#exult@irc.freenode.net logs for 6 Jan 2002 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage

[00:00:59] <Kirben> hmm are some pentagram files missing ? u8pal.pal and u8shapes.flx
[00:01:09] <wjp> those are U8 files
[00:01:15] <Kirben> oh
[00:01:54] <Kirben> shape/map viewers compile but not idea if they works
[00:01:59] <Kirben> not=no
[00:10:18] <wjp> they worked for Dominus & Colourless, so I guess they should work
[00:12:23] <wjp> eek... one simple palace screen already draws ~600 globs
[00:13:01] <wjp> (ok, maybe not so simple)
[00:15:19] <Fingolfin> hm, why is display.cc loading u8shapes.flx? seems it only does that to display the count of shapes... wouldn't it be nicer if U8ShapeManager had a method for that?
[00:15:38] <wjp> could be a leftover from when shpdisp.cc and display.cc were one file
[00:16:09] <wjp> yeah, it is.
[00:16:15] <wjp> just remove those lines :-)
[00:16:20] <Fingolfin> looks like it
[00:16:28] <Fingolfin> sure thing :)
[00:18:25] <Fingolfin> the glade problem was fixed by me today :)
[00:18:37] <Fingolfin> that is, wait a sec
[00:18:42] * Fingolfin grumbles
[00:19:27] <Fingolfin> i didn#t fix it not checking for glade, but I did fix it not using the right -I options
[00:19:47] <wjp> that whole Makefile.am looks wrong
[00:19:49] <Fingolfin> which reminds me, I should test exultstudion again, and the gimp plugin, too
[00:19:52] <Fingolfin> yeah
[00:19:59] <wjp> what is that HAVE_GTK thing for anyway?
[00:20:31] <wjp> it probably was for some gtk shapeviewer or something, but we don't build that anymore
[00:20:42] <wjp> do we?
[00:20:43] <Fingolfin> hm might be...
[00:20:54] <wjp> oh, we do
[00:21:05] <Fingolfin> and we do check for glade it seems? exultstudio isn't built when glade misses;
[00:21:16] <wjp> we do check for glade, yes
[00:21:41] * Fingolfin looks at mapedit/Makefile.am and is a bit confused
[00:24:13] <Fingolfin> wjp: do you think it is a good idea to prefix all source files with u8 ? it helps to distinguish it from the exult soure for now, sure, but once you start adding all the game play features it'll get in the way
[00:24:27] * Fingolfin ducks behind Kirben and tries to look invisible
[00:24:35] <wjp> Fingolfin: no, I don't think it's a good idea
[00:25:04] <wjp> in fact, I was planning to remove them, but someone rushed me into 'releasing' them :-)
[00:39:16] <Fingolfin> =)
[00:43:27] <Kirben> can I dcc win32 makefile for pentagram stuff ? so you can add it to cvs ?
[00:43:53] <Kirben> or do I have access to pentagram cvs too ?
[00:44:14] <wjp> you should have write access
[00:48:25] <Kirben> worked :)
[00:53:26] <Fingolfin> wjp: so... if I was to add a class for shape files... would I add that in a new subdir, "files" where u8flx would go later, too? or what should I do? :)
[00:54:25] <wjp> I'd wait with subdirs for a while :-)
[00:57:20] <Fingolfin> 'k
[01:06:51] <wjp> that hole in map 5 is strange
[01:07:59] <wjp> the glob that should be there is displayed too high, like for the other holes
[01:11:23] <Fingolfin> wjp: if you eve rename the source files, I'd appreciate if the filename and class name match (that includes case)
[01:11:47] <wjp> yeah, I don't know what possessed me to have a different case
[01:11:52] <wjp> it annoys me constantly now :-)
[01:17:24] <wjp> coordinates of the road leading up to the hole:
[01:18:20] <wjp> 6143, 23039, 48 -> 6143, 24063, 48 -> 6143, 24575, 48 -> 6143, 25087, 48 -> 6143, 25599, 48 ->
[01:18:41] <wjp> 6143, 26111, 48
[01:18:52] <wjp> x stays the same, y increases by 512 each step
[01:19:09] <wjp> then the chunk that should fill up that hole: 5632, 26112, 48
[01:19:34] <wjp> x and y both 512 too low...
[01:19:44] <wjp> s/chunk/glob/
[01:22:58] * Fingolfin suddenly has a wonderful idea how he can a) make everybody mad at him once more and b) increase the project activity ranking
[01:23:17] <wjp> what, change everything to \r endlines?
[01:23:24] <Fingolfin> * Copyright (C) 2000-2001 The Exult Team
[01:23:24] <Fingolfin> ^^^
[01:23:25] <wjp> update copyright headers to include 2002?
[01:23:28] <wjp> right :-)
[01:23:28] <Fingolfin> erhmm
[01:23:31] <Fingolfin> yeah =)
[01:26:25] <wjp> now... the strange thing is that somehow all glob items need a x,y coordinate adjustment of -512
[01:30:16] <wjp> hmm, they're not 512 too low, but 511 too low
[01:32:53] <wjp> maybe I should make globs 'snap to grid'
[01:34:22] <wjp> bingo
[01:34:44] <wjp> palace fixed too
[01:34:59] <wjp> and I got rid of those pesky -512 adjustments too
[01:37:05] <wjp> ok... that leaves 9 'free bits' in the x and y coordinates of globs
[01:37:24] <wjp> I wonder if they used those 18 bits for any flags... knowing U7's file formats it's possible :-)
[01:40:44] <wjp> ok... committed
[01:41:14] <wjp> I hope that all of the remaining glitches are caused by the shape ordering being too simple
[01:41:28] * Fingolfin updates and tries
[01:41:52] <wjp> (tall objects behind short objects are drawn in the wrong order)
[01:43:56] <Fingolfin> uhmmm
[01:44:04] <Fingolfin> did you forget to add something to CVS?
[01:44:07] <Fingolfin> like globdisp? :)
[01:44:10] <wjp> possibly...
[01:44:16] <wjp> oh, I added it but didn't commit...
[01:44:33] <Fingolfin> ok, I will run "make display" for now then
[01:44:42] <wjp> done
[01:45:01] <Fingolfin> that's much better (the palace I mean)
[01:45:11] <wjp> yeah, indeed
[01:46:37] <Fingolfin> still not perfect. Esp. the building in front of the palace bridge looks weird. e.g. the left most window of it...
[01:47:07] <Fingolfin> or similiarily on the right end of the building :) both ends look weird
[01:47:34] <wjp> yes, that should just be the ordering
[01:47:41] <wjp> same for the bridge
[01:47:44] <Fingolfin> ok
[01:47:52] <Fingolfin> and the tables next to the waypoint, I assume
[01:48:04] <wjp> although I don't get why that bridge section behaves different from the rest
[01:49:58] <Fingolfin> hum..
[01:50:04] <Fingolfin> take a look at glob 238 or 239
[01:50:19] <wjp> yeah, I saw those :-)
[01:50:33] <wjp> 225 too
[01:50:49] <Fingolfin> he
[01:51:09] <Fingolfin> I wonder if that has any special meaning? :)
[01:51:56] <wjp> :-)
[01:52:05] <wjp> I wonder if they're used anywhere
[02:15:40] <wjp> oops... past 3 am already
[02:15:48] <wjp> I guess I should go
[02:15:53] <wjp> bye
[02:16:06] <-- wjp has left IRC ("[x]chat")
[02:31:14] <-- Fingolfin has left IRC ("42")
[03:17:44] --> FoRcE has joined #exult
[03:18:29] <-- FoRcE has left #exult ()
[04:24:44] <-- Kirben has left IRC ("System Meltdown")
[06:36:34] --> Kirben has joined #exult
[06:36:34] --- ChanServ gives channel operator status to Kirben
[06:37:49] <-- Kirben has left IRC (Client Quit)
[06:37:53] --> Kirben has joined #exult
[06:38:31] <-- Kirben has left IRC (Client Quit)
[06:38:34] --> Kirben has joined #exult
[07:39:30] <-- matto has left IRC ("Play Dragon's Lair in linux - http://www.daphne-emu.com - Developers welcome :)")
[09:00:27] --> sbx has joined #exult
[09:22:32] --- sbx is now known as sbx|dvd
[10:12:55] <sbx|dvd> cya kirben
[10:12:55] <-- Kirben has left IRC (Read error: 104 (Connection reset by peer))
[10:12:56] <-- sbx|dvd has left IRC ("Push push push, add jmp mul pop, push call sub mod, in smth say...")
