[00:10:30] <wjp> yeah
[00:12:21] <sbx> okay, going to take a break
[00:12:24] --- sbx is now known as sbx|afk
[00:26:23] * wjp posts a seriously annoyed reply to Ignitus' latest post
[00:29:34] <wjp> Darke: yes, I'm pretty sure 'huge's were originally 'struct's
[00:36:16] * Darke reads forum, he hasn't checked it yet this morning. *gasp!*
[00:36:21] <wjp> omg!
[00:36:26] <wjp> how could you! ;-)
[00:36:35] <wjp> oh, wait, your morning :-)
[00:36:38] <wjp> slightly less bad
[00:36:41] <wjp> ;-)
[00:37:04] <Darke> *nod* Thanks. I was pretty sure about the huge/structs thing, since I really couldn't think of any other reason they would exist. *grin*
[00:37:27] <wjp> of course, decompiling them would be 'interesting' :-)
[00:38:36] * Fingolfin reads forum, he hasn't checked it yet this last two months *gasp!*
[00:38:51] <wjp> now where is that kick button? :-)
[00:38:55] <Darke> Yeah. That's why I'll need a 'translation table' to say 'this struct is actually three words', etc. That's where I hit up against with my decompiler and haven't gotten anywhere, I've been trying to work out the best way to do it, whilst coding the compiler as a distraction. *grin*
[00:38:57] --- ChanServ gives channel operator status to Fingolfin
[00:39:00] <Fingolfin> don't kill me, please =)
[00:39:07] * Darke giggles.
[00:46:31] <Fingolfin> after some reading in the forum, I can suddenly understand why wjp might reply annoyed to a post by Ignitus =)
[00:47:04] * Darke snickers.
[00:47:27] <wjp> you should see the post in question
[00:47:41] <Fingolfin> which thread?
[00:47:56] <wjp> top one (Julia Portrait)
[00:48:24] <wjp> hm, compile errors on uwadv
[00:49:29] <Fingolfin> ouch
[00:50:00] <Fingolfin> he does realized that Dominus is author of the FAQ, team member, and probably 10e12 times better informed than he, doesn't he?
[00:50:13] <wjp> apparently not
[00:54:16] <wjp> no matching function for call to `vector<Uint16, allocator<Uint16> >::at (unsigned int &)'
[00:54:22] <wjp> hm, that sounds strange
[00:54:31] <wjp> shouldn't that function exist?
[00:54:59] <Darke> Umm... yeah, I think so. Just a sec and I'll hit the text.
[00:56:35] <Darke> Yeah. It should have the same effect as vector<>, but it throws an exception if it's out of bounds.
[00:56:52] <Fingolfin> yup
[00:57:06] <Fingolfin> that is
[00:57:10] <Fingolfin> unsigned int != Uint16
[00:57:22] <Fingolfin> some C++ compiler can be very ... strange... in this issue
[00:57:40] <Fingolfin> well in this case, Uin16 != unsgined int
[00:57:44] <wjp> but Uint16 is the vector template type, unsigned int is size_t
[00:57:45] <Fingolfin> so the compiler is right to complain
[00:57:57] * Fingolfin hits his head
[00:57:57] <Fingolfin> aye
[00:58:02] <Fingolfin> I am a bit sleepy at 3 Am sorry =)
[00:58:07] <wjp> yeah, same here :-)
[00:58:13] <Fingolfin> it should exist, dunno why it would error
[00:58:24] <Fingolfin> at least it does exist in my C++ libs
[00:58:26] <Darke> Compiler error?
[00:58:44] <Fingolfin> you could check your vector header ...
[00:58:46] <wjp> well, it _is_ gcc 2.96, so... :-)
[00:59:00] <Fingolfin> I remember we had such a problem (or was it with operator?) with my vec.h class
[00:59:20] --> Kirben has joined #exult
[00:59:20] --- ChanServ gives channel operator status to Kirben
[00:59:27] <wjp> hi Kirben
[00:59:37] <Fingolfin> wjp: indeed, look at vec.h
[01:00:42] <Kirben> Hi
[01:00:51] <wjp> right...
[01:00:59] * Fingolfin scratches his head, he forgot what that #ifnded was supposed to fix exactly
[01:01:10] <Fingolfin> a missing at() ?
