IRC logs for #opendune on OFTC at 2011-06-18
⏴ go to previous day
02:22:33 *** fjb is now known as Guest5054
06:48:55 *** Alberth has joined #openDune
06:48:55 *** ChanServ sets mode: +o Alberth
10:34:25 <Alberth> variable_39F2 looks buggy.
10:34:25 <Alberth> I have g_global->variable_39F2 = iconMap[iconMap[ICM_ICONGROUP_FOG_OF_WAR] + 16]; <-- FOW has 16 tiles, so this is one further
10:34:25 <Alberth> tile->overlaySpriteID = g_global->variable_39F2 & 0x7F; <-- euhm, wtf? :)
10:36:05 <Alberth> in reality 39F2 is the pre-built concrete slab sprite ID
10:49:55 <TrueBrain> happens a lot; funny aint it? (NOT)
11:29:43 <TrueBrain> bloomSpriteIDs <- how can 1 uint16 be plural?
11:30:27 <TrueBrain> t->overlaySpriteID = g_global->wallSpriteIDs & 0x7F; <- many of those masks are unneeded, as overlaySpriteID is only 7 bits long :D
11:32:00 <TrueBrain> and is the patch large? It is what it should do, in my opinion
11:39:13 <Alberth> 10 235 2 tile indexes: Spice bloom tiles. <-- there are 2 bloom sprites :)
11:39:46 <Alberth> same with wall and landscape
11:39:47 <TrueBrain> but can 1 uint16 point to two IDs? :)
11:39:56 <TrueBrain> I expected a pointer :)
11:40:00 <Alberth> it is the first of a sequence
11:41:01 <Alberth> it is a pointer, except in numeric form
11:41:37 <TrueBrain> just saying it looks odd :D
11:46:19 <Alberth> you are correct, it is a spriteID.
11:49:02 <Alberth> that also means that code like if (t->groundSpriteID == g_global->bloomSpriteIDs || t->groundSpriteID == g_global->bloomSpriteIDs + 1) return 14; is wrong; it should check the icongroup numbers, not the sprite numbers
12:06:46 <DorpsGek> SVN: alberth (r1809) -Add: Names for icon groups, renamed a few sprite related variables.
12:07:10 <TrueBrain> is that '14' then not the icongroup?
12:07:41 <Alberth> no, and it is also not a structure number. so nice :)
12:07:56 <TrueBrain> we have been wondering what does numbers are
12:07:59 <TrueBrain> they are repeated a few times
12:09:03 <Alberth> my guess is some abstracted tile-type, but a separate enum
12:12:54 <Alberth> (I got distracted from cleaning up the emu_ax uses by those numbers, and from there to the icon group numbers :) )
12:15:16 <TigerP> I it normal that when keeping the cursor at the edge of the screen it does not scroll?
12:15:23 <Xaroth|Work> Alberth: the viewer wasn't made to display HG patches :P
12:15:35 <TigerP> I thought it worked that way in DOS
12:15:49 <Alberth> TigerP: there is a setting in the game controls iirc
12:16:28 <Alberth> but I may be confused with another game
12:16:49 <TigerP> yes, auto scroll is off
12:17:20 <TigerP> but when put to on, it still does not work
12:33:39 <Alberth> hmm, trunk crashes :( opendune: src/driver.c:601: Drivers_All_Init: Assertion `Drivers_Music_Init(music) == music' failed.
12:36:12 <TigerP> it seems that the emu library can not handle the shift key
12:36:36 <TigerP> [EMU] ERROR: unhandled key 12F
12:36:45 <DorpsGek> SVN: alberth (r1810) -Fix (r1809): Comma at end of enumerator list caused a -pedantic warning.
12:37:32 <Alberth> TigerP: yes, number of supported keys is quite limited, probably since it is all temporary code
12:41:05 <Alberth> with auto-scroll on, if you move to the edge, and click, the display moves to the far side of the map
12:41:54 <glx> * @note This documentation was inspired by information obtained from 'Arrakis Research Company', http://www.junkyard.dk/ . <-- it's also at www.dune2.dk :)
12:43:00 <Alberth> ok, but junkyard.dk was in the html file I have here at the local file system :)
12:45:18 <TigerP> Alberth: that is the same for auto-scroll on and off
12:46:36 <Alberth> oh. I don't remember ever playing with autoscroll on, I guess I found it useless in DOS too :)
12:55:41 <Alberth> I solved the crash mystery (a non-svn version did not crash here). Apparently, lack of data files crashes on the music driver initialization assert :p
12:57:25 <Alberth> TigerP: you may want to file a bug report
13:16:52 <TigerP> but maybe I should test first how it works in the real version
15:24:12 *** fjb|mobile has joined #openDune
16:05:46 <Alberth> glx: do you use the GENERATE_HTMLHELP of doxygen?
