#uwadv@irc.freenode.net logs for 8 Nov 2002 (GMT)

Archive Today Yesterday Tomorrow
Underworld Adventures homepage


[01:21:37] <-- Fingolfin has left IRC ("dream well")
[03:21:12] --> Coren__ has joined #uwadv
[03:21:12] <-- Coren_ has left IRC (Read error: 104 (Connection reset by peer))
[03:35:14] <-- Coren__ has left IRC (vinge.freenode.net irc.freenode.net)
[03:36:16] --> Coren__ has joined #uwadv
[03:48:54] --> Coren_ has joined #uwadv
[03:49:37] <-- Coren__ has left IRC (Read error: 113 (No route to host))
[05:14:21] --> al-khaaliq has joined #uwadv
[05:21:22] <-- al-khaaliq has left #uwadv ()
[09:22:59] --> QQtis_ has joined #uwadv
[09:23:00] <-- QQtis has left IRC (Read error: 54 (Connection reset by peer))
[12:29:10] --> Telemachos has joined #uwadv
[12:29:10] --- ChanServ gives channel operator status to Telemachos
[12:29:19] <Telemachos> hello there.
[13:09:45] --> John_Who has joined #uwadv
[13:09:51] <John_Who> hello
[13:14:28] <-- John_Who has left IRC ("Client Exiting")
[13:28:29] <-- Telemachos has left IRC (Read error: 110 (Connection timed out))
[15:44:44] --- QQtis_ is now known as QQtis
[15:44:44] --- ChanServ gives channel operator status to QQtis
[16:24:35] --> wjp has joined #uwadv
[16:24:35] --- ChanServ gives channel operator status to wjp
[16:26:01] * Coren_ waves to wjp.
[16:26:11] <wjp> hi
[16:26:24] <Coren_> How goes?
[16:26:50] <wjp> pretty good, thx :-)
[16:27:42] <Coren_> Joyjoy: LoW is now sourceforged. :-)
[16:29:49] * wjp guessed the URL correctly :-)
[16:30:11] <Coren_> Alternately, going to the old url would have gotten you there as well. :-)
[16:30:19] <wjp> that would require me to know the old one ;-)
[16:30:31] <wjp> (although I probably have that lying around in my bookmarks somewhere)
[16:30:52] <Coren_> You do, actually, although you might not remember it offhand; Michael gave it to you some time ago on this very channel. :-)
[16:31:11] <wjp> yeah, I've visited your old page before
[16:31:29] <Coren_> The new one looks nicer too. Black is sooo passe'. :-)
[16:31:34] <wjp> your left 'frame' is kind of big on your page
[16:32:12] <Coren_> Hrm; it shouldn't be much over 130 pixels or so-- Its driven by the sourceforge logo size which is, IIRC, 125 pixels wide or so.
[16:32:24] <wjp> it's 2-3 times as wide as the SF logo
[16:32:29] <Coren_> Oh?
[16:32:34] <wjp> but I guess that's because the right frame doesn't need any more space
[16:33:00] <Coren_> Doesn't do that with Konkeror; but I admit I haven't looked at the site with any other browser.
[16:34:21] <Coren_> I'm glad I moved to the sf servers /before/ I freshmeated. I already have some 500 hits coming from freshmeat in the past 8 hours.
[16:34:50] <wjp> :-)
[16:38:58] <Coren_> Now all that's left it wait for the sf people to import my cvs tree.
[16:39:14] <Coren_> I don't want to loose sync with it before it even gets there. :-)
[16:40:13] <Coren_> Sa, my understanding is that you're the Linux guru for uwadv?
[16:42:13] <wjp> I wouldn't say guru, but I do maintain the linux port, yes :-)
[16:42:17] <Coren_> You familiar with autoconf at all? I've never had a use for it yet (I VERY rarely write non Linux code) and so I'm completely unfamiliar with it myself.
[16:42:40] <wjp> yeah, I'm somewhat familiar with it
[16:42:52] <Coren_> I want (for obvious reasons) to move LoW to autoconf. Think you can spare the cycles to give me a hand?
[16:43:32] <Coren_> Right now; I have a very working but amazingly ugly makefile that builds for linux and cross-compiles for win32.
[16:44:07] * wjp nods
[16:46:11] <Coren_> I'm fairly certain I'll be able to keep it working once it's been setup in the first place; but I don't feel confortable trying to do it 'right' in the first place.
[16:47:52] <Coren_> What's your sf username? I'll add you to the project so you can do this whenever you have the time.
[16:48:08] <Coren_> Not like there's any rush.
[16:48:41] <wjp> wjpalenstijn
[16:48:53] <wjp> brb
[16:52:57] <wjp> b
[16:53:54] <Coren_> You're all set. I expect the CVS tree will be in sometime during the day, the support request has been assigned already.
[16:54:17] <Coren_> Like I said; there's no rush. Do this when you have the time and feel like it.
[16:54:25] <Coren_> Thanks a bundle. :-)
[16:55:26] <wjp> I'll be away all of next week, so it might take a while :-)
[16:55:46] <wjp> maybe tomorrow if the CVS tree is in place already
[16:57:03] <Coren_> Heh. *I* can keep building with my hairy makefile, that's the most important bit. It's just that I fear people emailing me with "what's that -bi686-linux-cygwin flag?" and such.
[16:57:16] <wjp> good quesion :-)
[16:57:20] <wjp> question, even
[16:57:52] <Coren_> *chuckles* Tells GCC to use my cygwin setup from within Linux for cross-compiling. :-)
[16:58:30] <Coren_> Kinda fun. I get to build the win32 executable with gcc under linux and test with with Wine. :-)
[17:00:32] <wjp> hehe :-)
[17:01:30] <Coren_> Ah; from what I can see from the sf tracker; there is only one ticket to handle by burley before mine. Looks like the tree will be in place very soon and I'll be able to check it out and resume the important things: working on the game itself. :-)
[17:15:30] <Coren_> This is strange. I have a significant number of visitors coming from www.hup.hu; a hungarian slashdot-like site. It looks like I have a writeup there but except for the name of my project I have /no/ idea what they are saying. :-)
[17:18:13] <Coren_> Although I can guess at "project honlapja", "screenshotok" and "1st person 3D grafikus engine-nel". :-)
[17:23:07] <Coren_> Hmmm. Babelfish doesn't speak hungarian either.