[10:39:50] --> Fingolfin has joined #exult
[10:39:50] --- ChanServ gives channel operator status to Fingolfin
[11:56:58] --- Fingolfin is now known as Fingolfin|afk
[12:10:26] --> Darke has joined #exult
[12:10:26] --- ChanServ gives channel operator status to Darke
[12:10:40] * Darke bows. "Hello."
[12:32:17] --- Fingolfin|afk is now known as Fingolfin
[12:32:22] <Fingolfin> hiya
[12:59:29] --> wjp has joined #exult
[12:59:29] --- ChanServ gives channel operator status to wjp
[12:59:32] <wjp> hi
[12:59:38] <Darke> Hello.
[13:00:55] <Fingolfin> hi willem
[13:01:04] <Fingolfin> I messed up your pentagram stuff, muwahahah
[13:01:07] <wjp> omg... that mouse in display sure slows things down :-)
[13:03:59] <wjp> looks good, though
[13:04:18] <Fingolfin> it doesn't slow down for me, but it is a bit *cough* hackish
[13:05:05] <wjp> hmm, it's really slow here
[13:06:13] <wjp> ...probably because of the full-screen update for every mouse move
[13:06:23] <Fingolfin> yeah
[13:06:27] <Fingolfin> that is what I meant with hackish :)
[13:06:51] <Fingolfin> it wouldn't be to hard to make it only update the relevants parts
[13:07:13] <Fingolfin> but ideally, I would like to add some sort of display manager, which manages a list of dirty rects...
[13:07:18] <wjp> yeah
[13:07:32] <wjp> but maybe the mouse should still be a special case
[13:07:49] <Fingolfin> yeah sure, I can add in some #ifdefs to turn it off :)
[13:08:07] <Fingolfin> or even a command line flag if I want to be especially cunning =)
[13:09:37] <Fingolfin> where does U8 store its text, btw?
[13:09:46] <wjp> usecode
[13:09:57] <Fingolfin> where does it store its usecode? :)
[13:10:02] <wjp> eusecode.flx
[13:10:12] <Fingolfin> ah that explains why I haven't seen anything
[13:10:26] <wjp> it isn't in the static dir
[13:10:40] <Fingolfin> I figured already
[13:10:53] <Fingolfin> no worries anyway
[13:11:33] <Fingolfin> you think it would be posssible to write a quick program that outputs a list of shapeids mapped to a descriptiojn as produces by usecode (as we talked about earlier today :)
[13:12:11] <wjp> well, Ryan wrote a function-name-lister
[13:12:34] <wjp> it isn't really a description, but it does contain the shape name in a somewhat encoded manner
[13:12:50] --> Colourless has joined #Exult
[13:12:54] --- ChanServ gives channel operator status to Colourless
[13:12:55] <wjp> the descriptions would be somewhat tricky, as they sometimes depend on framenum
[13:12:56] <Fingolfin> when you talk about the devil =)
[13:13:02] <wjp> indeed :-)
[13:13:11] <Fingolfin> I would hack something but I can't w/o the file
[13:13:11] <wjp> hi devil^H^H^H^H^HRyan
[13:13:19] <Darke> Hi Colourless! <grin> They appeared to be trying to summon you.
[13:13:23] <Fingolfin> hi ryan :)
[13:13:29] <Colourless> hi
[13:13:43] <Colourless> those evil beings
[13:13:57] <Fingolfin> the thing is, it would be nice to have such a mapping for some things i am playing around with. framenum dependency isn't such a big issue
[13:14:08] * Darke kicks his black wax candles under the couch.
[13:14:08] <wjp> ...or in the case of NPCs the usecode functions for displaying text get quite complex, depending on tons of flags
[13:14:43] <Fingolfin> hmm I see
[13:14:57] <Colourless> good thing you didn't mess up your summon spell. i hear that the spell to summon me can cause quite a bit of damage if it goes wrong :-)
[13:15:04] <wjp> another thing is that it only contains description for some of the more 'interesting' items; not for floortiles, building tiles, etc...
[13:15:24] <Colourless> p.s. you forgot the red candles
[13:15:27] <Fingolfin> Colourless: we nearly swapped the violet candle with the pink-with-green-spots one!
[13:15:48] <Colourless> hee
[13:16:06] <Fingolfin> wjp: the idea is it would give a start; the file could then be extended manually by volunteers =)
[13:16:21] <wjp> :-)
[13:17:01] <Colourless> fingolfin: i have no idea what would have happened then.
[13:18:50] <Colourless> so, why does it appear that 'someone' is attempting to turn pentagram into some sort of game engine?? :-)
[13:19:58] * Darke was just wondering where he'd stick the u8opcodes.txt file, in the exult/ source tree with ucxt, or the pentagram/ one?
[13:20:19] * Darke is, of course, not trying to turn pentagram into some sort of game engine...
[13:20:21] <Fingolfin> Colourless: nobody would do that. But a good map viewer should have some mouse steering, too, no? :)
[13:20:46] <Colourless> :-)
[13:24:06] <wjp> and when you have mouse steering anyway, it really wouldn't be that hard to have the avatar move around too
[13:24:08] <Colourless> i found that a debug build of dispay with no optimisations is really really slow :-)
[13:24:19] <wjp> yeah, it is :-)
[13:24:41] <wjp> ...and once you can let the avatar move around, NPCs and monsters should be easy too
[13:24:52] <Colourless> hehe
[13:25:11] <wjp> and then you're just a usecode interpreter away from a game engine :-)
[13:25:35] <Colourless> first though we should get it to read the files from the proper directories
[13:26:02] <Darke> And you need to have the sound play, just to have some music in the background whilst you looking at all this...
[13:26:47] <Colourless> well, the exult midi player already reads xmis :-)
[13:27:50] <Fingolfin> display is slow right now cause on every mouse move it does a full screen blit -> bad, cough
[13:27:55] <Fingolfin> easy to fix, though
[13:28:10] <Colourless> i think we really need clipping first :-)
[13:28:17] <wjp> Darke: yes, of course
[13:28:24] <Colourless> i was actually thinking about doing that tonight
[13:28:29] <Colourless> clipping that is.
[13:28:37] <wjp> Shouldn't be too hard to do
[13:29:09] <Colourless> instead of slowing the entire thing down by putting the clipping in the drawing function, i was thinking of writing a second one that does clipping, and is only used if the shape needs clipping
[13:29:11] <wjp> Colourless: did you see the palace, btw?
[13:29:20] <Colourless> not since you made changes
[13:29:42] <wjp> Colourless: good idea
[13:29:47] <wjp> do we do this in exult too, btw?
[13:30:07] <Colourless> wjp: no :-)
[13:31:47] <Colourless> what map is central tenebrae?
[13:31:58] <Colourless> nm, found it
[13:32:16] <Colourless> that is so much better. it almost looks perfect
[13:33:09] <wjp> the step from 'almost perfect' to 'perfect' is kind of a nightmare, though...
[13:33:33] <Colourless> i can imagine]
[13:34:04] <Colourless> does map 0 contain all the npcs???
[13:35:05] <wjp> u8mapfmt.txt: "All of the NPC information is stored in map 0 of NONFIXED.DAT"
[13:35:20] <Colourless> i guess that's a yes then
[13:35:46] <wjp> I wonder what he means by "all of the NPC information"
[13:36:48] <Colourless> whoa, there is some serious mouse lag
[13:36:57] * wjp nods
[13:37:03] * wjp is completely innocent, though :-)
[13:37:44] <Colourless> heh, in a really complex spot i got it to about half a second
[13:37:53] <Fingolfin> wjp: can you tell me what the -p option for objdump does? all the online man pages I found don't mention it, but dpkg-shlipdeps uses it...
[13:38:41] <wjp> -p, --private-headers Display object format specific file header contents
[13:38:59] <Fingolfin> ah ok thx
[13:44:14] <Colourless> can someone tell me why this is a little 'silly'
[13:44:14] <Colourless> if (xoff > 32767) xoff -= 65536;
[13:44:14] <Colourless> if (yoff > 32767) yoff -= 65536;
[13:44:54] <wjp> where did I do that?
