IRC logs for #opendune on OFTC at 2011-08-02
⏴ go to previous day
00:00:25 <DorpsGek> SVN: truebrain (r2189) -Codechange: mark the DOS related variables in g_global as removed, as they got removed long ago
00:02:29 <DorpsGek> SVN: truebrain (r2190) -Codechange: some more fixing g_global
00:05:21 <DorpsGek> SVN: truebrain (r2191) -Add: removed a timer from g_global
00:12:03 <DorpsGek> SVN: truebrain (r2192) -Add: one more timer, and some cleanup (yes, more :p)
00:12:15 <TrueBrain> I am walking the g_global from top to bottom atm :p
00:12:18 <TrueBrain> but I am getting bored
00:14:19 <DorpsGek> SVN: truebrain (r2193) -Codechange: some more variables were long not used
00:18:31 <DorpsGek> SVN: truebrain (r2194) -Add: use strings directly; no reason to point to the 16bit world to get them
00:19:23 <TrueBrain> 175 variables left in g_global
00:21:09 <DorpsGek> SVN: truebrain (r2195) -Codechange: the drawing pointers are long unused
00:23:09 <TrueBrain> yes, I love my stats :D
00:25:54 <TrueBrain> time to get some sleep I guess
00:26:13 <TrueBrain> not much more I can do for OpenDUNE :D
00:26:39 <TrueBrain> (that is, I am too bored to do more g_globals :p)
00:37:56 <DorpsGek> SVN: truebrain (r2196) -Add: some more g_global cleaning, another timer, read-only, write-only, you name it, it is all in here.
00:38:02 <TrueBrain> picked out the easy ones :D
00:38:10 <TrueBrain> right; time to go beddybed
00:42:22 <DorpsGek> SVN: truebrain (r2197) -Codechange: the last two csips marked as active in g_global are in fact no longer used. No more active csips in g_global!!
00:42:25 <TrueBrain> well, a good-night present :D
02:44:35 <DorpsGek> SVN: glx (r2198) -Codechange: some cleanup in MSData (removed write-only, ...)
08:38:29 *** Alberth has joined #openDune
08:38:29 *** ChanServ sets mode: +o Alberth
09:48:01 <Alberth> am busy in the input data to extract the various tables
09:48:26 <TrueBrain> I did mark the readonly variables yesterday, as I got annoyed that I didn't know what were tables :D
09:49:11 <TrueBrain> not really :) Just hope you updated :P :D
09:49:18 <TrueBrain> I would hate to give you conflicts :(
09:49:34 <Alberth> yeah I saw your cleanup actions, very good work
09:50:15 <Alberth> although I would not have bothered with unifying the removed variable names :)
09:50:34 <TrueBrain> I was annoyed my script didn't work :D
09:50:39 <TrueBrain> and wanted to know how much was really left
09:51:00 <TrueBrain> while doing I did realise it was kinda pointless, but meh :)
09:51:12 <Alberth> it was too late already :)
09:52:00 <Alberth> but it is all really moving very fast
09:52:26 <TrueBrain> I am really surprised by it tbh
09:59:06 <Alberth> what do you expect at 40+ commits / day :p
10:33:18 <TrueBrain> guess I will have to wait for glx :)
10:49:45 <TrueBrain> ugh, I should stop working on that patch :D
10:49:47 <TrueBrain> it grows too big :p
10:49:53 <TrueBrain> with changes that should in different commits tbfh :p
11:02:08 * Alberth wonders what happened to keymap_capslock[62] and keymap_caps_shift[62]
11:02:23 <Alberth> they are noted as 'read-only', but I cannot find the references
11:11:39 <TrueBrain> I resolved them once
11:11:43 <TrueBrain> but we never implemented capslock
11:11:49 <TrueBrain> (like most special keys)
11:11:53 <TrueBrain> same goes for numlock for example
11:12:02 <Alberth> Program terminated with signal 11, Segmentation fault. <-- your patch
11:13:13 <TrueBrain> anyway, I resolved the keymaps as I thought it might be nice
11:13:16 <TrueBrain> not sure if they are useful at all
11:13:35 <Alberth> src/mt32mpu.c:630:2: warning: overflow in implicit constant conversion [-Woverflow] <-- and I have this warning
11:14:16 <TrueBrain> do you have the latest version of my patch?