[17:23:51] <wjp> hm, there are several hungarians visiting at my uni :-)
[17:25:08] <Coren_> I'm not entirely certain why the article, which seems to be talking about LoW, sports a Q3 logo either. Might be a generic "3d game" category logo or something.
[17:25:20] <wjp> yeah, probably
[17:25:34] <wjp> hm, no, games
[17:25:48] <wjp> if you click on the icon the top story is on freeciv
[17:26:40] * Coren_ nods. I had just tought of that as well.
[17:28:10] <Coren_> Aha. Found what seems to be a site that translates hungarian.
[17:29:05] <Coren_> Yep. Machine translated, but parsable. :-)
[17:29:16] <Coren_> http://www.tranexp.com:2000/InterTran?type=url&url=http%3A%2F%2Fwww.hup.hu%2Fmodules.php%3Fname%3DNews%26file%3Darticle%26sid%3D2021&text=&from=hun&to=eng&Submit=TRANSLATE
[17:30:22] <wjp> ugh
[17:30:31] <wjp> that's pretty badly translated :-)
[17:30:37] <Coren_> Err, don't bother following the link. It just /appeared/ parsable. about 50% of the words didn't translate at all.
[17:32:33] <Coren_> Hmmm. Hungarian is a romans language, I might have better luck translating to french.
[17:33:50] --> DredgeRupert has joined #uwadv
[17:33:51] <DredgeRupert> hi
[17:33:56] <Coren_> Nope. Actually, the french output manages to be worse.
[17:33:56] <DredgeRupert> is it playable yet?
[17:34:20] <DredgeRupert> is the linux version playable?
[17:34:35] <DredgeRupert> ?
[17:34:43] <wjp> hi
[17:34:45] <wjp> bbl, dinner
[17:35:00] <wjp> I think that depends on your definition of 'playable' ;-)
[17:35:01] <DredgeRupert> yes or no?
[17:35:27] <wjp> you can walk around, but there's no conversation or combat yet
[17:38:51] <DredgeRupert> that's 1 and 2?
[17:40:58] <Coren_> <shameless plug>Actually, uwadv just does UW1 right now, but my project (http://low.sourceforge.net) does UW1</shameless plug>
[17:41:10] <Coren_> Err, does UW2
[17:41:13] * Coren_ smiles.
[17:48:10] <DredgeRupert> but neither can let you play
[17:49:18] <Coren_> Well, no. That will come after further development. Recreating a game of those magnitudes is a very complex task. Harder, in fact, that writing a new one from scratch. But keep an eye on both projects, they will eventually allow you to play both.
[17:49:58] <Coren_> Take a look at exult, which did just that for Ultima VII. It has gotten to version 1.0 (fully playable with no showstopping bugs) after four years of development.
[17:50:21] <Coren_> They did a very, very good job of it too.
[17:50:39] <DredgeRupert> gl?
[17:51:32] <Coren_> Not 1.0; but I seem to recall reading that the 1.1 development branch is working on GL assisted rendering-- note that U7 is isometric, not first person, so it's not as much of an issue.
[17:51:33] <-- DredgeRupert has left IRC ("I like core dumps")
[17:51:47] <Coren_> That was rude.
[18:01:14] <wjp> yup
[18:10:48] <Coren_> Typical luser. "Gimme gimme gimme gimme!"
[18:11:06] * wjp nods
[18:11:24] <wjp> impatient and not wanting to help one bit
[18:11:41] <Coren_> Erp. 900 hits in <10h. I wanted more people to test, I got my wish. :-)
[18:13:00] * Coren_ chuckles. "uwadv got 247 hits referred from my page in that same period."
[19:18:13] <Coren_> Aha! Voila. The CVS tree is there.
[20:35:15] --> Fingolfin has joined #uwadv
[20:35:49] <Fingolfin> hi
[20:40:27] <Coren_> Hello, Fingolfin.
[20:41:59] --> yot has joined #uwadv
[20:43:31] <Coren_> HI yot.
[20:43:48] <wjp> Coren_: grender.cc:280: `glPointParameterfvEXT' undeclared (first use this function)
[20:44:02] <wjp> it looks like it's declared in an #ifdef WIN32
[20:44:13] <yot> hi
[20:45:57] <Coren_> wjp: Lemme check.
[20:52:29] <Coren_> wjp: I don't get it: Look at line 278: static PFNGLPOINTPARAMETERFVEXTPROC glPointParameterfvEXT = 0;
[20:52:34] <Coren_> It's declared right there!
[20:53:29] <Coren_> And it builds on my box.
[20:53:50] <Coren_> Hey, wait a minute. I tought you were building under Linux? Why is WIN32 defined for you then?
[20:53:55] <wjp> it isn't
[20:54:09] <wjp> that function is used right under the #endif too
[20:54:20] <Coren_> Ah! Got it.
[20:54:21] <wjp> line 280
[20:54:43] <Coren_> Hmmm. Actually, this should be defined because it's guarded by #if defined(GL_DISTANCE_ATTENUATION_EXT)
[20:55:19] <Coren_> Try this: add #include <GL/glext.h> right after #include <GL/glut.h>
[20:55:49] <Coren_> glext.h isn't supposed to be needed for ARB and EXT extensions, but your particular setup might insist on it.
[20:56:20] <wjp> same
[20:56:47] <Coren_> Wait, are you using an old Mesa gl.h? There are known holes in it.