[13:45:13] <wjp> oh, there
[13:45:13] <Colourless> line 62 of u8shape.cc
[13:45:58] <wjp> hmm, shapes probably don't get that big, do they? ;-)
[13:46:24] <Colourless> you should just be able to change the int xoff and int yoff to shorts
[13:46:40] <wjp> ugh
[13:46:50] <wjp> ignore that last comment
[13:47:07] <wjp> yeah, that would probably work
[13:47:38] <Colourless> :-)
[13:47:41] <Colourless> does work :-)
[13:48:25] <wjp> there should still be plenty of room for optimizations in there
[13:49:26] <Fingolfin> indeed
[13:49:43] <Fingolfin> but well, it's just a map viewer, don't bother to much
[13:57:59] * Darke gets the feeling that we'll be calling this 'map viewer', 'just a map viewer' for the rest of eternity...
[14:02:02] <Colourless> what 2 files do i need to change so i can display shapes to the edge?
[14:03:02] <wjp> u8shapes.cc, l 40
[14:03:18] <wjp> u8glob.cc, l 72
[14:03:34] <wjp> u8shapeman.cc, l 114
[14:03:53] <wjp> u8map.cc, l 97
[14:04:34] <Colourless> u8map is actually somewhat ok
[14:04:53] <wjp> yeah, but 50 may not be enough
[14:05:43] <wjp> I wonder how fast it will be in 320x200
[14:05:46] <Colourless> really each shapes width should be used
[14:06:01] * wjp nods
[14:06:05] <Colourless> at the moment i'll set the border to 50 pixels off the edge
[14:06:16] <wjp> did I already add a bounding rect. function to U8shape?
[14:06:17] <Colourless> it can be changed later
[14:06:32] <Colourless> there is get_dims
[14:06:41] <Colourless> at some other functions as well
[14:07:00] <wjp> get_(xoff|yoff|width|height)
[14:07:04] <Colourless> yeah
[14:07:46] <Colourless> actually, i'll use them
[14:07:54] <wjp> we should probably add a function that checks if any part of the shape is inside a given rectangle
[14:08:22] <wjp> or have a function return the bounding box and add a generic rectangle intersect function
[14:08:31] <Colourless> yeah we should. i'll add it right now :-)
[14:08:34] <Fingolfin> yeah
[14:08:40] <Fingolfin> I am already doing some bounding in Mouse.cc
[14:08:58] <Fingolfin> would be nice if that was in a nice class or so, like the exult Rectangle class maybe...
[14:09:08] <Colourless> yeah that what i was thinking
[14:09:09] <wjp> well, it is GPL, isn't it? ;-)
[14:09:30] <Colourless> but can we rename it:
[14:09:48] <Colourless> windows headers have a structure using the same name
[14:09:58] <wjp> yes, I ran into no end of trouble with that with the original windows port
[14:10:56] <Fingolfin> Rectangle is a bit longish, too, something shorter would be nice
[14:11:10] <wjp> PentagramsOwnCustomRectangle?
[14:11:27] <Fingolfin> how about Rect, and we introduce a Pentagram namespace? :)
[14:11:51] <Colourless> Rect should be ok
[14:17:08] <wjp> what do we do with the copyright headers, btw? Just add your names too? (or create a 'Pentagram team' entity?)
[14:26:35] <Darke> Just reuse the 'Exult team' entity?
[14:27:11] * Darke thinks he should probably drop the relevant copyright notes in the ucxt/ code as well.
[14:56:40] <Colourless> ok, the clipping code seems to work fine
[14:56:56] <Colourless> now all i've got to do it get it to properly display the shapes
[14:57:06] <Colourless> that is, display the shapes on the edges
[15:23:17] <Colourless> ok, done
[15:28:52] <Colourless> committed
[15:30:54] <wjp> nice
[15:31:14] <wjp> looks a lot better this way
[15:31:25] <Colourless> many time better :-)
[15:31:33] <Colourless> s/time/times/
[15:32:45] <wjp> strange... the roof appears to be drawn in the wrong order
[15:32:56] * wjp wonders if he changed anything locally
[15:33:10] <Colourless> i get a few roof errors in places
[15:33:31] <wjp> it's the entire roof here
[15:33:37] <Colourless> the castle?
[15:33:51] <wjp> yeah. Any plane, in fact
[15:33:57] <wjp> bridge too
[15:34:15] <wjp> I must've accidently screwed something up
[15:34:30] <Colourless> what, you didn't screw up on purpose??!?!?
[15:34:38] <wjp> ssshh! ;-)
[15:37:14] <wjp> hmm, you reversed part of the compare operator
[15:37:41] <Colourless> that was probable because i had changes still here since yesterday
[15:38:31] <Colourless> so, it was me who screwed up on purpose :-)
[15:40:56] --- Fingolfin is now known as Fingolfin|away
[15:42:47] <wjp> it's pretty fast in 320x200 fullscreen
[15:43:19] <Colourless> so you've put in fullscreen modes?
[15:43:28] <wjp> just tried it for a second
[15:44:12] <Colourless> u8 looks far too small at 1024x768
[15:47:54] <wjp> yeah
[15:48:18] <Colourless> i know... SCALING! :-)
[15:49:18] <wjp> :-)
[15:49:20] * Darke watches it being integrated into the main exult tree with a 'commandline' option to allow it to be used as a u8 map viewer... <grin>
[15:49:39] <wjp> hehe :-)
[15:49:52] <Darke> Come on... you know you want too. <grin>
[15:49:57] <Colourless> heh, u7 and u8 are far too different :-)
[15:50:49] <wjp> we just have to add enough abstraction to make them the same on a higher level ;-)
[15:51:11] * Darke noddlenods, he was just about to say what wjp said... except less clearly. <grin>
[15:51:37] <Colourless> hehe.
[15:52:08] <wjp> Seriously though, what's next?
[15:52:35] <wjp> hmm, I guess for me renaming all files/classes is next
[15:52:43] <Colourless> int main(int argc, char *argv[])
[15:52:44] <Colourless> {
[15:52:44] <Colourless> if (argc >= 1 && !strcmp(argv[1], "-u8")) return pentagram_main(argc-1, argv+1);
[15:52:44] <Colourless> else return exult_main(argc, argv);
[15:52:44] <Colourless> }
[15:52:50] <Colourless> is that enough abstraction? :-)
[15:52:52] <wjp> lol
[15:53:00] <Darke> Works for me. <grin>
[15:57:45] <Colourless> so can you tell me what to change in the shape ordering code?
[15:58:04] <wjp> reverse the last two comparisons
[15:58:33] <Colourless> only the last 2>
[15:58:36] <wjp> I'll just commit it in a few minutes. I'm changing the whole file structure anyway
[15:58:49] <wjp> but, yeah, only the last 2
[15:59:26] <Colourless> that is so much better :-)
[16:03:47] <wjp> ok, classes & files renamed, and everything compiles again
[16:13:31] <wjp> ...and committed
[16:14:00] <Colourless> in u8 usecode, how does callis work?
[16:14:22] <wjp> 0x0F?
[16:14:45] <Colourless> i think so
[16:15:09] <wjp> 0F xx yy zz
[16:15:17] <wjp> xx = nr. of bytes from stack to be passed to function
[16:15:19] <wjp> yy = nr. of function
[16:15:21] <wjp> zz = ?
[16:16:02] <Colourless> ok. uscode function 562 is for the music eggs. it calls intrinsic BB which i would think is fairly obviously music
[16:16:15] <wjp> could also be that yyzz = usecode function
[16:16:48] <wjp> yes... makes sense
[16:17:05] <wjp> any idea what 11 would be?
[16:17:10] <Colourless> not at the moment
[16:17:21] <wjp> it's used a lot, IIRC
[16:17:37] <Colourless> it's obviously getting something
[16:18:12] <Colourless> it seems to be getting a property
[16:18:15] <wjp> object flags, I think
[16:18:47] <Colourless> because it gets the property, and if it's not zero, it passes it to BB
[16:18:48] <wjp> a lot of places and the return value with a bit
[16:19:23] <wjp> (note: I only guessed the 'bitwise_and' opcode because of this behaviour)
[16:19:48] <Colourless> i'm guessing then the music number is store in the object flags
[16:20:23] <wjp> yes
[16:22:16] <Colourless> so what were the events they you worked out?
[16:22:33] <wjp> we could probably figure out where this object flag is stored in fixed.dat from the music numbers
[16:22:51] <wjp> events? 0 = single click, 1 = double click
[16:22:58] <Colourless> yeah
[16:23:17] <wjp> haven't really looked further
[16:23:29] <wjp> should be easy to get "attacked" and "killed" events
[16:24:16] <Colourless> well, the music egg is event '4'
[16:25:34] <wjp> proximity?
