[00:02:09] --- Darke|zzZ is now known as Darke
[00:45:01] --> Kirben has joined #exult
[00:45:01] --- ChanServ gives channel operator status to Kirben
[02:41:16] --> SB-X has joined #exult
[02:41:21] <SB-X> hello again
[02:42:05] <Darke> Hello.
[02:44:13] * SB-X loads Exult to continue his game.
[02:47:20] * Darke watches SB-X exult.
[02:49:53] <SB-X> you pervert
[02:49:59] <SB-X> :P
[02:53:09] <SB-X> :-)
[02:53:13] <SB-X> didnt you get a new job?
[02:53:22] * Darke looks innocent.
[02:53:36] <Darke> Yep. It's saturday, so I'm actually online. *grin*
[02:54:26] <SB-X> Oh alright.
[03:03:48] * SB-X wonders why they had to move all the shrines in U7 and confuse us.
[03:04:43] <Darke> Because? *grin*
[03:05:08] <Darke> Besides, the shrine locations in U7 are in the correct location, it's just U6, etc that are wrong since I played U7 first. *grin*
[03:05:40] <SB-X> Well, I did find it.
[03:05:59] * SB-X tries to put Lady Tory's baby Riky into his bag -> "WON'T FIT"
[03:06:57] * SB-X stuffs him in Iolo's pack amongst the bolts.
[03:07:26] * SB-X kicks Iolo's pack repeatedly to make the crying stop.... aah there. Peace and quiet.
[03:09:05] * Darke snickers.
[03:10:56] <SB-X> What does the rune of honor look like?
[03:11:57] <SB-X> its a chalice or a sword
[03:12:09] <Darke> Not a clue. *grin*
[03:12:48] <SB-X> i've been passing the runes around the the proper shrines
[03:39:19] <SB-X> am i still online?
[03:39:48] <Darke> Yep.
[03:40:05] <SB-X> thank you
[03:40:10] * SB-X reaches into his goodie bag.
[03:41:11] * SB-X gives matto|wookin a keychain.
[03:41:41] * SB-X gives exultbot a notepad with at least half of its pages still in place.
[03:41:50] * SB-X gives Kirben a big pink marshmallow.
[03:42:05] * SB-X gives Darke an empty goodie bag.
[03:42:51] * SB-X finds a fresh turnip in his back pocket and tosses it into Darke's goodie bag.
[03:43:05] <SB-X> :-)
[03:43:25] * Darke snickers. Thanks.
[04:35:56] * SB-X reunites the pirate mates.
[04:35:59] <SB-X> "Blacktooth gives you a big hug, then turns away to look for Mole."
[04:38:32] * Darke grins. Nice work Avatar!
[04:41:25] <SB-X> "Thou must be kidding me! Blackie misses me?" ... "I thank thee, stranger, for imparting this information to me." With that, Mole turns away from you, doing a little jaunt.
[04:41:47] * SB-X nods and agrees with Darke.
[04:42:23] <SB-X> thats what an Avatar is for anyway :D
[04:42:23] * SB-X grumbles under his breath.
[04:42:54] <SB-X> who needs to save the world from giant muppets, and find ritualistic murders, when he can reunite old friends? :P
[04:43:03] <SB-X> s/murders/murderers/
[04:43:26] * Darke nodnodnods.
[06:00:32] --> EsBee-Eks has joined #exult
[06:29:21] <-- SB-X has left IRC (Read error: 113 (No route to host))
[09:59:38] * EsBee-Eks goes to sleep.
[09:59:42] <-- EsBee-Eks has left IRC ("Ultima! Ultima! Ultima!")
[10:17:49] --> Fingolfin has joined #exult
[10:17:56] <-- Kirben has left IRC ("System Meltdown")
[10:17:57] <Fingolfin> yo
[10:17:59] --- ChanServ gives channel operator status to Fingolfin
[10:18:45] <Darke> Hi.
[10:20:17] * Fingolfin did implement cubic interpolation, it sounds indeed a lot better than the current code, at least IMHO
[10:22:04] <Darke> Cool.