[20:57:16] <wjp> [wjp@aldur /usr/include/GL]$ rpm -qf gl.h
[20:57:16] <wjp> XFree86-devel-4.2.0-72
[20:57:48] <Coren_> grep it, see if you can find glPointParameterfv in it
[20:58:06] <wjp> yeah, it is
[20:58:16] <wjp> well, glPointParameterfvEXT is, anyway
[20:58:20] <Coren_> Ok, then I'm confused.
[20:58:31] <Coren_> Yes, that's expected since the file defines the _EXT version.
[20:58:45] <wjp> it's in several #ifdef's, though
[20:58:51] <Coren_> Hey, wait a minute. This file should not be including glut.h anymore, but gl.h.
[20:59:02] <Coren_> That might be the problem right there
[20:59:33] <wjp> still the same
[20:59:43] <Coren_> Can you DCC me your gl.h?
[21:00:37] <wjp> the declaration is inside a #ifdef GL_GLEXT_PROTOTYPES
[21:00:49] <wjp> and that isn't defined anywhere, AFAICT
[21:01:01] <Coren_> That's the first time I see that define.
[21:01:05] <Coren_> Oddness.
[21:01:26] <Coren_> Define it before the #includes. I see no reason why that should harm other setups.
[21:01:40] <wjp> oddly that still doesn't work
[21:01:52] <wjp> oh, wait, it needs to be before gl.h too
[21:01:58] <Coren_> Yes. :-)
[21:02:22] <wjp> I guess gl.h includes glext.h?
[21:02:37] <Coren_> As of GL 1.1 it's supposed to by default.
[21:02:55] <Coren_> glext.h is old crud dating from 1.0
[21:02:55] <yot> btw
[21:02:56] <wjp> ok, next up: bison trouble :-)
[21:02:57] <yot> wjp
[21:03:17] <yot> I heard changed the windows sound output from waveout to dsound
[21:03:19] <yot> is that right?
[21:03:32] <wjp> I think so, yes
[21:03:43] <yot> just if you want to know
[21:03:43] <wjp> the bug in SDL that forced us to go to waveout should be fixed now
[21:03:50] <yot> it isn't !
[21:03:51] <yot> :(
[21:04:05] <yot> the sound still bugs for me
[21:04:12] <yot> if I set the output to waveout
[21:04:13] <wjp> do you have SDL 1.2.5?
[21:04:17] <yot> everything's fine
[21:04:21] <yot> SDL 1.2.5a
[21:04:52] <Coren_> bison trouble? If you feel that's a potential Hard Problem there is no reason to not include the output files themselves-- it is unlikely anyone will want to mess with the grammar files since they are used for parsing .ase models.
[21:05:17] <wjp> does your bison create a ase_parse.cc.h?
[21:05:36] <Coren_> Yes.
[21:05:41] <yot> wjp: all I want to say is you should add a note in the readme if someone has trouble with the sound there should be an option for waveout
[21:05:42] <wjp> mine doesn't :-)
[21:05:43] <yot> just an option
[21:06:00] <Coren_> What does it create because of the -d?
[21:06:12] <wjp> possibly ase_parse.hh
[21:06:24] <wjp> timestamp on that looks like it was newly created
[21:06:42] <Coren_> That's not part of the CVS so that must be it. What an odd name for it.
[21:06:45] <wjp> yot: yeah, probably. I have no idea what would need to be done for that, however
[21:06:52] <wjp> yup, weird name
[21:07:11] <Coren_> yot: Why don't you go file a bug on the sourceforge project page? This way you're sure it'll get the right attention.
[21:07:43] <Coren_> wjp: that .hh file contains a simple typedef and a lot of #defines?
[21:07:59] <wjp> yeah
[21:08:30] <Coren_> wjp: That's definitely the one then. It just needs to be renamed to ase_parse.h
[21:08:59] <Coren_> Might want to change the makefile entries to something like:
[21:09:05] <Coren_> ase_parse.cc ase_parse.h: ase_parse.y
[21:09:09] <Coren_> bison -d -o ase_parse.cc ase_parse.y
[21:09:20] <Coren_> @mv -f ase_parse.cc.h ase_parse.h
[21:09:37] <Coren_> #mv -f ase_parse.hh ase_parse.h
[21:09:43] <Coren_> Err, @mv...
[21:09:52] <wjp> make aborts on the first error
[21:09:54] <Coren_> So it'll quietly rename it if it gets either name.
[21:10:06] <Coren_> -f means 'try if you can, don't bother if you can't'
[21:10:09] <Coren_> so no error.
[21:11:01] <wjp> sure about that?
[21:11:15] <Coren_> Sheesh; I had no idea just building the thing on a different setup than mine would be that troublesome. I *really* needed to autoconf the whole thing.
[21:11:19] <Coren_> Yes.
[21:11:32] <wjp> seems to return false
[21:11:43] <wjp> if mv -f ase_parse.cc.h ase_parse.h -f /tmp/sehtiowe /tmp/bsieth; then echo ja; fi
[21:11:45] <wjp> does not output 'ja'
[21:11:50] <Coren_> Hmmm. Not supposed to.
[21:11:55] <wjp> (dutch for yes, incidentally :-) )
[21:12:07] <Coren_> Hrm.
[21:12:11] <wjp> hm, that copy paste went slightly wrong
[21:12:21] <Coren_> I had understood the line anyways. :-)
[21:12:31] <Coren_> Well, *my* mv returns true.
[21:12:48] <Coren_> Oh, wait, just add a '-' in front of both mv lines
[21:12:54] <Coren_> That tells make to ignore errors. :-)
[21:13:08] <Coren_> -@mv -f ase_parse.cc.h ase_parse.h
[21:13:35] <wjp> ugh :-)
[21:14:10] <Coren_> I would, of course, like to know why the heck your bison behaves differently than mine.
[21:14:23] <Coren_> [marc@vampire low]$ bison --version
[21:14:26] <Coren_> GNU Bison version 1.28
[21:14:32] <wjp> 1.35 here
[21:14:41] <wjp> it's possible mine recognizes the .cc extension
[21:14:52] <Coren_> Likely, even.