[01:01:22] <Fingolfin> my 2.95.2 gcc does have at in vector, though, at least it seems so
[01:01:27] <Fingolfin> hi Kirben
[01:01:49] <Darke> Hi Kirben.
[01:02:44] <sbx|afk> hi
[01:02:50] <sbx|afk> kirben
[01:02:52] --- sbx|afk is now known as sbx
[01:03:19] --- Darke is now known as Darke|afk
[01:04:04] <Fingolfin> well I better go to bed now
[01:04:08] <Fingolfin> cya folks!
[01:04:12] <-- Fingolfin has left IRC ("good night everybody")
[01:04:41] <wjp> bed... yes...
[01:04:43] <wjp> good idea :-)
[01:04:46] <wjp> night
[01:04:59] <-- wjp has left IRC ("Zzzz...")
[01:14:05] --- sbx is now known as sbx|afk
[01:15:44] <-- sbx|afk has left IRC ("brb")
[01:43:03] --> sbx has joined #exult
[01:43:46] <sbx> hmm
[01:43:48] --- sbx is now known as sbx|away
[03:06:08] <-- sbx|away has left IRC (Read error: 104 (Connection reset by peer))
[03:08:58] --> sbx has joined #exult
[04:19:38] <-- Nikoras has left IRC (Read error: 104 (Connection reset by peer))
[04:49:09] <-- sbx has left IRC (Read error: 104 (Connection reset by peer))
[05:13:54] --- Darke|afk is now known as Darke
[05:17:03] --> sbx has joined #exult
[05:17:27] <sbx> ey
[05:17:29] <Darke> Hello again. *grin*
[05:17:30] <sbx> hi Darke
[05:17:51] <sbx> did you fall asleep?
[05:18:28] <Darke> Nope. Had to hop off to uni to return a book.
[05:20:23] <sbx> did you see the problem Artaxerxes had compiling rwregress?
[05:20:37] <sbx> silly sdl-config
[05:21:25] * Darke noddles. Agreed. It is quite dumb.
[05:27:07] <sbx> did you see the most recent forum post by wjp? and ignitusdragonn
[05:28:48] <Darke> Not recently. Only the one this morning.
[05:29:33] <sbx> oh yeah you were here
[05:29:43] <sbx> ignitusdragonn replied though
[05:29:51] * Darke noticed.
[05:29:53] <sbx> not surprisingly
[05:29:57] <Darke> Nope.
[06:58:43] --- sbx is now known as sbx|afk
[06:58:47] <sbx|afk> cya
[07:00:28] <Darke> Bye.
[07:49:15] --> Dominus has joined #exult
[07:49:35] --- ChanServ gives channel operator status to Dominus
[07:49:41] <Dominus> hi
[07:50:20] <Dominus> Kirben: if you have some time I have a little problem with make -f makefile.mingw with Msys
[07:50:32] <Kirben> what problem ?
[07:50:52] <Dominus> the install option does not read my u7path
[07:51:05] <Darke> Hi Dominus.
[07:51:39] <Dominus> makefile.win32 does use the changed u7path on my system makefile.mingw does not
[07:52:33] <Kirben> What path are you using ?
[07:52:52] <Dominus> f:\Spiele\u7
[07:53:30] <Kirben> Try f:/Spiele/u7
[08:20:23] <Dominus> sorry was on the phone with my sister
[08:20:28] <Dominus> I'll try that now
[08:22:37] <Dominus> strip Exult.exe -o C:/Ultima7/Exult.exe
[08:22:37] <Dominus> c:\programme\div\curios\mingw\bin\strip.exe: C:/Ultima7/Exult.exe: No such file or directory
[08:22:38] <Dominus> make: *** [install] Error 1
[08:23:08] <Dominus> allthough if I type set in msys it says U7PATH=f:/Spiele/Exult
[08:26:57] <Kirben> Weird, it works fine if I can U7PATH in Makefile.mingw here.
[08:27:55] <Dominus> well if I change u7path in the makefile it should work but doesn't your system setting override it?
[08:28:44] <Kirben> No, I don't think it would
[08:29:00] <Dominus> because it does for the makefile.win32
[08:30:15] <Dominus> or a suggestion: leave U7PATH= commented out in makefile.mingw and just write that it is needed for various install options
[08:32:51] <Dominus> also this seems to have failed on my system: if not exist $(U7PATH)\data md $(U7PATH)\data
[08:34:20] <Kirben> Well there needs to be a default install location, everyone should not need to alter the makefile.
