16:46] <-- QQtis has left IRC (capek.freenode.net irc.freenode.net)
[05:15:22] <-- Coren_ has left IRC ("using sirc version 2.211+KSIRC/1.2.1")
[07:51:52] --> Eld^zz has joined #uwadv
[07:52:07] --- Eld^zz is now known as Eldron
[13:39:39] --> yot has joined #uwadv
[14:44:29] --> Eldron- has joined #uwadv
[15:11:32] --> Coren_ has joined #uwadv
[15:11:38] * Coren_ wavies.
[15:40:30] <-- Eldron- has left IRC ()
[15:52:07] --> wjp has joined #uwadv
[15:52:07] --- ChanServ gives channel operator status to wjp
[15:52:19] <wjp> hi
[15:58:58] <Coren_> Hey WIllem.
[15:59:07] * Coren_ is steal beating over this configure.ac
[15:59:28] <Coren_> It actually works, right now, but doesn't faoil when it should; I'm building in some smarts.
[16:02:16] <wjp> AC_PREREQ(2.54)?
[16:02:21] <wjp> that's pretty cutting-edge :-)
[16:02:52] <Coren_> Well, I presume I might want to bring this down some, I have no idea which features I am using that might require 2.54 or not.
[16:03:20] <wjp> I think 2.53 is where most current major distros are
[16:03:40] <Coren_> Hmm. I have a couple of -L and -I I want to use when AC_CHECK_LIBing, you know who I can do this?
[16:03:54] <Coren_> how, even
[16:04:36] <wjp> use the last argument of AC_CHECK_LIB?
[16:04:54] <wjp> I'm not sure if -L/-I are allowed there, but maybe :-)
[16:05:13] <Coren_> Hmm, doesn't that just add libraries to the link and not library locations?
[16:05:25] * Coren_ tries adding the -Ls preemptively to LIBS
[16:09:18] <wjp> that'll work
[16:09:42] <Coren_> Ah. That works.
[16:11:03] <Coren_> Ah, no-- configure adds the library it wants to test *before* $LIBS so it never sees the -L in time.
[16:11:16] <wjp> yes, that's true
[16:11:45] <Coren_> Hmmm. Maybe LDFLAGS will do.
[16:12:25] <wjp> LDFLAGS is passed to the linker before LIBS
[16:14:30] <Coren_> Ah, yes, that works.
[16:14:56] * Coren_ adds saving/restoring of ldflags
[16:16:03] <wjp> I guess this is one of those things that would warrant writing a new test :-)
[16:16:15] * Coren_ nods.
[16:16:21] <Coren_> If only I knew how to do that.
[16:16:25] * Coren_ reads the doc.
[16:16:28] <wjp> ;-)
[16:16:39] <wjp> uwadv uses CPPFLAGS, it seems
[16:16:56] <Coren_> Appropriately enough, there is a section "Writing Tests"
[16:17:13] <wjp> oh, but that's because linking doesn't use CXXFLAGS
[16:19:52] * wjp decides to ignore the docs for now and reads the sdl.m4 file
[16:20:15] <Coren_> That one is a bad example because it relies on an executable to supply paths.
[16:20:47] <wjp> yeah, but wrapping a test in another test is pretty easy, so you just need some basic syntax
[16:20:47] <Coren_> AC_DEFUN([AC_CHECK_LIB_PATH], [
[16:20:47] <Coren_> saved_ldflags="$LDFLAGS"
[16:20:47] <Coren_> LDFLAGS="$3$LDFLAGS"
[16:20:48] <Coren_> AC_CHECK_LIB($1, $2)
[16:20:48] <Coren_> LDFLAGS="$saved_ldflags"
[16:20:49] <Coren_> if x$have_lib$1 != xno
[16:20:50] <Coren_> then
[16:20:52] <Coren_> LIBS="$3$LIBS"
[16:20:54] <Coren_> fi
[16:20:56] <Coren_> ])
[16:20:57] <wjp> exactly
[16:21:04] <Coren_> That's mine right there. :-)
[16:21:10] * Coren_ tries it, now. :-)
[16:21:37] <wjp> no space between $3 and $LDFLAGS?