[21:15:09] <wjp> hm, I need smpeg?
[21:15:40] <Coren_> If you want MP3 music. That's the kind of thing the ./configure script would deal with idealy. :-)
[21:16:17] <wjp> hm, mp3? that would explain why RH8 didn't package it :/
[21:16:37] <Coren_> RedHat having a bone to pick with mp3s?
[21:16:46] <wjp> patent issues or something
[21:16:55] * Coren_ shakes his head.
[21:17:05] * wjp builds smpeg
[21:17:19] <Coren_> Oh, which GCC are you using BTW?
[21:17:26] <wjp> 3.2
[21:17:30] * Coren_ nods.
[21:17:47] <Coren_> Good. No surprises there then. I wonder how it will cope with <3.2
[21:18:41] <Coren_> That'd mean RH8 doesn't even have xmms?
[21:18:48] <wjp> it does, but no mp3 support
[21:18:49] <Fingolfin> are there demo files for UW available somewhere? I was asked to try and see if uwadv still works on OS X, but I am 400km from home and don't have my UW files here :-)
[21:19:23] <Coren_> Fingolfin: I know there is a demo, but I have NO idea where it can be found. Contains the first level IIRC.
[21:19:43] <wjp> it's downloadable as a DOS .exe somewhere :/
[21:19:51] <Fingolfin> gah
[21:20:00] <Fingolfin> well, I could of course first port DOSbox <g>
[21:20:20] <wjp> oh joy, smpeg doesn't want to build
[21:20:22] <Fingolfin> don't have a PC emulator installed here either (this is just a small little poor laptop :-)
[21:20:29] <Coren_> wjp: Huh?
[21:20:42] <wjp> undefined refs to new and delete while linking
[21:20:46] <Coren_> wjp: Lemme make a tarball of the lib and includes for you.
[21:20:47] <wjp> it seems to think it's C or something
[21:20:59] <wjp> no need, I'll just get a prebuilt rpm from somewhere
[21:21:14] <Coren_> wjp: Ah, a common 3.2 bug when C++ programs are compiled with 'gcc' instead of 'g++' You can fix the makefiles for that.
[21:21:24] <wjp> yeah, I suppose I could
[21:21:50] <Coren_> I had to do that with like, gazillions of packages. I don't use a distro, I roll my own entirely from source.
[21:22:11] <wjp> :-)
[21:22:17] <wjp> must be painful :-)
[21:22:48] <Coren_> Actually, not really because then I *know* absolutely every bit works with absolutely everything else and everything is installed exactly where and how I want it.
[21:22:55] <Coren_> Keeping it maintained is not that hard.
[21:23:30] <Coren_> Except when I decide to do a glibc upgrade. Then I have about 5 days of 100% cpu usage while I recompile the entire universe. :-)
[21:23:42] <wjp> :-)
[21:24:40] <wjp> do you use any system to manage packages?
[21:25:22] <Coren_> No; but I wish I did. By now, I'd need to start over. I'm just really careful. :-)
[21:27:12] <wjp> I've recently started using a kind of package management thingie called 'graft' at the uni for software we install on our NFS mount for 'extra' software
[21:27:23] <wjp> (where 'extra' == not installed locally on each machine)
[21:27:54] <wjp> works kind of ok for our (relatively small) amount of packages; not sure how well it would scale
[21:28:32] <Coren_> Well, given that I personally install every single piece of software on my box it's not that hard to maintain; I tend to know what is installed at any time.
[21:28:52] <Coren_> Any success with building LoW?
[21:29:08] <wjp> yeah, it just finished building
[21:29:14] <wjp> smpeg should work now too
[21:29:53] <wjp> it probably shouldn't segfault on startup, right? :-)
[21:31:07] <wjp> looks like I need some more files
[21:32:27] <wjp> do I need to merge the original data dir and the low one?
[21:33:49] <Coren_> Yes.
[21:34:13] <Coren_> You may want to do a cvs update, too, I just added code to be polite when files are missing rather than just go boom.
[21:36:25] * wjp downloads 3d files
[21:36:36] * Coren_ chuckles.
[21:36:54] <wjp> do I need the media files too?
[21:36:58] <Coren_> You can actually skip the media files if you want to avoid the 20-odd meg download; you'll just have no music and no sound.
[21:37:19] <wjp> I can live with that for now I guess :-)
[21:37:40] <wjp> although I guess 20Mb won't take all that long
[21:38:06] <Coren_> Depends. If you're stuck on a 56k it's painful. :-)
[21:38:15] <wjp> yes :-)
[21:38:31] <wjp> I'm getting 45Kb/s now
[21:38:31] * Coren_ checks that, in fact, LoW still copes well with missing sound and music.
[21:38:59] <Coren_> Ah, yes. It whines on the console but works.
[21:39:30] <Coren_> audio: precached 0 sounds
[21:39:30] <Coren_> audio: sound system initialized
[21:39:30] <Coren_> Starting game... go! go! go!
[21:40:21] <wjp> yes, and "Illegal instruction" *boom* after that :/
[21:40:29] <Coren_> What CPU do you have?
[21:40:36] <wjp> athlon
[21:40:50] <Coren_> You may want to remove the -sse from the CFLAGS and rebuild then.
[21:40:53] * Coren_ grins.
[21:40:58] <wjp> evil :-)
[21:41:13] <Coren_> Actually, rom CCCPU
[21:41:32] <Coren_> Probable culprits are -mmmx and -msse2
[21:41:47] <wjp> -mcpu=pentium4 doesn't help either I guess :-)
[21:41:50] <Coren_> Forgot those are heavily tuned for my P4. :-)
[21:42:15] <Coren_> Actually, -mcpu just changes some of the code layout for scheduling and behavior; not the instructions themselves.
[21:42:24] <wjp> oh, ok
[21:42:29] <Coren_> So it can be suboptimal, but would still work.
[21:42:37] <wjp> ok, I actually see something now :-)
[21:42:44] <wjp> 25fps
[21:43:13] <Coren_> That's the framelock, so it means you get max performance.