[16:25:53] <Colourless> might be
[16:26:03] <wjp> or maybe just a generic 'egg trigger'
[16:26:10] <wjp> depends on how eggs are implemented
[16:26:19] <Colourless> there are lots of eggs
[16:26:48] <Colourless> each egg type seems to have it's own function
[16:28:01] <Colourless> of course the monegg 1153 uses events 7 and 10
[16:28:16] <wjp> monster egg?
[16:28:22] <Colourless> could be
[16:29:20] <Colourless> the 'eggs' the i can see are
[16:29:32] <Colourless> MONEGG
[16:29:55] <Colourless> (0x49D) (STALEGG)
[16:29:55] <Colourless> (0x49E) (EARTHEGG)
[16:30:04] <Colourless> that should be
[16:30:04] <Colourless> 1181 (0x49D) (STALEGG)
[16:30:04] <Colourless> 1182 (0x49E) (EARTHEGG)
[16:30:11] <Colourless> 1184 (0x4A0) (STMEGG)
[16:30:12] <Colourless> 1185 (0x4A1) (GUARDEGG)
[16:30:15] <Colourless> 1188 (0x4A4) (FISHEGG)
[16:30:23] <Colourless> 1194 (0x4AA) (DOOREGG)
[16:30:23] <Colourless> 1195 (0x4AB) (GHOSTEGG)
[16:30:23] <Colourless> 1196 (0x4AC) (GHOULEGG)
[16:30:23] <Colourless> 1197 (0x4AD) (SKELEGG)
[16:30:27] <wjp> STALEGG and EARTHEGG both use 7 and 10
[16:30:37] <wjp> stmegg 7
[16:30:42] <wjp> guardegg 7
[16:30:51] <wjp> fishegg 7,10
[16:30:52] <Colourless> 1202 (0x4B2) (PALEGG)
[16:31:00] <wjp> dooregg 7,10
[16:31:00] <Colourless> 1212 (0x4BC) (THURGEGG)
[16:31:07] <wjp> ghostegg 7,10
[16:31:16] <wjp> ghoulegg 7,10
[16:31:19] <Colourless> 1223 (0x4C7) (BOOKEGG)
[16:31:26] <wjp> skelegg 7,10
[16:31:34] <wjp> palegg 7
[16:31:44] <wjp> thurgegg F
[16:31:45] <Colourless> that's all i can see for now
[16:31:54] <wjp> gravegg 7
[16:31:59] <wjp> bookegg F
[16:33:10] <Colourless> well, i'm not sure what good that really does except say that eggs use 7 and 10
[16:33:31] <wjp> and 4
[16:33:33] <wjp> (music egg)
[16:33:51] <Colourless> yeah
[16:34:34] <wjp> I wonder why there are things like "ghostegg", "ghoulegg", but no "kithegg", "eyebeastegg", "demonegg"
[16:34:53] <wjp> or maybe these are just for non-hostiles
[16:35:18] <Colourless> i would guess that examining the functions might help to reveal that
[16:35:36] <Colourless> what about a torax too
[16:36:05] <wjp> good point
[16:37:01] <wjp> it would help if we knew some more intrinsics :-)
[16:38:33] <Colourless> yeah
[16:40:16] <wjp> heh, look at the big red 'G' in front of Salkind's house :-)
[16:40:26] <wjp> (map 41)
[16:44:41] <Colourless> yeah
[16:44:54] <Colourless> guardian voice?
[16:45:05] <wjp> could be
[16:46:19] <Colourless> the roof is being drawn quite wrong there
[16:46:24] <wjp> yeah
[16:46:54] <wjp> there's one big flaw in the current compare function: it sorts primarily on z-coord
[16:47:36] <wjp> because of that, any tall object behind a short object will be drawn after that short object
[16:50:43] <Colourless> what is globdisp?
[16:50:56] <wjp> displays globs
[16:52:08] <Darke> The viewer's readme mentions a 'nonfixed.dat', where does this come from? My copy of u8 doesn't seem to have it.
[16:52:10] <wjp> nothing spectacular, but I needed it for debugging some glob-related issues last night
[16:52:20] <wjp> nonfixed.dat is from gamedat
[16:52:48] <Darke> gamedat has 0 files in it...
[16:53:50] <wjp> you can probably extract one from u8save.000
[16:54:17] <Colourless> it's just a flex?
[16:54:27] <wjp> u8save.000? no
[16:54:47] <Darke> The filename is in there... I just need to find an unarj program.
[16:55:41] * Darke guesses it's arjed anyway, since there's a dearj.exe program in the root directory and no obvious use for it.
[16:55:49] <wjp> ultima8.001 is an arj, yes
[16:57:47] <wjp> once you have u8save.000:
[16:57:52] <wjp> dd if=u8save.000 of=nonfixed.dat skip=47 bs=1 count=242096
[17:01:36] <Darke> Thanks. <bow>
[17:04:24] * Darke oooohs, it's rather pretty, he's never actually _seen_ u8 before...
[17:05:59] <wjp> yeah, u8's graphics are rather good
[17:07:50] * Darke wonders how difficult it would be to map u7's tileset to their 'equilivant' (or close) tiles in u8.
[17:08:50] <Colourless> probably fairly difficult
[17:08:59] <Colourless> they do have different tile sizes
[17:11:51] * Darke nods.
[17:12:36] <Colourless> you know, you missed out on all the fun of the clippingless version of display. it was a guessing game of 'when will it crash next' :-)
[17:14:52] <wjp> :-)
[17:15:14] <-- Colourless has left IRC ("Got Coloured")
[17:15:26] <Darke> I'd rather not. <grin> I'm in the process of trying to write a conf/ file parser alike that Does What I Want(tm). It's enough of a guessing game of 'when will it crash next'.
[17:15:27] --> Colourless has joined #Exult
[17:15:27] --- ChanServ gives channel operator status to Colourless
[17:15:33] <Colourless> err opps :-)
[17:15:43] <Darke> (repose) I'd rather not. <grin> I'm in the process of trying to write a conf/ file parser alike that Does What I Want(tm). It's enough of a guessing game of 'when will it crash next'.
[17:16:30] <Colourless> but you would have gotten some cool looking graphics... in the part of the screen that actually displayed them :-)
[17:16:45] <wjp> hmm?
[17:17:12] <wjp> I never really noticed any screen corruption
[17:17:13] <Colourless> the large black border? :-)
[17:17:26] <wjp> oh, that's what you meant
[17:17:29] * wjp reparses sentence
[17:17:32] <Colourless> the lovely jagged edges on it :-)
[17:17:49] * Darke snickers.
[17:18:47] <Colourless> do you still have the screenshot from yesterday?
[17:19:05] <wjp> yeah
[17:19:11] * Darke seems to remember seeing that screenshot.
[17:19:13] <wjp> http://www.math.leidenuniv.nl/~wpalenst/u8display.jpg
[17:19:44] * Darke sees... or rather doesn't see. <grin>
[17:29:21] <Darke> Hmm... time to sleep. Night all. <grin> Good luck with your 'map viewer'!
[17:29:29] <wjp> goodnight
[17:29:37] <-- Darke has left IRC ()
[17:30:11] --- wjp is now known as wjp|dinner
[17:31:36] --> matto has joined #exult
[17:39:52] --- Fingolfin|away is now known as Fingolfin
[17:40:02] <Fingolfin> rehi
[17:40:39] <Colourless> hi
[17:41:38] <Fingolfin> hm, these makefiles are getting quite unhandy...
[17:42:27] <Colourless> which makefiles?
[17:44:54] --- wjp|dinner is now known as wjp
[17:45:08] <wjp> auto{conf|make} time?
[17:47:56] <Fingolfin> actually
[17:48:13] <Fingolfin> I am right now wondering about this; and I am reading once more the "Recursive Make Considered Harmful" document
[17:48:40] <Fingolfin> I'd like to try and see if it can be applied..... using autoconf would be good in any case, but maybe we can do w/o automake? :)
[17:48:49] <wjp> speed, dependencies, ...
[17:49:08] <wjp> do you have an URL for that document?
[17:50:34] <Fingolfin> http://www.pcug.org.au/~millerp/rmch/recu-make-cons-harm.html
[17:50:40] <wjp> thx
[17:51:52] <wjp> lengthy :-)
[17:52:11] <wjp> do you know of any existing automake-like solutions?
[17:52:12] <Fingolfin> you don't have to read it all, but I find it very interesting, so I recommend it.