[16:21:57] <Coren_> That's because $3 is in the form " -L /foo/bar " by then.
[16:22:08] <wjp> k :-)
[16:23:56] <Coren_> fixed 'if x$foo' to 'if test x$foo'
[16:24:21] <Coren_> Hey, cool, that works. :-)
[16:29:35] <Coren_> New configure.ac commited.
[16:29:57] <Coren_> See if that will actually allow LoW to build unaided on your box?
[16:32:01] <Coren_> The ugliness with bison has been corrected, too.
[16:32:25] <yot> wjp
[16:33:59] <wjp> it doesn't know AC_GNU_SOURCE but that's probably because I have 2.53
[16:34:09] <wjp> hi yot
[16:34:09] <yot> eventually you landed at RE right? :>>>
[16:34:15] <yot> I told you 4 month ago :)
[16:34:21] <wjp> I landed at RE?
[16:34:27] <yot> reverse engineering
[16:34:34] <wjp> yeah, I still know what it means ;-)
[16:34:39] <wjp> I don't get what you mean, though
[16:34:51] <yot> devlog of 2002-11-04
[16:35:02] <Coren_> I'm not entirely certain AC_GNU_SOURCE is, in fact, required actually.
[16:35:24] <yot> or did I get something wrong?
[16:35:45] * wjp is pretty much confused now
[16:36:03] <yot> am I wrong?
[16:36:05] <yot> sorry then
[16:36:11] <yot> but that's the way I understood it
[16:36:11] <wjp> I have no idea what you're trying to say
[16:36:39] <yot> hmm?
[16:36:47] <Coren_> Ahhh. Coolness. -host=i686-pc-cygwin produces a working low.exe
[16:36:49] <wjp> Coren_: oh.. right... that glPointParameterfvEXT again
[16:37:01] <yot> Decompiled code also helps to understand meaning of most imported functions <- what does that mean then?
[16:37:11] <Coren_> Ah, damn. Forgot about that. Do you remember what define it wanted?
[16:37:17] <wjp> #define GL_GLEXT_PROTOTYPES
[16:37:24] * Coren_ nods.
[16:38:09] <Coren_> wjp: update your setup.H from CVS
[16:38:10] <wjp> yot: where's that?
[16:38:17] <yot> the devlog
[16:38:22] <yot> 2002-11-04
[16:38:31] <wjp> of uwadv?
[16:38:34] <yot> http://uwadv.sourceforge.net/index.php?page=devlog
[16:38:41] <Coren_> yot: He's referring to decompiled conversation code.
[16:39:00] <wjp> aaahh...
[16:39:03] * wjp gets it now
[16:39:11] <wjp> yes, that's uw's conversation script
[16:39:15] <yot> so it's a script debugger?
[16:39:52] <wjp> Coren_: ok, compiles now
[16:42:28] <yot> wjp: you should update the screenshots more often :)
[16:43:20] <wjp> vividos handles the webpage, actually :-)
[16:43:53] <yot> but you also have access to :)
[16:44:13] <yot> and if I say you I mainly mean the team ;p
[16:51:00] <Coren_> Incidentally, wjp, that version you got should have mouselook enabled by default. :-)
[17:23:01] <-- yot has left IRC ()
[17:43:43] * wjp hmms... segfault on startup again
[17:45:33] <wjp> ack
[17:45:52] <wjp> it seems I put my uw2 data files in a different directory than I remembered :-)
[18:11:14] * Coren_ grins.
[18:11:48] <Coren_> I should update the README to mention that you want to set the path to where you really have the UW2 install and not just where it might be. :-)
[18:13:26] * Coren_ wonders.
[18:13:47] <Coren_> Waitaminit... it shouldn't segfault, it should exit with a verbose message about not finding the uw2 files.