16:07:12 <Alberth> good, then I can disable it in favor of two other settings
18:03:52 <DorpsGek> SVN: alberth (r1811) -Add: Improve doxygen support.
18:04:22 <Alberth> sorry, I changed all files :p
18:04:53 <TrueBrain> path in @file? Well, if it works :P
18:07:12 <TrueBrain> going to see how difficult it is to remove csip from WSA stuff
18:07:15 <TrueBrain> just to start somewhere
18:07:30 <Alberth> I hope it works for you too, I used doxygen 1.7.4, fairly new
18:08:06 <Alberth> and else we find out it will not be simple ;)
18:08:40 <TrueBrain> can just move them up one layer I guess
18:39:19 <TrueBrain> minor image distortion
18:47:06 <TrueBrain> also, what do you think?
18:50:49 <Alberth> sorry, I was busy chatting with my mother :)
18:51:19 <TrueBrain> I just posted it 3 minutes ago :P
18:59:32 <TrueBrain> I can't find the bug ... hmm ..
19:00:53 <Alberth> you are doing too many things at the same time, imho
19:01:05 <TrueBrain> some renaming that does nothing
19:02:03 <Alberth> you also merge calls, which is not needed for the change itself
19:03:18 <Alberth> first hunk of mentat.c, you eliminate the memBlock variable by merging both function calls
19:03:40 <TrueBrain> yeah, of course. About time tbh :P
19:04:16 <Alberth> sure, the step in itself is good, but it messes up the view of what you are really changing
19:04:31 <TrueBrain> it changed anyway, as it required an emu_get_memorycsip
19:04:35 <TrueBrain> so that wouldn't have helped
19:04:46 <TrueBrain> hmm .. I am afraid WSAHeader is now too big or something
19:05:02 * TrueBrain replaced more 33 for WSAHeader
19:06:31 <Alberth> char filename[13]; <-- that will be padded, most likely
19:26:15 <TrueBrain> I know it has to be something simple
19:26:31 <TrueBrain> guess I just revert and try it again ... I now know small stuff to clean up first
19:28:47 <TrueBrain> just hope it has nothing to do with larger buffers etc
19:30:31 <TrueBrain> think WSAs have their palette embedded
19:38:46 <DorpsGek> SVN: truebrain (r1812) -Fix: split out the Format40 functions to their own file
19:43:44 <Alberth> Map_B4CD_0750 seems to return the type of landscape.
19:44:22 <Alberth> g_global->variable_24B8[] contains the following values: 0=normal sand, 1=partial rock, 5=mostly rock, 4=entirely rock, 3=sand dunes, 2=full sand dunes, 7=partial mountain, 6=full mountain, 8=spice, 9=thick spice
19:46:50 <glx> Alberth: that's what it does yes :)
19:47:33 <DorpsGek> SVN: truebrain (r1813) -Fix: some cleanup for WSA
19:48:05 <Alberth> have a nice evening :)
20:00:07 <DorpsGek> SVN: alberth (r1814) -Add: Remove global variable_24B8.
21:10:20 <DorpsGek> SVN: alberth (r1815) -Codechange: Rename Map_B4CD_0750 to Map_GetLandscapeType, and introduce an enum for its return values.
21:10:55 <Alberth> gives nice insights in other unknown functions :)
21:11:07 <Alberth> but not today any more; good night
21:13:26 <Alberth> it's a really nice 'smiley' :)
21:13:46 <Xaroth|Work> more a wave really, o7 is a salute
21:14:16 <Xaroth|Work> ,.|.. is a middle finger
22:21:33 <TrueBrain> right, I was right, the WSAHeader size messes stuff up
22:28:25 <TrueBrain> LOL! WSAHeader length was assumed in header of the WSA
22:29:02 <DorpsGek> SVN: truebrain (r1816) -Fix: move WSAHeader to wsa.c (its locally required only) and make the code independant of the length of WSAHeader
22:52:12 <glx> removed csip32 and it works ?
22:55:50 <TrueBrain> passed the initial what I did
22:56:33 <DorpsGek> SVN: truebrain (r1817) -Add: C-ified all emu_ stuff out of the WSA. It is now (almost) completely without emu_
22:56:38 <TrueBrain> some minor stuff left
22:57:04 <DorpsGek> SVN: truebrain (r1818) -Fix: save before commit etc
23:00:24 <TrueBrain> I am happy it works
23:00:27 <TrueBrain> think I can remove Tools_MAlloc
23:00:30 <TrueBrain> but will try that later :D
23:03:09 <glx> tools_malloc is still used for music or voices
23:03:35 <TrueBrain> I meant for the WSA ;)
continue to next day ⏵