[10:25:16] <Fingolfin> now I only have to clean it a bit, and make sure to find out who wrote the mixer code in ScummVM to add him in a credit comment or so... I did write the cubic part, but he wrote the fixed point resample originally (or ripped it originally :-). So I only own 50% of that code
[10:25:23] <Fingolfin> of course it's all GPL
[10:26:52] * Darke noddles and grins.
[11:03:46] * Fingolfin thinks that maybe it would be nice to split loooong docs on the web page (like FAQ, docs ...) into multiple pages. We do that for the Fink docs, too: there are multip page, single page and txt versions of most docs, all generated from the same XML source
[11:43:43] --> wjp has joined #exult
[11:43:43] --- ChanServ gives channel operator status to wjp
[11:43:45] <wjp> hi
[11:48:06] <Darke> Hi.
[11:52:03] <Fingolfin> hi
[11:52:09] * Fingolfin will commit his cubic code in a moment
[11:52:49] <Fingolfin> Hey, exactly one month since my last commit =)
[11:54:07] <wjp> :-)
[11:54:20] <Darke> Nice timing. *grin*
[11:54:41] * Fingolfin also looked into those Mac OS X bug reports. He can't reproduce any of them with current CVS
[11:55:02] <Fingolfin> ok, in CVS. Guardian sounds somewhat better now, IMO (in direct compariosn
[11:55:05] <Fingolfin> comparision even
[11:55:17] <wjp> did you try both CVS branches, btw?
[11:56:50] <Fingolfin> no I am only working on head
[11:57:01] * Fingolfin hasn't looked at the branches at all yet
[11:59:52] <Fingolfin> what is the branch name?
[11:59:56] <Fingolfin> devel-1-00 ?
[12:00:30] <wjp> yeah
[12:00:33] <Fingolfin> hm
[12:02:15] * Fingolfin checks it out
[12:02:24] <Fingolfin> it seems there were quite some changes to Audio.cc since the branch
[12:02:34] <Fingolfin> since SDL_mixer is now being used
[12:02:40] <wjp> yeah
[12:02:43] <Fingolfin> alas my cubic change should work in the old branch, too
[12:02:51] <Fingolfin> but I'd prefer if first somebody could test it <g>
[12:03:00] <wjp> hm, I wouldn't put it in devel-1-00
[12:03:13] <wjp> that branch is in bug-fix-mode, so to speak :-)
[12:03:17] <Fingolfin> ok
[12:03:32] <Fingolfin> less work for me, so I don't object :-)
[12:05:27] <Fingolfin> hm, has SDL_mixer and rate conversion code of it's own, or does it leave that completly to the... simple code in SDL itself?
[12:06:47] <wjp> not a clue :-)
[12:07:17] <Fingolfin> /* Build the audio converter and create conversion buffers */
[12:07:17] <Fingolfin> if ( SDL_BuildAudioCVT(&wavecvt,
[12:07:22] <Fingolfin> I guess that means it uses SDL
[12:09:39] <Fingolfin> hrm
[12:10:07] <Fingolfin> it seems the main reason they don't use "accurate" rate conversion is that some audio drivers only accept buffers which have a size that is a power of 2...
[12:10:20] <wjp> :/
[12:10:26] <Fingolfin> and in any case
[12:10:34] <Fingolfin> the conversion system uses "filters"
[12:10:42] <Fingolfin> which are layered
[12:10:48] <Fingolfin> not the most efficient way to do this :-)
[12:11:46] <wjp> hm, no, doesn't sound too efficient
[12:12:11] <Fingolfin> first they convert endianess, then sign, then 8<->16 bit, then mono/stereo, then rate, each iterating over the full data
[12:12:39] <Fingolfin> and of course the mono/stereo convertes have to know if the data they get is 8/16 bit etc...
[12:12:39] <wjp> hm, sounds like a lot of double work is being done
[12:12:55] <Fingolfin> I think, somebody should completely rewrite SDL_audiocvt.c :-)
[12:13:08] <Fingolfin> of course, that would require testing on a lot of platforms, and I am not sure if it would be accepted... hm :-)
[12:15:45] <Fingolfin> of course one still has to somehow seperate the functionality into seperate functions; there are far to many combinations of these to implement each case seperatly
[12:15:59] <Fingolfin> i.e. you have to convert up to 32 intput formats to 32 output formats.