[18:15:48] <Coren_> *** Unable to access the UW2 data files. Check the README file.
[18:15:48] <Coren_> last error: No such file or directory
[18:16:01] <Coren_> Like this.
[18:18:10] <wjp> I don't get that
[18:18:26] <wjp> video: rendering with RIVA TNT2/AGP/3DNOW!
[18:18:26] <wjp> Fatal signal: Segmentation Fault (SDL Parachute Deployed
[18:18:26] <wjp> [wjp@aldur /data/cvs/low]$
[18:18:45] * Coren_ thinks.
[18:19:11] <Coren_> Might be caused by finding /some/ files and not others. I think that if I find the first few I just go ahead and presume they are all there.
[18:19:29] <Coren_> set setup.path.uw2 "/home/marc/uw2"
[18:19:50] <Coren_> Do you get the 'setup: Loading settings from low.ini' message at all?
[18:19:57] <wjp> no
[18:20:03] <Coren_> boggles.
[18:20:08] <Coren_> Err, 'cvs update'
[18:20:12] <Coren_> Might help. :-)
[18:20:27] <wjp> do you have a cvs mailing list?
[18:21:18] <wjp> oh, btw, I don't think you should have configure in CVS
[18:21:26] <Coren_> Err, no. But I use CVS the 'canonical' way; commiting every change that works and builds before starting on something else-- often means several commits per day so that'd be one heck of a high volume list. :-)
[18:21:45] <wjp> that's what mail filters are for :-)
[18:22:29] <Coren_> I feel configure should live in CVS if only because not everyone who wants to keep up with the CVS tree has autoconf handy.
[18:22:52] <wjp> but it gives a conflict every time I do a cvs update
[18:23:11] <Coren_> Aha. Point. Don't autoconf. :-)
[18:23:15] <wjp> if you want to use CVS you should have basic devel tools like autoconf
[18:23:24] <Coren_> Yeah, I guess that's a good point.
[18:23:52] <Coren_> Really think I should remove it?
[18:24:01] <wjp> IMHO, yes
[18:24:18] <Coren_> I'll take your word for it since you are my official autoconf guru. :-)
[18:24:28] <wjp> *cough* :-)
[18:24:46] * Coren_ adds configure to .cvsignore
[18:25:20] <Coren_> Actually, I'll add all the autoconf artifacts.
[18:26:28] <Coren_> You'll note I have not used automake at all; I *really* didn't like the Makefiles it creates. Too FSF. :-)
[18:26:39] <wjp> :-)
[18:26:55] <wjp> it kind of limits what you can do with them too
[18:27:18] <wjp> (them = Makefiles)
[18:27:19] * Coren_ nods. The 'default' targets are nice, but I can do this too. :-)
[18:27:34] * wjp hmms
[18:27:39] <Coren_> There. CVS is now autoconf-artifact friendly.
[18:28:19] <Coren_> Talking about that, I still need to add a 'distclean' target to the Makefile.in
[18:28:36] * wjp moves some data files around
[18:28:59] <Coren_> Yeah, well, I backtracked on my previous requirement of merged data/ directories.
[18:29:02] <Coren_> :(
[18:29:06] <wjp> I noticed :-)
[18:29:31] <Coren_> I think it's more well-behaved this way.
[18:29:40] <wjp> yeah, I agree
[18:30:09] * Coren_ notes that --with-uw2=/path during configure fills in low.ini
[18:32:07] <wjp> it currently only reads low.ini from the current dir?
[18:32:26] * Coren_ nods.
[18:32:36] <Coren_> I haven't decided on the right set of places to look for it yet.
[18:33:07] <Coren_> Probably eventually something like ~/.low ./low.ini ${varpath}/low.ini but I don't know what to do about win32
[18:33:47] <Coren_> I doubt I want to mess with the registry; but that might be the only viable option.
[18:33:52] <wjp> it'd probably be nice if it would read all of those in a specific order, and override settings when reading a new one
[18:34:05] * Coren_ nods.