[08:34:27] <Dominus> actually it didn'T fail as u7path isn't there :-)
[08:34:30] <Kirben> That line only exists in makefile.win32
[08:34:43] <Dominus> it exists in makefile.mingw as well :-)
[08:35:01] <Dominus> install: $(EXEC)
[08:35:01] <Dominus> strip $(EXEC) -o $(U7PATH)\\$(EXEC)
[08:35:01] <Dominus> if not exist $(U7PATH)\data md $(U7PATH)\data
[08:35:01] <Dominus> copy mapedit\exult_studio.glade $(U7PATH)\data
[08:35:14] <Kirben> exult 1.1.0cvs ?
[08:35:19] <Dominus> yep
[08:35:35] <Kirben> oh thats the studio makefile
[08:35:35] <Dominus> and if there is a c:\ultima7 dir it works
[08:35:45] <Dominus> oh, oops, you are right
[08:36:00] <Kirben> bbl, dinner
[08:36:09] <Dominus> see you then
[08:38:37] <Dominus> Kirben: when you are back, if we have this default location at least we should also have this location "made" on install so it doesn't fail because it can'T find the dir...
[09:06:26] <Dominus> and I forgot that make -e overrides environment variables in make (I really knew that once :-)
[09:11:35] <Kirben> good point, done.
[09:11:48] --> wjp has joined #exult
[09:11:48] --- ChanServ gives channel operator status to wjp
[09:11:50] <Dominus> thanks
[09:11:55] <wjp> hi
[09:12:00] <Darke> Hi wjp.
[09:12:30] <Dominus> Kirben: in msys how do you paste from memory (copy is just done by marking something)?
[09:12:34] <Dominus> hi wjp
[09:12:45] <Darke> Dominus: Tried the middle mouse button?
[09:12:46] <Kirben> middle mouse button
[09:12:57] <Kirben> and copy just by selection
[09:13:15] <Dominus> well, I don't have a middle mouse button...
[09:13:39] <wjp> try both buttons at once
[09:13:43] <Darke> Dominus: Press both buttons at once.
[09:13:44] * Dominus looks at the docs...
[09:13:46] <wjp> or shift-insert
[09:14:42] <Dominus> nope that doesn't do it either...
[09:19:02] <Kirben> Those are only options, check http://www.rxvt.org/manual.html
[09:29:25] <Dominus> ah, that is not nice, apparently msys only pastes something that was copied in msys not outside of it...
[09:30:06] <Dominus> oops
[09:30:11] <Dominus> strangely it works now...
[09:30:21] <Dominus> shift-insert btw
[09:31:46] <Kirben> ok I divided studio file up now too.
[09:32:05] <Dominus> in mingw and win32?
[09:32:11] <Kirben> yes
[09:32:39] <Dominus> so the run-pkg-config.exe is no longer needed?
[09:33:16] <Kirben> thats still needed if you use windows console.
[09:33:37] <Dominus> Kirben: now that I fianlly use msys you do this. Not fair! :-)
[09:33:58] <Kirben> Would have done it earlier if you have asked...
[09:34:19] <Dominus> hm, actually it did work before, didn't it?
[09:35:59] <Kirben> yes, the install method in studio file was added for windows console since I only used dist myself.
[09:37:06] <Dominus> ah, yes, I remember, I think I had asked you for that (ages ago). Now I know why it reported an error when I tried it on msys just now :-)
[09:52:39] <Dominus> Kirben: in makefile.studio.mingw clean:.... is no longer needed
[09:54:20] <Kirben> I will leave both clean and allclean as options but forgot to change that section.
[09:54:31] <Dominus> k
[10:57:41] * Darke disappears for the night, if you're unlucky, you'll see him in a few hours. *grin* Night!