[21:43:44] <wjp> drops to 15-20 when walking around in the throneroom
[21:44:02] <Coren_> That's to be expected. Them Granny models are EXPENSIVE in cpu time.
[21:44:13] <wjp> granny? :-)
[21:44:17] <Coren_> Granny 3d
[21:44:27] <Coren_> That's the format EA uses for UO and U9
[21:45:08] <Coren_> It's actually very nice skeleton-based IK model code; and the file format is sane.
[21:45:30] <Coren_> Although I obviously had to whip my own since I can't affort the $15K liscence to buy theirs.
[21:46:14] <Coren_> Chances are, I'm doing the mesh deformation in a very suboptimal way, but those do well for the moment.
[21:46:47] <Coren_> You can fiddle around with F12 to change your light source. Take a peek at my 'lantern' illumination (two steps from the default).
[21:47:30] <Coren_> If you go to 'map 5' (in the console) you can go see my particle system and ambient sounds around the waterfalls.
[21:48:09] <wjp> hm, it segfaulted when I pressed enter on the console
[21:48:27] <wjp> (surprisingly I guessed the right key for accessing the console ;-) )
[21:48:30] <Coren_> Err, I might not actually check that you have typed a command.
[21:48:41] <Coren_> And just try to parse the empty line. :-)
[21:48:49] <wjp> how do I leave the console again?
[21:48:54] <Coren_> '~' again
[21:49:13] <Coren_> Might need to hit shift to get an actual '~' mind you.
[21:49:25] <wjp> hm, I could've sworn that was my first try
[21:49:28] <wjp> ah well, works now :-)
[21:49:40] <Coren_> Tell me what you think of the waterfalls. :-)
[21:49:59] <wjp> any way to increase gamma?
[21:50:14] * wjp is lost in the dark :-)
[21:50:44] <Coren_> 'set s.v.g <gamma>'
[21:50:58] <Coren_> Or you can hit F12 repeatedly until you get to 'daylight spell' brightness. :-)
[21:51:30] <Coren_> I think gamma changes are only honored on video restart, though, like when you toggle between fullscreen and windowed.
[21:52:21] <Coren_> 'set' will show you the console variables
[21:54:45] <Coren_> So, did you find your way to the waterfall on map 5 yet? :-)
[21:54:51] <wjp> yup, just found it :-)
[21:55:18] <wjp> nice effect
[21:55:34] <Coren_> The moongates (map 69) also use the particle system.
[21:55:36] <wjp> ugh, gamma goes down to 1 when I leave the window with the mouse :-)
[21:55:45] <Coren_> That's an XGL limitation.
[21:55:53] <wjp> (which is a good thing, but not for admiring the fx :-) )
[21:56:08] <wjp> ooh, moongates are nice
[21:56:37] <wjp> the actual gate could use a bit more texture, though, I think
[21:56:43] <wjp> but the sparkles are very nice :-)
[21:57:20] <wjp> I have to go for a while; bbl
[21:57:25] <Coren_> See ya.
[21:59:09] <Coren_> (I call the moongate sparkles my 'buzzing flies' effect) :-)
[22:03:42] <Fingolfin> trying to compile uwadv from HEAD cvs, linking fails, with missing symbols:
[22:03:44] <Fingolfin> ua_conv_graph::init(std::vector<unsigned short, std::allocator<unsigned short> >, unsigned short, unsigned short, std::vector<std::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::allocator<std::basic_string<char, std::char_traits<char>, std::allocator<char> > > >&)
[22:03:44] <Fingolfin> ua_conv_graph::process()
[22:04:22] <Fingolfin> any hint?
[22:11:18] <Fingolfin> maybe I have to rerun autogen.sh, considerig that nothing in source/conv was built.. weird
[22:21:28] <Coren_> Hmmm. That's SDL woes; what GCC version are you using?
[22:25:21] <Fingolfin> Apple GCC 3.1
[22:25:36] <Fingolfin> but the compile didn#t even compile the files
[22:25:47] <Fingolfin> so I think it's more likely to be a build system issue
[22:25:59] <Fingolfin> I just rerun autogen.sh and configure, we'll see if that helps
[22:26:18] <Coren_> That might help. If it's a cvs checkout, you might want to try to 'make distclean' first.
[22:29:55] <Fingolfin> another question.. why is it forcing me to link against X11? I don't want to use X11 after all... but it automatically uses the X11 GL etc., even though my SDL is *not* based on X11, it's the cocoa backend
[22:31:51] <Fingolfin> also, GL/gl.h won't work on OS X, if you want to use native OpenGL+SDL. Rather either OpenGL/gl.h has to be used, or more portable, SDL_opengl.h
[22:33:11] <Fingolfin> is any part of uwadv actually based on X11 ?
[22:33:42] <Fingolfin> if so, I'll not even go on trying, it's not worth it; if no, why is configure.in explicitly looking for X11 (AC_PATH_X) ?
[22:36:20] <-- QQtis has left IRC ("partytime!!!")
[22:36:30] <Coren_> Sorry, was afk.
[22:37:51] <Coren_> Actually, I don't think any part of uwadv wants X11 specifically; but under X11 SDL requires linking to -lX11-- that might just an artifact of that and you probably can rip the link directive out entirely.
[22:39:28] <Fingolfin> I will have to modify quite some parts of configure.ac I guess
[22:39:53] <Fingolfin> is there any problem with using SDL_opengl.h instead of GL/gl.h, by the way (I am not sure if it was added in SDL 1.2.3 or 1.2.4)
[22:40:18] <-- wjp has left IRC (Remote closed the connection)
[22:40:19] <Coren_> Hmmm. Good question. I see no ill apriori.
[22:40:40] <Coren_> I think it even predates 1.2.3 actually.