11:14:20 <TrueBrain> 630 doesn't really add up
11:15:47 <Alberth> for (i = 0; i < 144; i++) data->variable_00B8[i] = 0xFF;
11:15:51 <TrueBrain> one of the downsides of reusing the same url
11:16:12 <TrueBrain> yeah, that was glx' work ... so I fixed that :D
11:16:16 <TrueBrain> but yeah, reload my patch plz :)
11:20:08 <TrueBrain> I just wish there was an easier way to push local patches
11:20:24 <TrueBrain> I stopped naming them for what I was doing, and add _1 etc to it, as for OpenTTD at one point I had 1000+ patches
11:20:28 <TrueBrain> it doesn't really work :P
11:21:27 <TrueBrain> what is 'index' in that frame?
11:21:39 <Alberth> I make patch queues, and upload them in separate directories, and remove those dirs after commit :)
11:22:07 <TrueBrain> that is a bit wrong :D
11:22:29 <TrueBrain> of course it is wrong, lolz
11:22:32 <TrueBrain> how could this ever work ...
11:23:24 <TrueBrain> new patch, or change line 402 to = -1
11:24:09 <DorpsGek> SVN: alberth (r2199) -Add: Removed keymaps from input segment.
11:25:34 <TrueBrain> can you please keep variables etc at the top of the file? It is kinda hard to read in the middle of functions :D (at least, my IDE doesn't show it in any pretty way :p)
11:26:07 <TrueBrain> "in the middle between function", grr @ english :P
11:26:36 <Alberth> me has speech from the intro
11:27:00 <TrueBrain> I believe the MPU is everything except voices :D
11:29:54 <Alberth> some part of the story is not told at all or too late
11:30:15 <TrueBrain> I believe glx also complained about that
11:30:27 <Alberth> ie the atreides aircraft scense has 2 airplanes normally, and now only one
11:31:02 <Alberth> the 'noble atreides text' was spoken while the scene of ordos was shown
11:31:23 <Alberth> the explosion at the end of the harkonnen scene was never played
11:31:24 <TrueBrain> that means my tables are wrong I guess
11:31:33 <TrueBrain> it took so much time to do that :(
11:31:41 <Alberth> probably timing is off
11:32:16 <Alberth> but sound in game is working, it just seems slower than it already was
11:32:32 <TrueBrain> but I see the intro on windows, with sounds
11:32:43 <Alberth> isn't that always the case with windows? :p
11:32:56 <TrueBrain> last attempt MSVC compile refused to run
11:33:00 <TrueBrain> it kept crashing in some weird ways
11:33:08 <TrueBrain> looking back it was a different DSP codesegment
11:33:18 <TrueBrain> the voice of the intro still makes me laugh
11:33:23 <TrueBrain> it is such a wrong voice for it
11:33:34 <TrueBrain> it misses ... depth
11:33:45 <Alberth> Assertion 's' failed at pulse/stream.c:1301, function pa_stream_write(). Aborting.
11:33:45 <Alberth> zsh: abort (core dumped) ./opendune <-- not entirely working thus
11:34:04 <TrueBrain> does it also happen without my patch?
11:35:41 <TrueBrain> well, I guess I can now test myself too :D:D:D
11:37:00 <TrueBrain> yeah, shit is very off
11:40:25 <Alberth> (same fatal assert without your patch)
11:40:32 <TrueBrain> btw, they use an inverted struct, really weird ... instead of struct[10], it is [10][<sizeof struct>]
11:40:42 <TrueBrain> k, so that is for glx to solve :D Hihi :)
11:41:01 <TrueBrain> doesn't happen on windows, so most likely some SDL problem
11:45:44 <DorpsGek> SVN: alberth (r2200) -Codechange: Move keymap tables to top of file for broken IDE.