[18:34:29] <Coren_> Right now, it's just a fairly ugly hack that is marginally less ugly than the very ugly hack of hardcoding everything. :-)
[18:34:37] <wjp> :-)
[18:35:20] <Coren_> But I used the opportunity to make all file opens go through my own interface so I'll be able to handle pak-style archives eventually.
[18:35:38] * wjp nods
[18:37:17] <Coren_> I think I did a nice job with the mouselook-- it was a bit clumsy at first but using CTRL to 'snap out' of it when needed plus the automatic temporary disable when manipulating gumps or objects just 'works' without being too confusing.
[18:37:53] <Coren_> Quake-style use-stuff-by-bumping-into-it justs didn't cut it for an RPG. :-)
[18:39:38] <Coren_> I'll probably add ALT-to-strafe before 0.3
[18:41:12] <Coren_> Does it still crash rudely with the CVS updated version?
[18:42:20] <Coren_> We received your address from a public area.
[18:42:20] <Coren_> We apologize if this letter has reached you by mistake.
[18:42:20] <Coren_> We'll not disturb you any more.
[18:42:26] <Coren_> Yeah, right.
[18:43:44] <Coren_> And the funny thing is, that particular piece of spam advertises spamming hosting quoting "tired of providers kicking you off for your marketing efforts?"
[18:47:30] <wjp> :-)
[18:47:52] <wjp> now that I properly set all paths it works ok, but let me screw something up again ;-)
[18:48:07] * Coren_ chuckles.
[18:49:08] <wjp> if I remove the data/models directory:
[18:49:11] <wjp> video: rendering with RIVA TNT2/AGP/3DNOW!
[18:49:11] <wjp> granny: attempting to initialize from granny.dat
[18:49:11] <wjp> Aborted
[18:49:22] <Coren_> Erm.
[18:49:34] <Coren_> Which means the exception doesn't get handled.
[18:49:55] <Coren_> That's odd because it is in mine. Did your CXXFLAGS eventually got to include -fexceptions?
[18:50:21] <wjp> yes
[18:50:28] <Coren_> Odd.
[18:50:38] <Coren_> What about fuxx0ring the path to UW2?
[18:50:44] <wjp> but let me do a clean make
[18:51:43] <wjp> setting the wrong uw2 path gives a segfault
[18:52:03] * Coren_ frowns.
[18:52:06] <Coren_> From a clean make?
[18:52:13] <wjp> yup
[18:52:28] <Coren_> Lemme try to check a fresh version out and build it.
[18:55:20] <Coren_> setup: Loading settings from low.ini
[18:55:20] <Coren_> video: attempting to set gamma to 1.3
[18:55:21] <Coren_> video: initialised 800x600 windowed video mode
[18:55:21] <Coren_> video: NVIDIA Corporation OpenGL version 1.3.1 NVIDIA 31.23
[18:55:22] <Coren_> video: rendering with GeForce2 MX/AGP/SSE2
[18:55:22] <Coren_> *** Unable to access the UW2 data files. Check the README file.
[18:55:24] <Coren_> last error: No such file or directory
[18:55:36] <Coren_> That's from a fresh co
[18:56:16] <Coren_> WTF could be going wrong?
[18:58:23] <Coren_> [marc@vampire low]$ gcc --version
[18:58:24] <Coren_> gcc (GCC) 3.2
[19:00:56] <Coren_> granny: attempting to initialize from granny.dat
[19:00:56] <Coren_> *** 3D models not found. Are you sure you downloaded the 3D and Media archives?
[19:00:56] <Coren_> last error: No such file or directory
[19:01:08] <Coren_> And that's with the models gone.
[19:19:35] <wjp> it works here too when I remove all the -f... options
[20:05:29] --> yot has joined #uwadv
[20:11:09] <Coren_> That is soooo odd. Theorically, it should *not* possibly work without -fexceptions-- I use exceptions to signal errors and -fexceptions turns handling on
[20:12:26] <wjp> exceptions are on by default in C++
[20:16:23] <Coren_> `-fexceptions'
[20:16:23] <Coren_> Enable exception handling. Generates extra code needed to
[20:16:24] <Coren_> propagate exceptions.