[12:16:21] <Fingolfin> or so :-)
[12:16:34] <wjp> oh, that's only 992 conversion functions :-)
[12:16:54] <Fingolfin> hehe
[12:16:59] --> wjp|gtk2 has joined #exult
[12:17:07] <Fingolfin> hi wjp|gtk2 =)
[12:17:11] <wjp|gtk2> hi ;-)
[12:17:22] <Fingolfin> how's the xchat 1.9 series?
[12:17:22] <wjp|gtk2> xchat 1.9.2
[12:17:38] <wjp|gtk2> I don't really like the font
[12:18:05] <wjp|gtk2> also the interline spacing is huge
[12:18:41] <wjp|gtk2> hardly anything fits into the channel window this way
[12:18:50] <Fingolfin> one could do the audio conversion this way: first convert all to say LE / unsigned / 16 / stereo, then convert to target format. Obviously not efficient either, but one then is down to only 62 (plus two no-op) conversion methods
[12:18:57] <Fingolfin> hehe
[12:19:53] <Fingolfin> in fact
[12:20:06] <Fingolfin> only 16+16 (one each is a no-op), plus a rate converter in the middle
[12:20:52] <Fingolfin> in fact even less
[12:21:00] <Fingolfin> 8bit doesn't need endianess conversion after all
[12:22:46] <-- wjp|gtk2 has left IRC ("I like core dumps")
[12:37:35] --> Kirben has joined #exult
[12:37:35] --- ChanServ gives channel operator status to Kirben
[12:43:38] <wjp> hi
[12:44:26] <Kirben> Hi
[12:45:19] <Fingolfin> hi
[12:57:30] <Fingolfin> in fact, this audio conversion code would be quite simple!
[13:01:50] <Fingolfin> First, convert intut data into 16bit unsigned data (in native endianess); this is trivial to do in one procesdues. Then, perform the rate conversion, taking care of mono/stereo. If both in and out are mono, convert the one channel; if in is mono, out is stereo, convert as mono but duplicate data; if stereo->stereo, convert each channel on its own; if stereo->mono, avarage the channels and convert that.
[13:02:15] <Fingolfin> so now we have a 16 bit unsigned stream of nativ endianess and the correct mono/stereo property
[13:02:23] <Fingolfin> and with the right rate
[13:02:44] <Fingolfin> now we can if necessary convert that to signed/8bit data, or a diferent endianess
[13:02:59] <Fingolfin> this can even all be done in one go, i.e. no need to have an intermediate buffer
[13:16:34] * Darke sleeps now. Night!
[13:16:38] --- Darke is now known as Darke|afk
[14:43:31] <wjp> *sigh*.. I'm going to do something dreadful... reinstall windows
[14:43:54] <wjp> bbl, hopefully :-)
[14:43:59] <-- wjp has left IRC ()
[15:32:57] <-- Kirben has left IRC (Read error: 104 (Connection reset by peer))
[16:02:00] <-- Fingolfin has left IRC ("42")
[17:40:57] --> wjp has joined #exult
[17:40:58] --- ChanServ gives channel operator status to wjp
[17:41:00] <wjp> hi
[18:24:38] <wjp> bbl again
[18:24:40] <-- wjp has left IRC ()
[19:55:59] --> Fingolfin has joined #exult
[19:56:25] <Fingolfin> re
[19:56:27] --- ChanServ gives channel operator status to Fingolfin
[21:27:49] --> wjp has joined #exult
[21:27:50] --- ChanServ gives channel operator status to wjp
[21:27:57] <wjp> hi again
[21:44:16] <Fingolfin> yo
[21:56:49] --> matto has joined #exult
[23:06:37] <-- matto has left IRC ("This feeling.. inside me. Finally found my life, I'm finally free. No longer torn in two. Living my own life by learning f)
[23:28:16] <-- Fingolfin has left IRC ("42")