[10:57:45] --- Darke is now known as Darke|afk
[10:59:20] --> Fingolfin has joined #exult
[10:59:48] <Fingolfin> yo
[10:59:58] <Dominus> hi
[10:59:58] --- ChanServ gives channel operator status to Fingolfin
[11:02:46] <wjp> hi
[11:06:06] <Dominus> http://mickmils.free.fr/Screens/ (Eriadain screenshots - NWN based Alter-U9)
[11:06:40] <Dominus> Pyros looks good
[11:11:54] <wjp> I have to go
[11:11:56] <wjp> bye
[11:12:00] <-- wjp has left IRC ()
[11:44:17] --- Dominus is now known as Dominus|away
[12:06:24] --> wjp has joined #exult
[12:07:06] <wjp> hi again
[12:13:33] --> Colourless has joined #Exult
[12:13:33] --- ChanServ gives channel operator status to Colourless
[12:13:37] <wjp> hi
[12:13:43] <Colourless> hi
[12:54:01] <Fingolfin> yo
[12:54:35] --- Colourless is now known as Cless|Away
[13:11:55] <-- Kirben has left IRC ("System Meltdown")
[13:27:21] <-- sbx|afk has left IRC (Read error: 104 (Connection reset by peer))
[13:41:39] --> sbx|afk has joined #exult
[14:27:28] --- Cless|Away is now known as Colourless
[14:43:23] --> vividos has joined #exult
[14:43:29] <vividos> hi!
[14:43:36] <Colourless> hi
[14:43:46] <wjp> hi
[14:50:31] <vividos> Colourless: the OpenGL scaling thing works like a charm, but when I upload a texture to the graphics card every frame, it gets really slow
[14:51:35] <Colourless> are you creating a new textures, or just continuely updating a single texture?
[14:52:08] <vividos> no, just calling glTexImage2D every frame for a bound texture object
[14:52:41] <vividos> the problem may be that it's a 512x256 sized texture :)
[14:52:47] <Colourless> hmm
[14:53:38] <Colourless> what hardware?
[14:53:57] <vividos> I tried it on my GeForce 2 MX, probably 32MB memory
[14:54:05] <Colourless> some cards don't like rectangular textures too much, and the driver might be scaling the texture
[14:54:25] <Colourless> 32 bit?
[14:54:37] <vividos> the texture? yes
[14:55:48] <Colourless> that really should be ok. have you tried deleting the texture after use, and creating a new one?
[14:55:55] <vividos> a test program showed that uploading a 256x256 is way faster than a 512x512 one
[14:56:11] <vividos> hmm, no. do you think that gives more performance?
[15:03:17] <vividos> btw, using GL_COLOR_INDEX as type for glTexImage2D and providing a palette via glPixelMapuiv is even way slower
[15:04:56] <Colourless> creating a new texture may give higher performance, since it's possible that updating a texture that already exists might cause a performance penalty. I just cant say though, it's not like i've ever done something like that
[15:05:22] <Colourless> using GL_COLOR_INDEX would be unwise, as you found out :-)
[15:06:08] <Colourless> it is best to supply the texture in the same format as is being used on the hardware
[15:08:17] <vividos> is the format the screen bitdepth I selected, or is it the bitdepth I used with SDL_SetVideoMode()?
[15:09:04] <wjp> gtg; bye
[15:09:06] <-- wjp has left IRC ("[x]chat")
[15:12:41] <vividos> hmm, SDL_SetVideoMode() seems to ignore what bitdepth I pass
[15:13:37] <Colourless> the format i was talking about was the texture internal format
[15:16:23] <vividos> how do i get that format? is it a glGetIntegerv() call?
[15:17:20] <vividos> creating and deleting the texture object every frame gives slower framerates than using one everytime
[15:17:36] <Colourless> ok
[15:18:23] <Colourless> i'm just talking about in your call to glTexImage2D, make sure internalformat and format are the same (i.e. make both GL_RGBA8)
[15:18:51] <Colourless> ah, ignore that
[15:18:57] <Colourless> format can't be GL_RGBA8 :-)
[15:19:17] <Colourless> format would be GL_RGBA with internal format being GL_RGBA8 :-)
[15:20:05] <vividos> I'm using GL_RGBA as texture format, but how can I determine that the internal format is GL_RGB8?
[15:20:17] <Colourless> just set it, your hardware supports it
[15:20:35] <vividos> ah ok :)
[15:21:08] <Colourless> you can use glGetTexLevelParameteriv to get the details about a texture
[15:22:56] <vividos> ah ok
[15:23:35] <Colourless> i don't know though if it will actually tell you what format is being used by the hardware
[15:24:30] * Dominus|away comes in and says: Underworld Adventures rocks
[15:24:37] * Dominus|away goes silently away again
[15:25:16] * vividos grins
[15:25:26] <Colourless> generally, i don't always like the way OpenGL works. It's too much of an abstraction from the hardware for my liking in places.