11:46:22 <TrueBrain> I really dislike searching for variables :P
11:48:12 <TrueBrain> I can't detect any issue, bah
11:48:27 <TrueBrain> the only thing that feels wrong are the static tables
11:55:35 <TrueBrain> hmm .. it seems just a bit random when it works or not
11:57:01 <TrueBrain> yup, now it works fine
11:57:07 <TrueBrain> just a bit unstable :)
11:57:39 <TrueBrain> owh, Alberth, what might help for you
11:57:46 <TrueBrain> to download the PAK updates of MrFilbble
11:58:30 <Alberth> and that helps in what way?
11:58:38 <TrueBrain> the VOC are known to be broken
11:58:44 <TrueBrain> so it is possible that pulse crashes over it
11:58:53 <DorpsGek> SVN: truebrain (r2201) -Add: move most of the mt32mpu variables to static local variables, and name some where possible
12:00:05 <TrueBrain> so it wouldn't surprise me if that fixes it; if so, it also explains why DOSBox plays Dune2 on windows just fine
12:00:18 <TrueBrain> as Windows clearly has no issue what-so-ever with the broken VOC files
12:03:57 <Alberth> but it never crashed before
12:04:16 <TrueBrain> it also never worked correctly before :D Hihi :)
12:04:31 <TrueBrain> I just wonder if it helps
12:04:37 <TrueBrain> what would be possible, is this:
12:04:44 <TrueBrain> normally an interrupt is called when the buffer runs out
12:04:47 <TrueBrain> this is now removed
12:04:51 <TrueBrain> as the buffer is a predefined size
12:04:58 <TrueBrain> it might be that at that place in the VOC is some special code
12:05:03 <TrueBrain> like 2 length bytes or what-ever
12:05:09 <TrueBrain> that would not be totally misread
12:05:29 <TrueBrain> of course this makes the VOCs broken, as they are not pure VOCs, but would work before the buffer interrupt was removed
12:05:40 <TrueBrain> but if that is true, MrFlibbles PAK should fix your crash :)
12:05:50 <TrueBrain> and if so, it is worth further investigation
12:06:09 <Alberth> haha, lets translate | to | :D
12:06:47 <DorpsGek> SVN: truebrain (r2202) -Add: we no longer care about IRQ, DMA and Port, and name DriverInfo fields
12:07:28 <Alberth> translateMap[14] == '|' so the character '|' is replaced by translateTo[14], which is..... '|' :D
12:09:45 <Alberth> perhaps there are other keyboard layouts where it matters
12:13:22 <Alberth> do we have compile time assertions?
12:13:39 <TrueBrain> works on ... most compilers
12:15:05 <DorpsGek> SVN: truebrain (r2203) -Add: extracted DSP and MPU DriverInfo
12:16:37 <DorpsGek> SVN: truebrain (r2204) -Add: stripped DriverInfo from unused fields (or useless fields)
12:16:45 <TrueBrain> I hope you don't mind, but I ahve been busy in the drivers :P
12:17:09 <glx> should be ok as I had no active patch :)
12:18:08 <DorpsGek> SVN: truebrain (r2205) -Codechange: byebye codesegment hacks
12:18:44 <TrueBrain> I removed the emu_dsp_cs and emu_mtu32_cs thingies :)
12:18:52 <TrueBrain> found out some nice things, the dirver is kinda nice :)
12:18:59 <TrueBrain> and! MSVC finally gives me a working binary here :D
12:19:44 <TrueBrain> list of TODO is getting REALLY small now :D
12:20:35 <DorpsGek> SVN: alberth (r2206) -Add: Move key translate maps out of input segment.
12:20:51 <TrueBrain> 6 variables left Alberth :D
12:21:24 <TrueBrain> glx: you btw made a buffer overflow which GCC warned about and MSVC did not; made me laugh :D
12:21:39 <TrueBrain> MSData->variable_00B8 is a double array, and you initialized it as such
12:21:44 <TrueBrain> but had it defined as 9 variables
12:21:51 <TrueBrain> for some reason, MSVC was perfectly fine with that :p
12:22:51 <TrueBrain> it is an inverted struct
12:23:07 <glx> I have the original source ;)
12:23:29 <TrueBrain> its a struct like: struct { a[16]; b[16]; c[16]; }
12:23:36 <TrueBrain> instead of struct { a; b; c; }[16
12:23:43 <TrueBrain> really odd, if you ask me :p
12:24:06 <glx> ctrl_log STRUC ;XMIDI sequence/global controller log
12:24:06 <glx> MODUL db NUM_CHANS dup (?)