[20:16:36] <wjp> yeah, I read that
[20:16:40] <Coren_> From 3.2's gcc.info
[20:16:55] <wjp> If you do not specify this option, GCC will
[20:16:55] <wjp> enable it by default for languages like C++ which normally require
[20:16:55] <wjp> exception handling
[20:16:58] <wjp> a few lines down
[20:17:04] <Coren_> However, you may need to enable this
[20:17:05] <Coren_> option when compiling C code that needs to interoperate properly
[20:17:05] <Coren_> with exception handlers written in C++.
[20:17:08] * Coren_ smiles.
[20:17:54] <Coren_> And exception handling just fails when I don't have -fexceptions on my box, the exact same way yours does when it is there.
[20:18:24] <Coren_> You have 3.2?
[20:18:38] <wjp> yeah
[20:18:39] * Coren_ frowns.
[20:18:54] <wjp> not the official gcc3.2, though, but RH8's
[20:18:57] <Coren_> Sounds like I'll have to cook a test up to reproduce the bug in configure.ac
[20:19:17] <Coren_> Erp. RH fuxx0red THAT too?
[20:19:28] <wjp> quite possibly :/
[20:19:51] <wjp> -fexceptions isn't the problem
[20:19:57] * wjp tries adding back another -f
[20:20:20] <Coren_> From my reading of the .info; -fexceptions could not possibly harm even if it's not needed. It smells like they reverted to 2.96 behavior where -fexceptions toggled exception support.
[20:21:35] <wjp> still works after adding -frtti
[20:21:54] * wjp is betting on omit-frame-pointer
[20:22:17] <wjp> yeah, that one does it
[20:23:08] <Coren_> WTF does omit-frame-pointer have do to with exception handling? Stack unwinding doesn't care about frame pointers.
[20:24:01] * Coren_ grumbles. Omit-frame-pointer gives me almost 5% performance boost.
[20:24:12] <Coren_> Oh, well. I'm nixxing it from the configure.ac
[20:25:14] <wjp> google gives a couple of hits on this
[20:25:21] <Coren_> So -fexceptions without -fomit-frame-pointer works for you?
[20:25:50] <wjp> in particular a developer.apple.com page says "The C++ Exception Handling stack unwind code can get confused if it needs to unwind past a stack frame that has been adjusted via the gcc builtin function alloca()" ... "Workarounds: either compile with "-fno-omit-frame-pointer""
[20:26:04] <wjp> (not sure which gcc version this refers too)
[20:26:05] <wjp> s/too/to/
[20:26:38] <Coren_> That sounds like a fundamental problem; chances are this is most (all) versions.
[20:27:27] <Coren_> I don't use alloca, but I make frequent use of variably-sized auto arrays which are implemented mostly the same way.
[20:28:12] * wjp wonders if he should add an --enable-athlonxp
[20:28:45] <Coren_> If there is considerable gain in your opinion from athlon-specific switches, go right ahead.
[20:29:44] <Coren_> The --enable-pentium4 gives 25% performance boost.
[20:30:03] <Coren_> Better scheduling and sse2 insn.
[20:34:57] <wjp> adding a couple of athlon-specific -m options seems to add some fps in the throne room
[20:35:14] <wjp> went from ~10 to ~12
[20:35:38] <Coren_> That's pretty good; the throne room is VERY cpu intensive with 3d cards that don't do geometry.
[20:35:56] <wjp> I guess a tnt2ultra doesn't do that yet? :-)
[20:36:29] * Coren_ chuckles.
[20:36:48] <Coren_> AFAIK only the GeForces Ti from nVidia are supported ones that do.
[20:37:11] * Coren_ wishes he could afford one.
[20:37:17] <Coren_> The woes of being a poor student.