[17:52:34] <Fingolfin> no. then again, if I look at what automake does, i wonder what all of it is really necessary for us
[17:53:02] <wjp> yeah, we hardly use any of its functionality in exult. Pentagram should use even less
[17:57:23] <wjp> "The physical memory of modern computers exceeds 10MB for small computers, and virtual memory often exceeds 100MB" *grin*
[17:58:21] <Fingolfin> hehe
[17:59:35] <wjp> ah... so that is what ":=" does...
[17:59:40] <Fingolfin> yeah :)
[17:59:42] <Colourless> what does it do?
[17:59:52] <wjp> immediate evalution, instead of delayed
[18:00:03] <wjp> evaluation, even
[18:01:55] <wjp> interesting article
[18:02:00] <wjp> (still reading...)
[18:10:09] <wjp> ok, so what do we do about this?
[18:10:18] * wjp is now convinced recursive make is evil :-)
[18:10:47] <Fingolfin> hehe
[18:10:50] <Fingolfin> I am already testing
[18:11:04] <Fingolfin> works fine it seems, but then, our test case is not exactly extensive yet =)
[18:11:18] <Fingolfin> I had to do some tweaking compared to the article, since we use c++
[18:11:27] * wjp nods
[18:12:56] <wjp> hmm, VPATH support through configure is important too
[18:14:18] * Fingolfin tries to remember whcih symbol represents the list of dependencies again
[18:14:32] <wjp> ?
[18:14:43] <wjp> oh, I see
[18:16:23] <wjp> $^, I think
[18:16:49] <wjp> or maybe $+ (depending on what you want to do with it)
[18:17:47] <wjp> looks like make can become about as cryptic as perl :-)
[18:19:22] <Fingolfin> well, IMHO the new Makefile I have here now is much more flexible
[18:19:33] <Fingolfin> much easier to maintain
[18:20:30] * Fingolfin adds clean target
[18:27:18] <Colourless> do the maps in u8 have any sort of defined boundry?
[18:29:04] <wjp> Colourless: no, not AFAIK
[18:29:19] <Colourless> hmm, interesting.
[18:29:37] <wjp> Fingolfin: looking forward to seeing it. Commit whenever you feel like it, broken or not :-)
[18:29:54] <Fingolfin> ok, I will do it now then, there is one annoying issue (or not)...
[18:32:19] <Fingolfin> in CVS now; has been working fine for me
[18:33:35] <Fingolfin> only gotcha is: it always wants to rebuild the .d files. Hence, if one removes them in the clean target, then on each "make clean" it would first build the .d files, then delete them... annoying. for now I don't delete the *.d files
[18:34:04] <Fingolfin> OTOH, adding a new source file/app target is trivial
[18:34:30] <Fingolfin> all *.cc files will be compiled automatically, if needed. If you want to add a source file to an app, just add "foobar.o" to the depends list of that app, that's all
[18:41:44] <Fingolfin> ha interesting
[18:41:57] <Fingolfin> it seems the GNU make docs have copied some of his text, and give similiar examples :)
[18:55:32] <wjp> you know, it would be nice if the .d files were stored in a .deps subdir or something
[19:05:43] <Fingolfin> hm, shouldn't be a big problem I think..
[19:08:43] <Colourless> wjp, how does your sorting algorithm work?
[19:09:27] <wjp> it just calls STL's sort on the list of ScreenShapes
[19:09:36] <Colourless> and?
[19:10:03] <wjp> sort uses ScreenShape::operator<
[19:10:27] <wjp> which currently sorts on z coordinate first, and on x+y after that
[19:12:05] <Colourless> hm, ok
[19:14:07] <wjp> Fingolfin: the way automake handles the dependency/clean problem, is that it uses "-include" to include the dep files, which doesn't seem to create the dep files when they're missing
[19:14:32] <wjp> you just need to recreate them at the right time, them
[19:14:36] <Fingolfin> I do use -include; the only difference with it is that it doesn't give a warning when it is missing a .d file
[19:15:15] <wjp> hmm, oh, right
[19:15:19] * wjp seems to be going blind
[19:17:51] <Fingolfin> anyway, there must be some trick to this
[19:18:04] * wjp is still trying to figure out an automake Makefile
[19:23:13] <wjp> it looks like it never explicitly builds the dep files
[19:23:25] <wjp> it just builds them as a side effect of the actual compilation
[19:24:03] <wjp> and since there's no rules for the .P file, it doesn't build it, and ignores it
[19:28:10] <Fingolfin> hm
[19:32:00] <Fingolfin> hm
[19:32:10] <Fingolfin> I could try to move up the Makefile on level, and then let it build the tools, too
[19:32:15] <wjp> Colourless: The 'broken' automaton only uses 'he'. The commander of the SS fort only uses 'she'.
[19:32:22] <Colourless> ah
[19:33:41] <Colourless> deleted the post :-)
[19:34:49] <wjp> "Elude me."... what a strange comment to make
[19:36:27] <Fingolfin> is there a way to get a list of maps for U8 ? all numbers I tried randomly (20,22,60,...) so far only yielded blackness, even though I scrolled around a lot
[19:36:49] <wjp> the 'divination' function has something like that
[19:37:04] <wjp> (usecode function, I mean)
[19:37:55] <wjp> I'll create a small file in docs/
[19:45:14] <wjp> docs/maps.txt, committed
[19:46:32] <Fingolfin> thx
[19:46:57] --> Dominus has joined #exult
[19:46:57] --- ChanServ gives channel operator status to Dominus
[19:47:01] <wjp> hi
[19:47:07] <Dominus> hi
[19:47:09] <Colourless> hi
[19:47:33] <Dominus> wow, not an empty house today :-)
[19:47:41] <Fingolfin> hi
[19:47:44] <Dominus> Exult won't compile for me still
[19:48:05] <Dominus> g++ -mwindows -o tools/expack.exe tools/expack.o U7file.o Flex.o IFF.o Table.o F
[19:48:06] <Dominus> lat.o utils.o listfiles.o crc.o -mconsole
[19:48:06] <Dominus> g++: utils.o: No such file or directory
[19:48:06] <Dominus> make: *** [tools/expack.exe] Error 1
[19:48:34] <Colourless> what version of gcc?
[19:49:52] <wjp> what does 'make utils.o' do?
[19:50:42] <Dominus> where is utils.o
[19:50:44] <Dominus> ?
[19:51:00] <Colourless> doesn't matter, just do it
[19:51:12] <Dominus> make: *** No rule to make target `utils.o'. Stop.
[19:51:34] <wjp> do you have a files/utils.o?
[19:51:46] <Colourless> .cc doen't you mean
[19:51:49] <wjp> no, .o
[19:51:58] <Dominus> yep, I do
[19:52:07] <Colourless> delete it
[19:52:08] * Fingolfin gets a parachute when viewing map 7
[19:52:22] <wjp> Fingolfin: me too
[19:52:31] <Fingolfin> good :)
[19:52:49] <Dominus> ah that did it
[19:53:04] <Colourless> i get a crash too
[19:53:06] <Dominus> shouldn't make clean do that for me?
[19:53:22] <Colourless> no, because that file isn't meant to exist
[19:53:27] <wjp> 'make clean' doesn't delete UCXT_OBJS
[19:53:34] <wjp> (files/utils.o is in UCXT_OBJS)
[19:53:59] <Colourless> the crash is being caused by 'd' passed to Glob constructor being 0
[19:55:07] <wjp> yeah
[19:55:24] <Dominus> thanks for "fixing" me up :-)
[19:55:34] <wjp> it might break again when you build ucxt
[19:56:04] <Dominus> wjp (and others): is the FAQ good to go (the Pentagram stuff in it needs your approval)
[19:56:22] <wjp> hm? can I preview it anywhere?
[19:56:24] <Dominus> wjp: so it is included in make tools
[19:56:38] <Dominus> the faq is in docs/faq.html
[19:57:43] <wjp> hehe, 1.5 is nice :-)
[19:58:09] <wjp> it's no longer just me doing things, though
[19:58:16] <Colourless> ok, mananged to prevent the map 7 crash
[19:58:43] <Dominus> wjp: I noticed :-)
[19:59:08] <Colourless> something is causing one of the globs to not be read
[19:59:10] <wjp> 1.4 might need some rephrasing...