12:24:09 <glx> C_LOCK db NUM_CHANS dup (?)
12:24:09 <glx> C_PROT db NUM_CHANS dup (?)
12:24:11 <glx> V_PROT db NUM_CHANS dup (?)
12:24:33 <TrueBrain> it is an inverted struct if you ask me :p
12:24:59 <DorpsGek> SVN: truebrain (r2207) -Codechange: dsp.c no longer depends on libemu
12:25:18 <TrueBrain> I will now removed the last bits of csip file access in mt32mpu.c
12:25:25 <TrueBrain> then I will remove the csip access in Driver
12:25:33 <TrueBrain> and then we are nearly done :)
12:25:50 <TrueBrain> but just so you know that I will be doing minor stuff in the driver for a bit, and then I am going to run away as far as possible from it :D
12:26:20 <glx> I have to work on the static variables in mt32mpu
12:30:01 <TrueBrain> others I had no clue about :p
12:41:24 <TrueBrain> this is weird ... according to this, music->content is a NULL csip pointer
12:41:55 <TrueBrain> meaning it tries to read a music file from 0x0 :P
12:42:30 <Alberth> 'reference on original functionality' ?
12:43:14 <TrueBrain> keep it for reference
12:43:19 <TrueBrain> I mean, we have capslock too, not?
12:43:43 <TrueBrain> subversion will tell us what it was when we need it :)
12:43:55 <TrueBrain> I just wanted to keep it to find out what the funcionality of keys wer ein Dune2
12:44:00 <TrueBrain> we haven't uncovered all, I am sure
12:44:08 <TrueBrain> but .. I don't really care tbh :D
12:45:48 <DorpsGek> SVN: alberth (r2208) -Add: Remove flags from input segment.
12:52:39 <DorpsGek> SVN: truebrain (r2209) -Add: analyze a file via real pointers
12:53:05 <TrueBrain> oeh, and it still works :P
12:54:12 <TrueBrain> had to introduce 2 GetCSIPs
12:54:18 <TrueBrain> but I am going to remove those now :)
12:57:39 <TrueBrain> TIMB is read, but never used?
12:59:08 <DorpsGek> SVN: truebrain (r2210) -Add: remove the final csip from mt32mpu
12:59:33 <DorpsGek> SVN: truebrain (r2211) -Add: mt32mpu.c no longer depends on libemu
13:00:13 <TrueBrain> I am totally done with mt32mpu.c btw
13:01:59 <glx> I'll remove TIMB and RNBR ;)
13:03:22 <TrueBrain> what is it used for?
13:03:32 <TrueBrain> other type of music files?
13:04:42 <glx> TIMB is never used in AIL
13:06:35 <DorpsGek> SVN: alberth (r2212) -Add: Removed mouse read buffer 063B from input segement.
13:09:50 <DorpsGek> SVN: truebrain (r2213) -Add: move sound/music/voice driver out of g_global
13:11:05 <DorpsGek> SVN: truebrain (r2214) -Add: the extension field from DriverInfo is unused
13:11:47 <DorpsGek> SVN: truebrain (r2215) -Codechange: look mom, I can set values and never read them again
13:13:32 <DorpsGek> SVN: truebrain (r2216) -Codechange (r2215): history repeating itself; twice!