[20:38:15] <Coren_> Though I'm fairly pleased with my box that can still hold ~16fps in the throne room.
[20:38:38] <Coren_> How does my first attempt at mouselook feel to you BTW?
[20:38:45] <wjp> quake-ish :-)
[20:38:53] <Coren_> Is that good or bad? :-)
[20:39:12] <wjp> I'm not sure
[20:39:50] * Coren_ shrugs. It's settable on or off-- I have had a number of people request this, I expect because they are vastly used to that way of moving around.
[20:40:05] <wjp> I'm vastly not used to it ;-)
[20:42:23] <Coren_> I am, but I'm still wondering if I'd use it myself or not-- I guess it kinda feels 'wrong' for UW2, but I can see why people would like it.
[20:47:25] --> vividos has joined #uwadv
[20:47:25] --- ChanServ gives channel operator status to vividos
[20:47:32] <Coren_> Hey Michael!
[20:47:48] <vividos> hi Marc!
[20:48:00] <Coren_> Still on schedule for that release? :-)
[20:48:27] <vividos> most stuff is done, yes
[20:48:40] <Coren_> I'll make sure I wait at least a week after yours to do my next since we appear to be sharing traffic.
[20:49:47] <Coren_> Besides, mouselook and a better build process are neat but not much meat for a 0.3. I'll prolly wait for preliminary conversations.
[20:50:31] <vividos> good. I'm anxious to see where uwadv gets the most hits from
[20:50:49] <Coren_> I can tell you ahead of time. Freshmeat, with linuxgames a close second.
[20:51:24] <Coren_> Though I expect I'll see a fair number a hits from your site during that rush the same way you had quite a few visit yours on my 0.2
[20:51:28] <vividos> I more think of freshmeat with retrogames.com next, then perhaps happypenguin. or so
[20:52:04] <Coren_> Really? Personally, I was surprised by linuxgames; I got about twice as many hits from them as happypenguin when they picked up my freshmeat announcement.
[20:52:05] <vividos> yes, that could be. do you have referer logging on the low sf page?
[20:52:22] <vividos> ok, don't be sure about linuxgames vs. happypengiun
[20:52:25] <vividos> penguin :)
[20:52:46] * Coren_ nods. Mostly ripped out of yours; but I collect cumulatives in a separate file that doesn't grow so fast.
[21:07:50] <vividos> hmm, what does AC_GNU_SOURCE?
[21:08:38] <Coren_> It defines _GNU_SOURCE; I don't actually think it's needed anymore.
[21:09:05] <Coren_> It helps with some #includes that define gnu extensions I no longer use.
[21:09:52] <Coren_> Actually, I'm going to remove it right now.
[21:10:13] <vividos> ah ok
[21:10:41] <Coren_> There. Commited.
[21:11:40] <vividos> wjp always manages to let people remove their ./configure from CVS. *bg*
[21:12:31] * Coren_ chuckles.
[21:14:22] <vividos> ah last SF mounted web space read-write again
[21:16:45] <Coren_> wjp: feel free to commit your --enable-athlonxp changes.
[21:16:52] <wjp> oh, right :-)
[21:16:55] * wjp got distracted :-)
[21:18:02] <wjp> what's your changelog policy?
[21:18:48] <Coren_> I add to it just before a release tag by examining the commit logs. Less error-prone this way.
[21:19:52] * vividos thinks about if uwadv should also have a hand-written "Changes" file or something ...
[21:19:58] <Coren_> So, basically, make your commit message descriptive and everything else will go smoothly.
[21:21:08] <QQtis_> vividos
[21:21:15] <QQtis_> I'm cruising down on level 8
[21:21:29] <QQtis_> stalactites should be attached to the ceiling :)
[21:21:56] <vividos> yes I know, and I thought they were
[21:22:58] <vividos> Coren: well, if you have hundreds of commits, a casual user wouldn't see the wood from the trees (don't know if that phrase exists outside of .de :)
[21:23:43] <Coren_> vividos: A casual user is expected to politely wait for a release with a nice, updated changelog. caveat cvsor. :-)
[21:24:04] <vividos> well, I do my changelog with cvs2cl.pl
[21:24:53] <Coren_> That only works for big commits; I am a compulsive commiter-- whenever any change is finished and builds/works I commit.