[19:59:25] <Dominus> he he
[19:59:52] <Colourless> :-)
[19:59:52] <Dominus> but I think with the 1.5 question it is kind of okay
[19:59:58] <wjp> Colourless: glob nr. 3070 is actually empty
[20:00:20] <Colourless> i've just put some checks in for null pointers
[20:00:38] <Dominus> I still think make clean should kill everything that gets in its way that gets generated by make tools
[20:01:04] <wjp> maybe we should clean up the static makefiles sometime
[20:04:03] <Dominus> wjp: please take also a look at readme.html if I got all the hidden funtions of exult.exe right
[20:04:52] * Fingolfin wonders where on map 13 anything is
[20:05:14] <wjp> Dominus: there's also a '-v/--version' option
[20:05:28] <wjp> and a -h/--help one
[20:05:36] <Dominus> ok, good
[20:06:17] <wjp> "nethertheless"
[20:06:44] <wjp> (in the -nocrc section)
[20:07:14] * Fingolfin find the loong line of items on map 14 interesting (2048,-464)
[20:07:19] --- Fingolfin is now known as Fingolfin|dinner
[20:07:39] <wjp> 40 has a line of foodstuffs
[20:07:51] <Dominus> aehm, -v and help do ntohing for me
[20:08:10] <Colourless> nether-the-less or never-the-less .... which should it be.... As far as I know, they are both used
[20:08:12] <wjp> I have a strong feeling I may be interpreting item coordinates the wrong way :-)
[20:08:31] <wjp> nethertheless? heh, never heard that one before... never mind then
[20:08:47] --> lex has joined #exult
[20:09:12] <Colourless> hello
[20:09:12] <Dominus> so I didn't have it wrong after all
[20:09:23] <wjp> hi
[20:09:29] <Dominus> hi
[20:09:33] <wjp> -v works for me
[20:09:40] <wjp> and so does -h
[20:09:48] <Dominus> maybe not on windows?
[20:09:58] <wjp> should work everywhere
[20:10:03] <Dominus> Colourless?
[20:10:26] <wjp> how do they 'not do anything' btw? just start exult normally or quit without doing anything?
[20:10:37] <Dominus> quit without doing
[20:10:49] <wjp> oh... anything in stdout.txt?
[20:11:11] <Dominus> nope
[20:11:16] <wjp> stderr? :-)
[20:11:19] <Colourless> dominus?
[20:11:42] <Dominus> in stderr only the platform and build stuff
[20:11:47] <wjp> that's -v
[20:11:51] <Dominus> Colourless: does it do anything for you
[20:11:59] <Colourless> i don't have exult built
[20:12:19] <Dominus> ah in stderr it shows the h as well
[20:12:30] <Dominus> ok, then I know what to write :-)
[20:12:52] <wjp> not exactly desired behaviour on windows :/
[20:13:03] <Dominus> right
[20:13:04] <Colourless> when why is stderr used?
[20:13:25] <Colourless> it's not exactly an error message, no?
[20:13:40] <wjp> should be cout or clog, I guess
[20:13:46] <Dominus> -game GAMENAME Set the game data name to play (refer to the documentation)
[20:13:52] <Dominus> hm
[20:14:08] <wjp> looks like the documentation is finally catching up ;-)
[20:14:10] * Colourless wonders if he should add in an option to allow the devlopers to build exult for the console in win32
[20:14:51] <wjp> Dominus: incidentally, -game doesn't work for anything except -buildmap
[20:15:08] <Dominus> yeah so the way I wrote it is fine
[20:15:40] <Dominus> actually it would make more sense if buildmap only responds when either -si or -bg is set
[20:15:58] <Dominus> and not -game
[20:17:35] * wjp nods
[20:20:43] <wjp> done
[20:20:51] <Dominus> cool
[20:20:56] <wjp> -game option is gone
[20:21:00] <Dominus> now I have to rewrite :-(
[20:21:04] <Dominus> he he
[20:21:18] <wjp> for now, anyway. It'll probably be resurrected when we allow custom games (someday)
[20:21:22] <Dominus> and buildmap needs -bg
[20:21:26] <Dominus> ?
[20:21:30] <wjp> yeah, -bg or -si
[20:22:26] <Dominus> I tired on friday morning but didn't have enough patience to actually build it :-)
[20:22:45] <Dominus> tried not tired altough I was tired :-)
[20:22:47] <wjp> stdout/stderr should show progress
[20:23:08] <wjp> it does take a while, doesn't it? :-)
[20:23:10] <Dominus> it was nice that all the map parts were there after I killed the process
[20:23:34] <wjp> you should start a new game before trying it, btw
[20:23:54] <Colourless> buildmap should take a param to specify which superchunk to start and, and which one to end at
[20:24:02] * Dominus understands what wjp means
[20:24:12] <wjp> might be nice, yes
[20:27:38] <wjp> hmm, I think I'll change all the options to the more-or-less-standard "--" if the option is longer than 1 letter, "-" if it's a single letter
[20:27:51] <wjp> (so "--buildmap", "--bg", etc...)
[20:28:20] <Dominus> ok with me :-)
[20:28:42] <Colourless> um i just started a game of SI, and things really scrwed up when all the items were meant to be exchanged after the storm... everything was put on the ground
[20:29:09] <wjp> hehe :-)
[20:29:27] <wjp> sounds like one of the fixes to all the give/take functions
[20:29:44] <Dominus> hm, works for me
[20:30:19] <Colourless> let me check
[20:31:09] <Colourless> ah ok, it's not a problem. i just had gotten my wieght to the maximum
[20:31:38] <wjp> *phew*
[20:32:50] <wjp> ok, cmdline options committed
[20:33:34] <Dominus> did you change the output of -h -v as well?
[20:34:08] <wjp> -h shows the proper syntax
[20:34:14] <wjp> I didn't change -v
[20:34:24] <Dominus> ok
[20:35:51] <Colourless> you are probably going to cause me conflicts in the changelog
[20:36:44] * wjp should remember to actually exit after a fatal error...
[20:41:15] <wjp> hmm, bug in the Args class
[20:43:27] <-- matto has left IRC (Connection timed out)
[20:45:16] <wjp> hmm, buildmap is broken :/
[20:45:29] <Dominus> oh
[20:47:04] <wjp> (because of Jeff's Gamemap changes)
[20:49:48] <-- lex has left #exult ()
[20:55:49] --- Fingolfin|dinner is now known as Fingolfin
[20:59:47] <wjp> Fingolfin: if you want to compare with automake's dependency generation, look for "%.o: %.cc" in exult's main Makefile
[21:00:11] <wjp> it looks somewhat similar
[21:00:26] <wjp> but, as I said earlier, it's done as a side effect of the actual compiling
[21:01:01] <Fingolfin> I don't have such a rule, but a .cc.o rule, and I already looked at it
[21:01:21] <Fingolfin> yeah, as a side effect, which has both good and bad effects
[21:01:30] <Fingolfin> good: it solves the "make clean" problem
[21:02:47] <Fingolfin> bad: it will regenerate dependencies unnecessarily in some cases. but that should be fairly minor
[21:03:08] <wjp> hmm, which cases?
[21:03:28] <Fingolfin> hm no wait forget what I said
[21:03:35] * Fingolfin was thinking stupid
[21:03:51] <Fingolfin> I am actually not sure I see any problem with it right now
[21:03:56] * Fingolfin goes to try it out
[21:04:57] --> matto has joined #exult
[21:09:54] <Dominus> someone care to fix makefile.win32 for pentagram again?
[21:10:08] <Dominus> if bnot don't worry :-)
[21:10:13] <Dominus> if not
[21:10:17] <Colourless> what's wrong with them?
[21:11:29] <Colourless> it works fine for me
[21:12:35] * Dominus whistles innocently
[21:12:59] * Dominus should take a look at what it does sometimes
[21:13:06] <Colourless> well, what 'was' your problem
[21:13:22] <Dominus> that it only built display and not the other stuff
[21:13:32] <Colourless> oh, i can fix that real quick
[21:13:35] <Dominus> with just a make -f maefile.win32
[21:13:52] <wjp> all: viewer/display.exe viewer/shpdisp.exe viewer/globdisp.exe
[21:14:05] <Dominus> if you fix it can you make it build with install as well (like exult makefile)?