[15:26:48] <vividos> hmm, what does a value of 0 for the internal format means?
[15:27:05] <Colourless> no idea.
[15:27:22] <vividos> never mind, have an OpenGL error
[15:32:09] <vividos> now it returns 1, but that definitely is not GL_RGB8
[15:34:20] <vividos> is the internal format the same parameter that I pass to glTexImage2D? seems so
[15:34:54] <Colourless> I really don't know. i'm not an open gl guru or anything. i've just used it a bit :-)
[15:35:11] <vividos> me too :)
[15:35:27] * vividos hope that the optimizer does it's thing for the rest of the code
[15:35:47] <vividos> hopes, not hop
[15:35:49] <vividos> hope
[15:37:20] <vividos> do you think to use OpenGL for pentagram?
[15:38:13] <Colourless> yeah I do :-)
[15:38:28] * vividos 's english is bad today :)
[15:39:17] <vividos> hmm ... exult has a fps limiter (or?). how many fps would be possible if exult could render as much frames as it wants?
[15:39:41] <Colourless> seriously I have no idea :-)
[15:40:48] <vividos> I have no idea about 2D graphics performance, but if textures were the speed issue with OpenGL, then using 2D would perhaps yield more speed.
[15:41:41] <Colourless> using textures is not a problem
[15:42:01] <vividos> I read somewhere that you have 70 MB of textures (or something)
[15:42:06] <Colourless> uploading textures maybe, but if can just upload and reuse, opengl will be much faster
[15:42:20] <Colourless> yeah, my texture usage isn't exactly efficent at the moment
[15:45:33] <vividos> mine too :)
[15:51:19] <vividos> TSSHP also seems to reuse your xmidi code
[15:52:12] <Colourless> tsshp uses a variation on the code that i developed for them
[15:52:39] <vividos> not the xmidi class from exult?
[15:52:59] <Colourless> 2 reasons 1) they use ANSI C and 2) they are not GPL
[15:53:09] <Colourless> they can't use what is in exult
[15:53:35] <vividos> you're right, they only use C
[15:55:14] * vividos looks into tsshp cvs
[15:59:11] <vividos> nice work!
[16:25:25] <vividos> have to go. bye!
[16:27:48] <-- vividos has left IRC ("http://uwadv.sourceforge.net/")
[16:37:17] <-- Fingolfin has left IRC ("bba reboot")
[16:44:20] --> Fingolfin has joined #exult
[17:36:24] --- Dominus|away is now known as Dominus
[17:37:01] <Dominus> Colourless: what does the zombie function do atm?
[17:52:58] --- ChanServ gives channel operator status to Fingolfin
[17:54:24] <Colourless> don't look at me. i don't know what it is
[17:54:32] <Dominus> hrmpf
[17:54:44] <Dominus> seriously, how far is it?
[17:54:52] <Colourless> all i know there is a zombie flag, and jeffs been doing lots of stuff that it would appear, is related to it
[17:55:45] <Dominus> in you opinion (in regard to the stuff in /opengl)
[17:55:57] <Colourless> it was an accident :-)
[17:56:08] <Colourless> jeff didn't mean to commit the opengl file with the zombie stuff :-)
[17:56:26] <Dominus> he he (but it sounds nice)
[17:57:00] <Colourless> if you check yesterday's logs you should see a few smart assed comments about it :-)
[17:57:15] <Dominus> yeah, but nothing substantial :-)
[17:58:59] <-- sbx|afk has left IRC (Read error: 104 (Connection reset by peer))
[17:59:26] <Colourless> well, what do you expect, we only have short attention spans. we had our minute or 2 of fun, and that's it
[17:59:48] <Dominus> huh? what are you talking about?
[17:59:56] <Dominus> ah the opengl stuff
[18:00:04] <Dominus> I already forgot I asked :-)
[18:01:52] --> sbx|afk has joined #exult
[18:44:53] --> wjp has joined #exult
[18:44:53] --- ChanServ gives channel operator status to wjp
[18:44:53] <-- Dominus has left IRC (Read error: 104 (Connection reset by peer))
[18:44:54] <wjp> hi
[18:46:13] --> azeem has joined #exult
[18:46:30] <Fingolfin> yo
[18:46:43] <Colourless> hi
[18:47:02] <azeem> hello, great exult hackers
[18:47:39] --> Dominus has joined #exult
[18:48:00] --- ChanServ gives channel operator status to Dominus
[18:48:21] <Dominus> that was strange, my pc bluescreened without me doing something "dangerous" (like using opengl)
[18:52:32] <Colourless> what error?