[22:41:03] <Coren_> But take anything I saw with a grain of salt; I'm not part of the uwadv team; I'm the low team. :-)
[22:41:10] --> wjp has joined #uwadv
[22:41:11] --- ChanServ gives channel operator status to wjp
[22:41:24] <Coren_> We work together on occasion but I'm not an insider. :-)
[22:41:35] <wjp> grrr, RH8's X is really unstable
[22:41:43] <wjp> it's crashed 3 times in the last 2 weeks now
[22:41:51] <Coren_> Urg. How ugly.
[22:42:46] <Coren_> wjp: Incidentally, what GL are you using? nVidia's?
[22:42:52] <wjp> yeah
[22:42:56] <Coren_> :nods.
[22:43:06] <wjp> I have a tnt2ultra
[22:43:34] <Coren_> With a bit of luck, I'll get some feedback on other cards someday. Up to now, 100% of the reported configs always have an nVidia card in them. :-)
[22:43:42] <wjp> :-)
[22:44:23] * wjp is peeking at the strace of the hanging X
[22:44:27] <Coren_> Sheesh. 1200 hits on my homepage already.
[22:44:50] <wjp> select() is returning ERESTARTNOHAND?
[22:45:00] <wjp> hm, I don't think I ever saw that error before
[22:45:07] * Coren_ RTFS.
[22:45:17] * wjp is grepping through some includes
[22:45:53] <Coren_> Got it. Bug in X: restarted system call new style.
[22:47:03] <Coren_> Shold be dealt with; there's even a comment about this having been done on X's account.
[22:47:19] <Coren_> * Update: ERESTARTSYS breaks at least the xview clock binary, so
[22:47:19] <Coren_> * I'm trying ERESTARTNOHAND which restart only when you want to.
[22:48:10] * Coren_ goes to check on Xfree86's developper site.
[22:48:20] <wjp> where is that copy/paste from?
[22:49:07] <Coren_> /usr/src/linux/fs/select.c
[22:50:38] <wjp> hm, I see
[22:51:00] * Coren_ is checking the Xfree devel cvs tree
[22:51:04] <wjp> :-)
[22:53:23] <Fingolfin> Coren_: SDL_opengl.h is already being used (I think I told Michael Fink about it last time already :-) Apparently it#s only the configure script that wants it
[22:53:45] <Fingolfin> going to disable the X11 & OpenGL checks in configure for now, a proper solution can come later
[22:54:13] <wjp> I've seen X crash in 2 different ways now; once in a loop with a SIGALRM and a select (returning EREST...); the other was in a loop where it was constantly getting SIGALRMs
[22:55:13] <wjp> kind of annoying; I hadn't seen a crash in months with RH72/73
[22:56:14] <Coren_> What exact release of XFree86 are you using?
[22:56:32] <wjp> RH's 4.2.0-72
[22:57:13] <wjp> you can never be sure when they exactly branched from XFree's CVS, unfortunately
[22:57:35] <wjp> I'd tempted to install a non-RH XFree86
[22:58:43] <wjp> s/'d/'m/
[22:59:33] <wjp> RH has a huge amount of patches in there
[23:00:09] <wjp> it's apparently essentialy a 4.2.1... *sigh*
[23:01:32] <wjp> lol, several patches have "FIXME: needs testing to make sure it works" comments :-)
[23:02:31] <Coren_> RH rapes most of their packages.
[23:02:40] <wjp> yeah, I know
[23:02:40] <Coren_> (Another reason why I don't do dist)
[23:02:58] <wjp> I find that for most RH packages I actually take a look inside
[23:03:04] <Coren_> I'm looking at the 4.3 pre code right now and see no reason why this should break.
[23:03:11] <Coren_> I expect it's a RH screwup.
[23:03:23] <wjp> very likely :-)
[23:03:50] <wjp> is there a specific directory or file that contains this code?
[23:04:03] <wjp> if so, I could check if there are any RH patches applied to it
[23:05:31] <wjp> sheesh, this is insane... there must be literally hundreds of patches here
[23:07:21] --> Dark-Star has joined #uwadv
[23:15:07] <wjp> *sigh*... the more I see of RH's packages the more I feel like switching distro's
[23:15:31] <wjp> I spent hours and hours getting an up-to-date tetex installed last week, since RH's is literally years old
[23:15:57] <wjp> and then there's the mess they made of gnome-terminal
[23:16:09] <wjp> Ctrl-C doesn't even work any more if you're running bash in one
[23:17:16] <Fingolfin> you should switch to OS X and use Fink <g>
[23:18:24] <wjp> heh :-)
[23:18:41] <wjp> you know, I should drop by in Eindhoven sometime and you can try to convince me ;-)
[23:19:12] <wjp> did anyone port Fink to Linux yet, btw? :-)
[23:23:42] <Dark-Star> BTW, what is redhat trying to achieve by making KDE and Gnome behave like each other?
[23:24:13] <Dark-Star> Recently I read that they're also trying to make all shells behave similar, so there's no difference between bash, csh , tcsh, ... anymore
[23:24:32] <wjp> lol, that wasn't serious :-)
[23:24:53] <wjp> that was a story on humorix.org or some site like that
[23:25:17] <wjp> that article also said they were merging vi and emacs :-)
[23:25:44] <Dark-Star> hmm.. ok... maybe I should have investigated it further ;-) But it sounded reasonable after their "merge" of KDE and Gnome...
[23:26:17] <wjp> as for making KDE and Gnome behave the same: it's apparently something about giving RH Linux a single 'look', making it more recognizable, and also about reducing support costs
[23:26:51] <wjp> I don't personally agree; but I can understand their motives
[23:26:52] <Dark-Star> so why don't they just ship one window manager and let users, who want the other one, download and install it seperately?
[23:27:01] <Coren_> It's understandable from a suport and corporate point of view; but making both uniform means crippling both as each have their own cute features distinct from the other
[23:27:05] <wjp> that would make them lose half their lose base :-)
[23:27:09] <wjp> s/lose/user/ ugh
[23:27:29] <wjp> Coren_: yeah
[23:27:32] <Coren_> And it would mean taking sides, something which is as dangerous in KDE vs Gnome as it would be in vi vs emacs. :-)
[23:27:46] <wjp> indeed
[23:28:08] <wjp> although arguably this way they risk alienating both gnome and kde users
[23:28:21] <Dark-Star> but all other distros manage to ship both desktops, too...