[21:14:14] <Dominus> wjp: I noticed it then :-)
[21:14:26] <wjp> install: all
[21:14:51] <Dominus> ah
[21:15:13] * Dominus does not understand all there is to understand but he is slowly learning
[21:15:39] * Dominus didn't know that you can do a install: all
[21:16:03] <wjp> no... I meant "install: all" should be the line in Makefile.win32 instead of "install:"
[21:16:07] <Fingolfin> what good for is an install target here? :)
[21:16:26] <Fingolfin> anyway, the changed ("normal", not win32) Makefile works fine, guess I'll commit it soon
[21:16:40] <Colourless> allinstall
[21:16:41] <Dominus> wjp: I noticed that now
[21:16:56] <Colourless> but install is still wrong
[21:17:07] <Colourless> i'll fix it
[21:17:15] <Dominus> thanks
[21:17:59] <Dominus> when is pentagram going to look int the appropiate subfolders (static and what else it needs)? :-)
[21:18:15] * wjp sighs... multiplying in F16 (finite field of 16 elms) is no fun at all...
[21:18:23] <wjp> Dominus: dunno :-)
[21:18:36] <Dominus> he
[21:18:50] <wjp> it's already quite an achievement that it actually tells you which file it can't open ;-)
[21:18:57] <Dominus> he he
[21:21:11] <Colourless> ok, committed
[21:23:39] <Fingolfin> commited, too
[21:23:55] <wjp> Colourless: you should put 'all' on top
[21:24:26] <Fingolfin> why does Makefile.win32 sometimes use / and someitmes \ ?
[21:24:40] <wjp> because it can?
[21:24:50] <wjp> dunno really
[21:24:50] <Colourless> yeah
[21:24:59] <Colourless> because it can :-)
[21:25:26] <Colourless> actually, it can cause problems in some places such as with 'copy'
[21:26:27] <wjp> Fingolfin: works great
[21:26:35] <Fingolfin> hmm, ok
[21:26:47] <Fingolfin> so now I need to break two things to settle my negative score again :)
[21:27:02] <wjp> :-)
[21:29:03] <Colourless> ok, i've put all at the to
[21:29:05] <Colourless> p
[21:33:10] <Dominus> mhm, should the arrow work in display?
[21:33:23] <wjp> no, it's just decorative
[21:33:26] <Dominus> mouse arrow I mean
[21:33:34] <wjp> yeah
[21:33:39] <Dominus> seriously?
[21:34:00] <wjp> yeah, it should just move around. You can't use it for anything
[21:34:09] <Dominus> ok then
[21:34:09] <Fingolfin> hey, come on, what do you expect?`:)
[21:34:27] <Dominus> mhm, from reading the logs it wasn't clear to me
[21:34:30] <Fingolfin> it could be changed to scroll, though.... no? :)
[21:35:00] <wjp> sure, that wouldn't be really hard
[21:35:11] <Fingolfin> Dominus: look, if it did anything, it would be useful.... that would complete ruin my score. this way, it sonly lows down redraw, so it's OK :)
[21:35:21] <Dominus> pfff
[21:35:24] <Dominus> LOL
[21:35:56] * Fingolfin notices his interesting typo, and wonders how he pulled that one
[21:36:45] <wjp> yeah, that 's' sure is out of place :-)
[21:37:15] <Fingolfin> sonly lows -> only slows :)
[21:39:43] * Colourless get back to writing a function to center the map on startup :-)
[21:40:56] <Fingolfin> how about a tiny overview at the top right (or left) ? :)
[21:41:06] <Fingolfin> and scalers sure would be useful *cough*
[21:42:09] <Colourless> no, that's too useful
[21:42:21] <Dominus> speaking of scalers: Colourless, Derek asked if or when you would integrate his updated/enhanced point (or whatver its name was) scaler
[21:42:40] <Colourless> yeah I guess i can
[21:48:12] * wjp wonders how far we'll fly up the SF activity ladder after today
[21:48:38] <Fingolfin> not much
[21:48:42] <Fingolfin> they revamped it
[21:49:09] <Dominus> in what way?
[21:49:53] <Fingolfin> it works a lot different than it used to in those days when we had super high activity ranking :)
[21:50:12] <Fingolfin> in fact, on one of my SRs I finally got some explanation from them how it supposedly works now
[21:50:17] <Fingolfin> Here's how the project weekly stats are computed:
[21:50:18] <Fingolfin> log (3 * # of forum posts for that week) +
[21:50:18] <Fingolfin> log (4 * # of tasks ftw) +
[21:50:18] <Fingolfin> log (3 * # bugs ftw) +
[21:50:18] <Fingolfin> log (10 * patches ftw) +
[21:50:19] <Fingolfin> log (5 * tracker items ftw) +
[21:50:20] <Fingolfin> log (# commits to CVS ftw) +
[21:50:23] <Fingolfin> log (5 * # file releases ftw) +
[21:50:25] <Fingolfin> log (.3 * # downloads ftw)
[21:50:27] <Fingolfin> ----------------------------------
[21:50:29] <Fingolfin> take the result, and the activity is the percentile of that
[21:50:31] <Fingolfin> project (0-100%) against all other projects who had a
[21:50:33] <Fingolfin> non-zero result.
[21:50:42] <Fingolfin> so cvs commits have the loweset weight in total
[21:51:03] * Fingolfin considers not using CVS directly anymroe but instead submitting patches for evrything :)
[21:51:23] <Dominus> hmm
[21:51:25] <wjp> eh? log (3 * a) = log 3 + log a...
[21:51:43] <wjp> so none of those coefficients do anything, right?
[21:56:00] <wjp> or am I missing something?
[21:58:58] <Fingolfin> hm, right... actually, i somehow never read log as being the log function... lol
[21:59:04] <Dominus> Fingolfin: OSX beta status request on the forum
[22:00:33] <Fingolfin> oh wow SF web site is down
[22:00:39] <Dominus> yep
[22:01:39] <wjp> Fingolfin: for a few hours now
[22:01:45] <Fingolfin> cool
[22:12:51] * wjp curses F16... computing the ggd of two degree 6 polynomials in F16[X] is _not_ something you want to do manually
[22:13:23] * wjp hopes Maple can do something about it...
[22:31:29] <wjp> cool, Jeff fixed --buildmap according to the ChangeLog
[22:38:50] <Dominus> nice, extra for my doc changes I bet :-)
[22:40:41] <wjp> Fingolfin: now that depfiles don't have their own target, that sed command can probably go, right?
[22:41:14] <Fingolfin> probably, yeah, lemme look
[22:41:39] <wjp> also, is it possible to both generate dependencies and compile in one gcc call?
[22:41:53] <wjp> it seems a bit wasteful to preprocess the .cc file twice
[22:42:00] <Fingolfin> hmmmm
[22:42:43] <Fingolfin> let's try :)
[22:43:17] <Fingolfin> but I somewhat doubt it
[22:43:29] <wjp> automake seems to do it
[22:43:39] <Fingolfin> ah wait, i see
[22:43:52] <Fingolfin> -MMD
[22:44:18] * wjp nods
[22:44:56] <wjp> can't use -MG then, though
[22:45:25] <Fingolfin> yes you can
[22:45:31] <Fingolfin> $(CXX) -Wp,-MMD -Wp,-MG $(CXXFLAGS) -c -o $@ $<
[22:45:38] <wjp> "`-MG' is not
[22:45:38] <wjp> supported with `-MD' or `-MMD'."
[22:45:39] <Fingolfin> the problem is how to put the .d file into .debs, though
[22:45:47] <Fingolfin> hm, it works for me
[22:46:20] <wjp> oh... of course... it would be silly to ignore missing files while actually compiling
[22:46:36] <Fingolfin> yeah
[22:46:37] <Fingolfin> also
[22:46:49] <Fingolfin> I think automake does *not* do it in a single run
[22:47:05] <wjp> it looks like you can specify an output file name behind the -MMD option
[22:47:36] <wjp> $(CXX) -Wp,-MMD,.deps/$(*F).pp
[22:49:04] <Fingolfin> yeah
[22:49:30] <Fingolfin> this is sooo l33t
[22:49:32] <Fingolfin> =9
[22:49:41] <wjp> :-)
[22:50:03] * wjp tries to figure out what tr ' ' '\012' < .deps/$(*F).pp | sed -e 's/^\\$$//' -e '/^$$/ d' -e '/:$$/ d' -e 's/$$/ :/' is supposed to do...
[22:50:38] <wjp> that 'tr' command replaces spaces with endlines?
[22:52:00] <Fingolfin> seems so
[22:52:27] <Fingolfin> shall I commit the streamlined Makefile?
[22:52:35] <wjp> sure
[22:52:43] --> Dominik has joined #exult
[22:53:06] <-- Dominus has left IRC (Read error: 104 (Connection reset by peer))
[22:53:13] <wjp> is $$ anything for make?