[21:25:02] <vividos> me too :)
[21:25:26] <Coren_> Make going 'oops' less painful when you break something. :-)
[21:25:37] <vividos> and I think it works for "source-level followers" of projects
[21:26:31] <Coren_> I'll go take a peek at cvs2cl.pl at any rate.
[21:26:46] <vividos> but the "ChangeLog" link on freshmeat is better filled with some other document :)
[21:30:01] <vividos> http://img.web.de/c/00/18/90/F5.420 <-- "before steering was invented"
[21:30:15] <wjp> :-)
[21:30:18] <Coren_> Heh. Somebody flamed me for using U9-style gazers. :-)
[21:30:25] <vividos> I read it :)
[21:32:04] <QQtis_> :)))
[21:34:27] * vividos thinks about ripping apart all FMOD stuff ...
[21:35:02] <QQtis_> what is that?
[21:35:20] <vividos> FMOD is a music playback library, almost like SDL_mixer
[21:35:57] <QQtis_> hm...
[21:36:13] <QQtis_> because eventually we might want to have SFX in 44khz, no?
[21:36:14] <QQtis_> :)
[21:36:31] <vividos> it's rather cumbersome to set up and isn't even used in current source, IIRC
[21:36:41] <QQtis_> hm..
[21:36:53] * QQtis_ out for a smoke
[21:37:28] <vividos> QQtis: no, the FMOD lib just plays back midi
[21:39:35] <wjp> hmm... are we even allowed to include fmod?
[21:40:42] <wjp> (wrt FMOD's license and the GPL, I mean)
[21:41:02] <vividos> it's free for non-commercial use
[21:41:24] <vividos> and besides it, every user can do what he likes at home :)
[21:43:46] <Coren_> I replied "Okay, you're right. What an heretic I've been. I'll use the U7 gazers instead-- green spiders are soo much cuter."
[21:44:08] <vividos> heh :)
[21:45:01] <vividos> hmm, should we check for SDL_mixer.h?
[21:47:59] <vividos> bbl
[21:50:20] * Coren_ does; I disable sound when SDL_mixer is not there.
[21:55:05] <wjp> Coren_: hm, that last test in configure.ac should probably be checking HAVE_LIBSMPEG, right? :-)
[21:55:45] <Coren_> Err, oops. Maybe? :)
[22:07:17] * Coren_ avoids GPL-related problems by avoiding the GPL altogether.
[22:08:15] <Coren_> I'm aprticularly proud of my LoW licence, which got accepted outright as Open Source by sf despite my having taken all of 2 minutes to create and write it. :-)
[22:25:17] <Coren_> I mees ot ytpe klie na sas otnigth.
[22:25:37] <wjp> heh, it's still readable ;-)
[22:25:56] * wjp hmms
[22:26:02] <wjp> I think I'll go to bed
[22:26:06] <wjp> night
[22:26:11] <Coren_> This early?
[22:26:16] <wjp> 23:25
[22:26:19] <Coren_> Oh, right, different continent. :-)
[22:26:22] * Coren_ waves.
[22:26:26] <wjp> :-)
[22:26:31] <-- wjp has left IRC ("Zzzz...")
[23:14:49] <vividos> b
[23:15:53] * vividos checks out the LoW license
[23:20:11] <vividos> DO NOT TAUNT HAPPY FUN BALL! :))
[23:21:08] <vividos> hmm, does your license mean that everyone that wants to contribute code has to be canadian citizen?
[23:52:44] <Coren_> No; it means that I, the copyright holder, am. :-)
[23:54:16] <Coren_> Did you get the happy fun ball reference? I would expect almost no non-americans would. :)