[23:28:25] <wjp> http://i-want-a-website.com/about-linux/articles/sep02/nullify.shtml
[23:28:34] <Coren_> Personally I use KDE, because I like konkeror and kmail a lot; and the integration is nice.
[23:28:47] * wjp prefers galeon and evolution :-)
[23:28:50] <wjp> but I like having the choice
[23:28:52] <Coren_> But I've used Gnome in the past and it was very nice too.
[23:29:43] <Coren_> I've noticed, however, there is a very strong correlation kde<->gnome vi<->emacs perl<->python
[23:29:44] <Dark-Star> I use KDE too, I think it has more "potential" (through the integration / DCOP interface). But that might be just an emotional choice, I don't know...
[23:30:03] <wjp> <-- gnome/emacs/perl
[23:30:20] <Coren_> Ok, so you're an exception. Heathen! :-)
[23:30:27] <wjp> ;-)
[23:30:38] <wjp> maybe I should try python sometime? :-)
[23:32:37] <Dark-Star> <--- SuSE fan :-)
[23:33:09] <wjp> yikes, scary
[23:33:35] <wjp> :-)
[23:34:32] <Coren_> Well, you know what I think of distros in general.
[23:34:33] <Coren_> :-)
[23:34:34] <Dark-Star> why? At least they write a consistent configuration tool. I remember having to mix Gtk/Curses/Webmin-tools on redhat to configure the system...
[23:34:54] <Dark-Star> ... OK, that was with Red Hat 6.2 I think, maybe that changed since then...
[23:35:11] <Coren_> My package install tool is 'make' :-)
[23:35:31] <wjp> doesn't having a single config tool make maintaining dozens of systems at once a bit annoying?
[23:35:40] <Coren_> And I have the most versatile and powerful configuration editor: vi
[23:35:41] <Coren_> :-)
[23:35:48] <wjp> modularity helps a lot for that
[23:35:52] <wjp> Coren_: hehe :-)
[23:36:16] <wjp> <emacs-mode>emacs is more versatily and more powerful!</emacs-mode> ;-P
[23:36:20] <wjp> s/tily/til/
[23:36:22] <Dark-Star> their configuration tool is modular. With every version, they add new modules (like NIS configuration and such)...
[23:36:28] <wjp> my typing is really going downhill :-)
[23:36:46] <Coren_> My vi speaks perl. Much better than lisp if you ask me. :-)
[23:37:00] * Coren_ chuckles, and pulls out the holy water.
[23:37:02] <wjp> (I must say I haven't used SuSE myself, btw. This is just what rubbed off from one of our sysops' dislike of suse)
[23:37:19] <Dark-Star> (define x (lambda () scheme-and-lisp-rule))
[23:37:23] <Dark-Star> :-)
[23:37:45] * Coren_ stuffs a closure up Dark-Star's butt. :-)
[23:38:18] <Fingolfin> wjp: feel free to drop by in Eindhoven any time =)
[23:38:26] <Fingolfin> wjp: or next weekend in Delft at the ACM contest =)
[23:39:00] <Coren_> Hmmm. I find myself at a juncture. Should I now implement inventory/object manipulation or breathe life into mobiles?
[23:39:12] * Fingolfin personally thinks the RH gnome/KDE unification isn't that bad; don't quite understand why so many people complain so much. as a MacOS user, who learned a long time ago that a unified GUI is a very nice thing to have, I think it's a clever move on the long run
[23:39:13] <wjp> ah, right, ACM
[23:39:22] <wjp> I'm not participating this year, I'm afraid
[23:39:41] <wjp> I'm in Germany from sunday to next saturday, so it's at a bit of a bad time
[23:39:51] <Coren_> Fingolfin: Yes, but the Right Thing is to pick one interface, not strip two down to their lowest common denominator.
[23:40:22] <Coren_> Fingolfin: I expect that is why most people were offended by RH's move.
[23:40:34] <wjp> picking one interface would be a very bad business-decision for RH, though
[23:40:54] <Fingolfin> lcd, not gcd? :-) anyway, what exactly was stripped out of great importance (serious question, not flame bait, I don't know the answer)
[23:41:00] <Coren_> wjp: Exactly. That is why they should have punted and left well enough alone and wait for the 'market' to settle.
[23:41:04] * Fingolfin agrees with wjp
[23:41:13] <Fingolfin> wjp: hehe
[23:41:52] <Coren_> Eventually, one of KDE or Gnome will 'win' once they both are mature enough to settle technical merit disputes. Trying to choose one over another is premature.
[23:41:59] <Dark-Star> but why does RH, one of the biggest linux distros (or _the_ biggest?) have trouble supporting 2 different desktops, while other distros support even more (fvwm and such...)??
[23:42:11] <wjp> others don't really support them
[23:42:30] <wjp> at least, I haven't seen many companies actually offer support contracts
[23:42:43] <Dark-Star> ok, maybe not fvwm (they're installing icons for the applications and that's it) but at least they support gnome _and_ kde...
[23:43:05] <Coren_> Fingolfin: Mostly integration among applications, actually. Some window manager features that some people rely on that are particular to one or another. Lots of small detail mostly, but it's in the details that one chooses one over the other usually.
[23:43:36] <wjp> yeah, it's usually just one or two features you've become accustomed to that keep you from switching
[23:43:51] <Fingolfin> but can't you re-enable those?
[23:43:53] <wjp> s/switching/not caring/
[23:44:32] <Coren_> Fingolfin: Presumably yes, but most people wouldn't want to have to reconfigure their settings just for the sake of humoring a distro, and understandably so.
[23:45:07] <Coren_> Fingolfin: And RH also has the nasty habbit of raping the source of the tools they 'customize' making reenabling things tricky for many.