[18:52:58] <Dominus> quite generic
[18:53:06] <Dominus> nothing that pointed to anything...
[18:56:52] <Colourless> probably heat related. ]
[18:57:27] <Dominus> that's what I suspect
[19:08:49] <Dominus> how do they expect us to buy a computer for $125 when we have already spent over $3 millions...
[19:08:58] <wjp> heh :-)
[19:09:25] <wjp> that must be quite a bit higher by now, btw :-)
[19:09:36] <Dominus> I know
[19:09:53] <Dominus> we are probably in the 4 million leage by now
[19:09:57] <Dominus> league
[19:13:27] <-- azeem has left IRC (Read error: 104 (Connection reset by peer))
[19:18:14] --> azeem has joined #exult
[19:23:24] --- sbx|afk is now known as sbx
[19:23:26] <sbx> hi
[19:23:42] <Colourless> hi
[19:23:48] <Dominus> hi
[19:42:44] <-- Darke|afk has left IRC (vinge.openprojects.net irc.openprojects.net)
[19:44:44] <-- Dominus has left IRC (Read error: 104 (Connection reset by peer))
[19:45:17] --> Darke|afk has joined #exult
[21:13:05] <sbx> "This email was intended to be received in HTML format. If you did not receive this email in the proper form, Click Here"
[21:13:50] * sbx wonders why they want him to click some text.
[21:17:34] <wjp> they added a link to the plain text version?! lol
[21:19:35] <sbx> well, it is HTML, but my MUA doesn't support that and I wonder why they would ask you to Click Here if your not seeing the mail in HTML format :)
[21:19:38] <wjp> oh, wait, there is no plain text version
[21:19:50] <wjp> well, that part is shown in html mode too
[21:20:33] <sbx> so thats the ironic thing
[21:20:44] <sbx> you have no reason to click it since your seeing it in the proper form :)
[21:20:58] <sbx> i am not so they want me to click it, but i cant :)
[21:21:42] <wjp> as if you would want to :-)
[21:22:33] <Colourless> what? it sounds like it could be a good deal :-)
[21:23:08] <sbx> my computer was $1000! i got ripped!
[21:24:31] <Colourless> damn right you did
[21:25:30] * sbx clicks the link right away! Oh wait, he can't because he doesn't see HTML.
[21:25:32] * sbx trashes it.
[21:39:58] * Colourless should probably leave
[21:40:02] <-- Colourless has left IRC ("cya")
[22:30:40] --- Darke|afk is now known as Darke
[22:31:17] <wjp> morning
[22:31:31] <Darke> Morning. *bow*
[22:34:05] <sbx> hello
[23:34:05] <wjp> hmm... I think I was supposed to backport some things from MAIN to devel-1-00, right?
[23:34:06] * wjp thinks
[23:35:55] <Darke> IIRC yes. They're probably listed in the logs if you can't think of them. *grin*
[23:36:15] <wjp> the '-c' option and the broken savegame name input, I think
[23:36:27] <wjp> not sure if I want to do the latter, though
[23:36:41] <sbx> :|
[23:42:55] <wjp> the problem is that it requires a small restructuring of the keyboard event handling
[23:43:15] <wjp> I'm not particularly happy about the way I solved it in the main branch yet, either
[23:44:12] <-- Fingolfin has left IRC ("night")
[23:47:35] <sbx> what is odd is why the guardian doesnt show up for me anymore
[23:48:06] <sbx> perhaps it is some setting in the cfg file
[23:48:22] <sbx> different from yours
[23:48:40] <wjp> hm, it gets to the butterfly and then stops?
[23:49:14] <wjp> could you try running it in gdb and see what it's doing when it's been stuck for 10 seconds or so?
[23:51:46] <sbx> alright
[23:55:18] --- Darke is now known as Darke|afk
[23:57:32] <sbx> (gdb) bt
[23:57:32] <sbx> #0 0x402d3d8e in __select () from /lib/libc.so.6
[23:57:32] <sbx> #1 0x4006c6ac in SDL_alarm_interval () at eval.c:88
[23:57:32] <sbx> #2 0xfffff3e8 in ?? ()
[23:57:32] <sbx> Cannot access memory at address 0x5318ec83