[22:53:26] <wjp> oh, dollar sign
[22:56:34] <wjp> hmm, I don't get why that sed script produces correct output
[22:56:34] <-- Dominik has left IRC (Read error: 104 (Connection reset by peer))
[22:58:00] <wjp> in fact, I don't even get why it's necessary at all
[22:58:57] <wjp> judging by exult's .deps/*.P files, it's supposed to add a "dependency : " line for each dependency of the object file
[22:59:50] <wjp> oh... oops... I do get why it works
[23:00:34] <wjp> Fingolfin: any idea why it creates those extra rules?
[23:02:16] <Fingolfin> I can't see any difference between .P and .Po except that .Po has some additional linebreaks...
[23:02:28] <wjp> .Po?
[23:02:46] <Fingolfin> ugh
[23:03:13] <Fingolfin> I have here .P and .Po files in the top leve .deps folder; and .P and .Plo in the subfolders (hm, automake is pro-palestina? :)
[23:03:22] <Colourless> ok, it's impossible to actually get the 'centre ' of the u8 maps. turns out that all the black areas are still filled with globs. so instead I decided to attempt to center on the music egg if there is one, which there should be
[23:03:38] <Colourless> it works better than nothing
[23:03:38] <Fingolfin> Colourless: good idea
[23:03:52] <wjp> Fingolfin: I only have .P files
[23:03:55] <Fingolfin> wjp: hm, what extra rules do you mean then?
[23:04:06] <Fingolfin> wjp: maybe because I use AM 1.5 ? what version do you have?
[23:04:14] <wjp> hmm, dunno
[23:04:24] <wjp> 1.4.8, could be it then
[23:04:34] <Fingolfin> grrr
[23:04:49] * Fingolfin deleted files/.deps to test what happens...
[23:05:00] <wjp> Bad Things(tm) ?
[23:05:06] <Fingolfin> major bad
[23:05:11] <Fingolfin> I can't recreate it anymore it seems
[23:05:29] <wjp> and if you just mkdir .deps?
[23:06:07] <Fingolfin> it says it has no rule to buil the .Plo files
[23:06:12] <Fingolfin> not even make clean works anymore
[23:06:16] <wjp> ouch
[23:06:38] <wjp> is automake 1.4.x -> 1.5 a 'safe' upgrade?
[23:06:55] <Fingolfin> for me it was in 98% of all cases
[23:07:03] <wjp> good enough
[23:08:15] <wjp> your Makefile looks quite clean now
[23:09:30] <wjp> hm, mapedit/Makefile.am:49: invalid unused variable name: `shapetest_LDADD'
[23:11:19] <wjp> ok, automake 1.5 does things quite differently
[23:11:54] <wjp> could the .P files be leftovers from 1.4?
[23:13:13] <Fingolfin> maybe yeag
[23:13:15] <Fingolfin> yeah even
[23:13:22] <Fingolfin> me does a complete clean & rebuild
[23:13:44] <wjp> it seems to use a 'depcomp' script
[23:13:55] <Fingolfin> yes
[23:15:45] <wjp> hmm, the automake 1.5 Makefile seems a lot faster when nothing needs to be done
[23:18:39] <wjp> hmm, I wonder... now there's no longer any unix shell stuff in the makefile, could it also be used as a windows makefile?
[23:19:12] <wjp> would probably need some changes for libraries & sdl-config
[23:19:14] <Fingolfin> maybe
[23:19:18] <Fingolfin> yeah
[23:19:23] <Fingolfin> and for exe suffix
[23:19:46] <wjp> maybe we should have a configure.bat ;-)
[23:19:54] <Colourless> heh
[23:21:42] <Colourless> so, how does clean work if there is no unix stuff?
[23:21:56] <wjp> Colourless: there's only one music egg in each map?
[23:22:09] <Colourless> seems that way
[23:22:18] <wjp> then event 4 would probably be OnLoad?
[23:22:27] <Colourless> yeah probably
[23:24:42] --> Cless has joined #Exult
[23:25:16] <-- Colourless has left IRC (Killed (NickServ (Nickname Enforcement)))
[23:25:31] --- Cless is now known as Colourless
[23:25:34] --- ChanServ gives channel operator status to Colourless
[23:31:55] <Colourless> you know, 'display' has advanced a hell of a lot over the last 24 hours
[23:32:51] --> Dominus has joined #exult
[23:32:52] --- ChanServ gives channel operator status to Dominus
[23:33:14] <Dominus> hi again
[23:33:21] <Colourless> hi
[23:33:43] <Fingolfin> rehi
[23:33:49] <Dominus> now if you add the avatar npc with frames you are almost there :-)
[23:34:55] <Colourless> heh, just draw the avatar at the center of the screen :-)
[23:35:18] <Fingolfin> Dominus: what about heights?
[23:35:28] <Fingolfin> collision detetion? :)
[23:35:40] <Fingolfin> detection even
[23:35:40] <Dominus> don't be so nitpicky :-)
[23:36:16] <wjp> Fingolfin: were those suggestions?
[23:36:32] * wjp remembers something about suggesting == volunteering... :-P
[23:36:42] <Fingolfin> I was just asking questions, I am not that stupid =)
[23:36:57] <Colourless> so are we attempting to get pentagram finished and released before exult is finished :-)
[23:37:15] <Fingolfin> =9
[23:37:15] <Dominus> it seems that way
[23:37:18] <wjp> lol
[23:37:42] <wjp> if the sanity of their usecodes is any indication, U8 should be less tricky to finish
[23:37:59] <Fingolfin> well, briefly before pentagram is "finished", I see this new project emerging: a U9 mapviewer =)
[23:38:09] <wjp> hehe :-)
[23:38:38] <wjp> and someone will start deciphering its usecode... write a disassembler...
[23:40:33] <Colourless> i think the u9 'map viewer' has to be authentic to the original and be really buggy :-)
[23:41:07] <Fingolfin> =)
[23:41:19] <Fingolfin> no problem, I can do the bug part if you like
[23:43:14] <wjp> hmm, Dancer replaced my broken getline with an equally broken getline
[23:43:33] <Colourless> heh
[23:43:34] <wjp> (or at least one that doesn't do what we want it to do)
[23:44:42] <wjp> hmm... Dancer claims mine doesn't work on the last line
[23:45:05] <wjp> as long as the file is terminated by an \r or \n it should work I think
[23:45:17] <Colourless> where is this code?
[23:45:20] <wjp> expack.cc
[23:45:26] <wjp> look for "BROKEN" :-)
[23:45:36] --> Dominik has joined #exult
[23:45:46] <Fingolfin> Dance... haven't seen him in ages
[23:45:50] <Fingolfin> rehi
[23:45:53] <Dominik> my connection is really buggy today
[23:45:56] <Dominik> tonight
[23:46:22] <Colourless> well, couldn't just check for eof?
[23:46:41] <wjp> after the first file.get(c)? sure
[23:47:21] <-- Dominus has left IRC (capek.openprojects.net irc.openprojects.net)
[23:47:21] <-- matto has left IRC (capek.openprojects.net irc.openprojects.net)
[23:47:43] <Colourless> no, instead of doing (c != '\r' && c != '\n') the following could be done
[23:48:10] <Dominik> hehe, what I wrote after I was disconnected: the exult beta stage: Jeff works on ExultStudio and breaks and fixes stuff, the rest of the team idles or works on Pentagram instead. Exult stays Bug Free (TM)
[23:48:13] <Colourless> c >= ' ' || c == '\t' && file.good()
[23:48:42] <Colourless> let me rewrite that
[23:49:22] <Colourless> the while would be replaced with
[23:49:22] <Colourless> while (i < size-1 && ( c >= ' ' || c == '\t') && file.good())
[23:49:32] <Colourless> that is the first one
[23:49:58] * wjp nods
[23:50:15] <Colourless> the second one would want 'similar' changes
[23:52:17] <Colourless> actually using !file.eof() would be better no?
[23:52:22] <wjp> why?
[23:53:00] <Colourless> no real reason, except we are trying to check for eof, but good will do that anyway
[23:53:06] <wjp> while (!(file.peek()>=' ' || file.peek()=='\t') && file.good())
[23:53:06] <wjp> ?
[23:53:29] <Colourless> yeah that seems ok
[23:56:30] <Dominik> I would really like to know why the MS Office XP setup thinks it is neccessary to delete the exe of my the mS powerpoint viewer
[23:56:40] <Dominik> MS seems to be the keyword here
[23:59:58] --> matto has joined #exult