IRC logs for #opendune on OFTC at 2011-06-04
            
00:26:42 *** fjb has quit IRC
00:43:09 *** glx has quit IRC
06:11:15 *** TinoDid|znc has joined #openDune
06:13:51 *** TinoDidriksen has quit IRC
06:13:51 *** TinoDid|znc is now known as TinoDidriksen
07:40:55 *** fjb has joined #openDune
07:47:50 *** fjb has quit IRC
07:59:01 *** fjb has joined #openDune
08:13:29 *** Alberth has joined #openDune
11:21:52 * Alberth needs a 80186 wizard :p http://www.fpaste.org/2Y48/ <-- with cs == s_input_local normally, and ds == g_global normally, but the push/pop mess that up, it seems, so what happens here?
11:25:15 <TrueBrain> ds = cs
11:25:40 <TrueBrain> it jsut doesn't make sense at all
11:26:21 <TrueBrain> had that when I converted too :D
11:27:20 <Alberth> in other words, a broken compiler?
11:27:38 <TrueBrain> or code :D
11:27:39 <TrueBrain> but yes
11:27:44 <TrueBrain> it reads some weird offseted
11:27:58 <TrueBrain> 29E8:700E
11:28:15 <TrueBrain> so ... 30E8:E or something?
11:31:03 <Alberth> disassembly says 29E8:700E is in the middle of some data, 01b3 is in a bunch of 0 bytes
11:33:29 <Alberth> l__09E2: <-- there it also messes with ds, but in the context of DOS interrupts
11:37:00 <Alberth> emu_addw(&emu_bx, 0xA9C);
11:37:00 <Alberth> emu_ax = emu_get_memory16(emu_ds, emu_bx, 0x0); <-- at line 1173
11:37:34 <TrueBrain> pretty normal
11:37:40 <TrueBrain> but yeah, check DOS interrupts what they change
11:40:04 <Alberth> hmm, a9c could be , it is unknown data
11:49:21 <Alberth> mouse code seems no to do such things
12:16:33 *** glx has joined #openDune
12:16:33 *** ChanServ sets mode: +o glx
12:16:59 <glx> hello
12:34:12 <Alberth> hello
12:34:40 <Alberth> such fun, cs__29e8 :p
13:10:06 *** fjb|mobil has joined #openDune
13:15:24 *** fjb|mobil has quit IRC
13:16:51 <Alberth> http://devs.openttd.org/~alberth/01b_fix_segments.patch <-- let's use cs when it accesses that segment :p
13:27:40 *** fjb|mobile has joined #openDune
13:30:51 <glx> hmm this stuff is weird
13:39:17 <glx> and it fells very wrong
13:40:29 <Alberth> yep :)
13:40:59 <glx> because ds should be 353F there
13:45:05 <Alberth> hmm, two of them don't do anything? lines1540 and 1550
13:45:42 <Alberth> near eeb
13:46:08 <Alberth> pull weird value, do xor, push, pop
13:47:01 <Alberth> somebody was very drunk here :p
13:49:30 <Alberth> btw I checked the value ds there by printing its value
13:49:49 <Alberth> (except one case)
13:53:56 <TrueBrain> yeah, I remember clearly battling with this
13:54:03 <TrueBrain> if you fix it, what you suggest it would mean, things break
13:54:11 <TrueBrain> it writes a random pieces of memory, and it 'works'
13:56:05 <Alberth> timing?
13:58:31 <TrueBrain> no, clearly the code it overwrites does no harm
13:58:37 <TrueBrain> but ... that is more luck than anything
13:59:12 <glx> like the missing 0xFFFE check in load_voices
14:01:43 <TrueBrain> owh, these are only reads
14:01:44 <TrueBrain> so no harm is done
14:01:48 <TrueBrain> just weird stuff is read :D
14:04:09 <Alberth> (15:54:50) TrueBrain: if you fix it, what you suggest it would mean, things break <-- fix in the sense reading from g_global?
14:04:15 <TrueBrain> yes
14:04:34 <TrueBrain> as now it reads read-only values :p
14:05:37 <Alberth> causing the cache with the previous value to be invalidated or so
14:08:20 <TrueBrain> and it builds on that from that point on :)
14:10:43 <glx> hmm something looks wrong in GameLoop_B4AE_0000()
14:10:55 <glx> atreides case for campaign 8
14:11:31 <glx> args are not in the same order as the other cases
14:11:43 * glx checks his conversion
14:13:01 <TrueBrain> might be intended?
14:13:09 <glx> but it's weird
14:13:14 <TrueBrain> shows the cutscenes ..
14:13:16 <TrueBrain> try it out :)
14:14:04 <glx> all are struct_19A8, struct_19F0, struct_1A2C, uint16 except this one
14:14:28 <glx> it does the opposite
14:17:52 <TrueBrain> so conversion error, or long standing bug?
14:19:11 <glx> conversion error
14:19:27 <TrueBrain> :(
14:23:05 <glx> probably nobody finished A8 since the conversion
14:23:18 <TrueBrain> :P
14:29:37 <DorpsGek> SVN: glx (r1754) -Fix (r1384): conversion error
15:54:28 *** fjb|mobile has quit IRC
16:08:59 *** fjb|mobile has joined #openDune
16:54:23 <DorpsGek> SVN: glx (r1755) -Add: C-ified MPU_ApplyVolume()
17:00:56 <Alberth> wow, the emu_Mouse_HandleMovementIfMoved() is making sense!!! :p
17:01:08 <glx> nice
17:17:22 * Alberth does not understand the compiler: emu_cl = s_input_local->activeInputMap[emu_di]; emu_andb(&emu_cl, emu_bl); then do nothing with cl or cx?
17:18:33 <Alberth> &emu_bl, emu_cl would make sense
17:57:11 *** fjb|mobile has quit IRC
18:01:33 *** fjb|mobile has joined #openDune
18:02:35 <Alberth> fjb|mobile: connection stability can be improved :)
18:03:21 <fjb> Mobile deveces power consumption can improve. :-)
18:25:02 <Alberth> is uint8 idx = ...; alright? or should it be uint8 idx; idx = ....; ?
18:55:18 <glx> we use both
19:12:34 <Alberth> ok, thanks
19:30:26 *** fjb has quit IRC
19:46:42 *** fjb has joined #openDune
19:48:30 <Alberth> good night
19:49:38 *** Alberth has left #openDune