13:13:38 <TrueBrain> +1 for cryptic commit messages :P
13:18:58 <Xaroth|Work> you can't +1 yourself :P
13:19:01 <Xaroth|Work> so I'll +1 for you :P
13:19:59 <DorpsGek> SVN: truebrain (r2217) -Add: moved out a few buffers from g_global
13:24:43 <DorpsGek> SVN: truebrain (r2218) -Add: some more driver related variables moving out of g_global
13:25:16 <DorpsGek> SVN: truebrain (r2219) -Add: one g_global variable felt left alone. Well, he was useless anyway
13:26:12 <DorpsGek> SVN: truebrain (r2220) -Codechange: MSBuffer no longer has to be packed
13:26:38 <TrueBrain> right, now the csip32s are exposed
13:33:37 <DorpsGek> SVN: truebrain (r2221) -Add: exchanged two csip32s for char *
13:39:21 <DorpsGek> SVN: alberth (r2222) -Add: Removed 0036 and 0058 tables from input segment.
13:46:20 <DorpsGek> SVN: alberth (r2223) -Add: Removed 01B9 table from input segment.
13:52:21 <glx> src/mt32mpu.c:231:2: warning: comparison is always true due to limited range of data type
13:52:21 <glx> if (s_mpu_controls[chan].volume != 0xFF) MPU_Send(0xB0 | chan, 7, s_mpu_controls[chan].volume);
13:55:05 <DorpsGek> SVN: alberth (r2224) -Add: Removed activeInputMap from input segment.
13:56:20 <glx> for me it should be -1 :)
13:57:26 <Alberth> C doesn't speak 2-complement numbers :)
14:00:13 <TrueBrain> I don't like that one
14:00:38 <Alberth> last chance for s_input_local :p
14:01:55 <DorpsGek> SVN: alberth (r2225) -Add: Remove s_input_local.
14:02:58 <Alberth> one emu_get_memory8 less :p
14:14:50 <DorpsGek> SVN: truebrain (r2226) -Add: some more work in removing csip32 from driver.c
14:15:28 <TrueBrain> it can be more pretty! :P
14:16:29 <TrueBrain> fixed Drivers_CallFunction in the most ugly way possible :D:D:D Haha :)
14:20:43 <DorpsGek> SVN: truebrain (r2227) -Add: move one more csip32 to 32bit/64bit world, and don't use custom code for what we have wrappers
14:22:25 <TrueBrain> Alberth: did you test the VOC pak stuff?
14:22:43 <DorpsGek> SVN: truebrain (r2228) -Add: one more csip32 is now gone
14:22:46 <TrueBrain> k k, was wondeirng :)
14:23:27 <DorpsGek> SVN: truebrain (r2229) -Codechange: Tools_Malloc(), I thank you, and I wave you goodbye.
14:24:20 <DorpsGek> SVN: truebrain (r2230) -Codechange: emu_Global_GetCSIP(), I never liked you, and with pride I can say: I no longer need you.
14:24:25 <TrueBrain> right, almost there
14:24:35 <TrueBrain> some static information left ...
14:25:07 <DorpsGek> SVN: glx (r2231) -Add: some more work on MPU static variables
14:25:25 <glx> a little easier to read :)
14:25:41 <TrueBrain> new space policy? :D
14:27:14 <DorpsGek> SVN: glx (r2232) -Fix (r2231): some unwanted spaces
14:29:01 <TrueBrain> do you know what thid DIGPAK is?
14:29:11 * Alberth does colourBorderSchema
14:30:16 <DorpsGek> SVN: truebrain (r2233) -Codechange: we don't use DIGPAK, so remove references to it
14:35:33 <TrueBrain> we only have 1 music+sound driver, right? They are shared etc?
14:35:36 <TrueBrain> and that won't change, right?
14:35:53 <glx> no reason for that to change
14:36:13 <TrueBrain> how do you want to call it
14:41:28 <glx> I think we could remove Driver->dcontent first :)
14:41:39 <TrueBrain> that was when I noticed
14:41:51 <TrueBrain> not as trivial as it sounds btw
14:42:27 <TrueBrain> problem is CallFunction
14:44:28 <TrueBrain> first the index, second the offset
14:44:33 <TrueBrain> but I have an idea .. hackish, but it might work
14:44:50 <TrueBrain> but I am doing it, so pfff :p
14:45:48 <TrueBrain> well, I am testing it, is a better way of putting it :p
14:47:22 <DorpsGek> SVN: alberth (r2234) -Add: Moved colour border schema variables out of global.