[23:45:23] <wjp> for our math department we're recompiling/reconfiguring quite a few RH8 packages now
[23:45:39] <Fingolfin> Coren_: the problem as I see it is that RH is trying to comfort people that want to use Linux as a desktop system and come from the windows world. I think those won't care about all the things you list as problems. On the same time, the UNix geeks that could actually fix those "problems" (which are only problems in their eyes), instead complain about RHs move
[23:45:54] <Fingolfin> raping source is bad of course
[23:45:55] <wjp> tetex, parts of gnome/kde; adding things RH kicked out such as mp3 support
[23:46:24] <wjp> Fingolfin: yeah, RH is shifting focus to 'The Desktop'
[23:46:31] <Coren_> Fingolfin: Perhaps. But Linux lives /because/ of the geeks; and a distro that looses support from what is, after all, the developer pool is a Bad Thing.
[23:46:33] <wjp> (whatever that may mean)
[23:46:45] <Fingolfin> unfortunatly, many geeks are very arrogant
[23:47:18] <Fingolfin> well, RH is trying to enlarge the Linux market; people on the one side argue that they want Linux to be used by "normal" users, yet on the other side don't like it when RH tries to achieve exactly that
[23:47:43] <Fingolfin> to non-UNix geeks, the split between Gnome and KDE is extremly irritating and annoying
[23:47:46] <Coren_> Fingolfin: You don't get the point; it's not WHAT they are trying to do that annoys people, it's HOW.
[23:47:51] <Fingolfin> it also splits the Unix community development resources
[23:48:02] <Fingolfin> Coren_: actually I do think I get the point =)
[23:48:33] <Coren_> Fingolfin: If they had chosen one desktop over another, they would have annoyed 50% of geeks, but not lost respect while achieving the same goal of a uniform interface.
[23:48:52] <Fingolfin> Coren_: and even though you claim it's the HOW you don't like, your initial reasons, and those I so far heard of most ppl who complain about this, are not the HOW reasons =)
[23:49:09] <Fingolfin> yes, maybe they should have chosen one desktop
[23:49:19] <Coren_> Fingolfin: And, when you think about it, RH did a common sense fault: what's the /point/ of offering a choice between two desktop environments if you iron out the differences between them in the first place?
[23:49:58] <Fingolfin> in fact I think about the best thing that could happen to Linux is that either Gnome or KDE dies, or they merge, but I don#t think either of these options will ever happen, resulting in a community that is split up foreever, arguing between itself =)
[23:50:02] <Dark-Star> Coren_: yes, that's _exactly_ what I was wondering ...
[23:50:35] <Dark-Star> Fingolfin: AFAIK at least the KDE folks seem to be interested in interoperability/compatibility with Gnome
[23:50:38] <Coren_> Fingolfin: That will happen eventually. (One dying). The problem is that both teams are now religiously entrenched in their differences and have no desire to cooperate.
[23:50:44] <Fingolfin> actually as I understand it, it is less to try to offfer the choice between KDE and Gnome they try to offer, rather they try to allow users to use Koffice *and* Gnumeric, with a similar UI
[23:51:12] <Fingolfin> Coren_: currently both systems seem to do very well, I can't even remotely see one gaining an upper hand, and both have strong supporters
[23:51:43] <wjp> I can't say I see either gnome or kde dying anytime soon
[23:51:53] * Dark-Star places his bets on KDE...
[23:51:58] * wjp hopes for gnome -)
[23:52:04] <wjp> s/-/:-/
[23:52:09] <Coren_> Fingolfin: Actually, I think KDE will be the long term winner regardless of technical merit or not for two reasons: Gnome is backed by RMS which is alienating more and more people as he descends into raving lunacy, and KDE has (marginally) better relations with developers.
[23:52:47] <Coren_> Technically, I think both are currently of equal value.
[23:53:52] <Dark-Star> to me it seems that KDE also has the better development design with DCOP (Embedding applications in each other Worked well for Windows too)
[23:54:17] <Coren_> That's IMHO the best technical strong point of KDE. Gnome has its own benefits as well.
[23:54:19] <Fingolfin> well, we do offer both KDE and Gnome installation with Fink, and I don't care to much about which one will win... and frankly from my experience, both have a long way to go before they can compete with OS X (in my eyes that is, I am not claiming this as a general truth, so there is no need for you to flame now :-)
[23:54:34] <Coren_> Fingolfin: I don't flame.
[23:54:39] <Dark-Star> the Gnome apps are somehow "isolated", i.e. everyone works out his own stuff... KDE seems more "integrated", i.e. all programs use the same KParts...
[23:54:41] <Fingolfin> Coren_: no you don't
[23:54:56] <Fingolfin> Coren_: I just wanted to say: "no need to start bashing me for saying I think OS X is better :-)"
[23:55:13] <Coren_> But KDE is more likely to win for simple political reasons: they are not annoying commercial developers which will, after all, make or break Linux for non-geeks.
[23:55:21] <Fingolfin> anyway, I gotta go to bed now, we can discuss more another day if you like :-)
[23:55:31] <Dark-Star> ok, see ya
[23:55:35] <-- Fingolfin has left IRC ("good night")
[23:56:26] <Coren_> So. Should I start mobile activity or object interaction? :-)
[23:56:50] <Dark-Star> start the easier task, I'd say . . .
[23:57:52] <Coren_> They are both of overall equal difficulty and length; which is why I was probing for opinions. I.e.: which would you like to see first?
[23:57:57] * wjp is staring at a race in scummvm, in the meantime
[23:58:17] * wjp votes for umm... hmm... good question :-)
[23:58:19] <Dark-Star> wjp: which game?
[23:58:23] <wjp> simon1
[23:58:32] <Dark-Star> ok, don't have that one :(
[23:58:52] * Dark-Star votes for Object Interaction...
[23:58:53] * Coren_ will bb in a few
[23:59:00] * Coren_ is away: Food!
[23:59:12] <Coren_> (Votes are logged anyways) :-)