14:50:33 <Xaroth|Work> this is going like crazy
14:51:05 <TrueBrain> meant mostly as 'temporary' solution
14:51:09 <TrueBrain> to get a bit of idea what is going on really
14:53:01 <TrueBrain> oeh, Unit() was wrong, fixed now :D
14:54:32 <glx> bool contentMalloced; /*!< Wether content pointer is the result of a malloc. */
14:54:32 <glx> + uint8 type; /*!< Type of driver. 0 = DSP, 1 = MPU. */
14:56:30 <Alberth> TrueBrain: dune2_speech_fix.rar <-- seems to prevent crash
14:56:39 <TrueBrain> Alberth: so I was right :D:D:D
14:56:59 <TrueBrain> glx: I have an hypotises: when using normal VOC, pulseaudio segfaults
14:57:06 <TrueBrain> Windows runs, as you hav enoticed
14:57:09 <TrueBrain> we know the VOCs are damaged
14:57:17 <TrueBrain> now pulse did work with interrupt
14:57:29 <TrueBrain> so now what if the VOC has some magic byte after the buffer length?
14:57:57 <TrueBrain> that at the exact position normally the buffer runs out, some magic byte is, which is skipped/
14:58:07 <TrueBrain> you worked on the code, hence the question, but is that anyhow possible?
15:02:47 <TrueBrain> removed the whole shissle
15:03:02 <glx> strcpy(driver->extension, "xmi"); <-- useless
15:03:12 <TrueBrain> we know music/sound is MPU, so let his init init MPU :)
15:03:45 <TrueBrain> glx: but did any of that VOC and buffer bla sound familair with what you read in the code?
15:03:50 <TrueBrain> or is it just bullshit?
15:03:52 <glx> + if (!Drivers_Init(sound, (char *)emu_get_memorycsip(driver->extension))) return 0; <-- why not "XMI" ?
15:04:07 <glx> or C55 (can't remember which is used
15:04:27 <TrueBrain> not everything in one patch :D
15:04:33 <TrueBrain> that is for another one
15:05:43 <TrueBrain> k k, will look into that later
15:05:56 <TrueBrain> first libemu :P I am going too fast for my own good :P
15:06:15 <glx> we read a block, then send it to the "callback"
15:06:31 <glx> when it's finished we get parse the next block
15:06:40 <DorpsGek> SVN: truebrain (r2235) -Add: refactored driver.c a bit, to get ride of some nasty API-based calls
15:07:15 <TrueBrain> what did I do wrong
15:07:48 <TrueBrain> it worked when I applied the patch
15:08:48 <TrueBrain> it does open the audio device here
15:08:51 <TrueBrain> just nothing comes out of my boxes
15:10:05 <glx> there's one nasty thing with voices, they use different sample rate so SDL_Audio is open/closed a lot
15:10:39 <TrueBrain> my last crash did hang
15:11:04 <TrueBrain> when you close it during voice playback
15:11:57 <TrueBrain> it might not have closed the driver correctly?
15:12:27 <TrueBrain> if it works for you
15:12:29 <TrueBrain> that is good enough for me :D
15:13:40 <glx> I think I know why it hangs
15:14:03 <DorpsGek> SVN: truebrain (r2236) -Add: long ago we left the possibility to pick your music/voice/sound driver
15:14:52 <DorpsGek> SVN: truebrain (r2237) -Codechange: wave goodbye to emu_get_memorycsip()
15:15:06 <glx> hmm no on SDL_Close() the right call is done
15:15:20 <TrueBrain> can you please help by testing if you can remove emu_flags.inf in drivers.c?
15:15:23 <TrueBrain> just call MPU_Interrupt?
15:15:28 <TrueBrain> see if everything is still in working order?
15:16:15 <DorpsGek> SVN: truebrain (r2238) -Add: remove the last of csip32
15:17:26 <TrueBrain> g_global and drivers.c :D
15:17:32 <TrueBrain> I can almost touch it! :P
15:17:35 <glx> seems to work without emu_flags
15:17:55 <TrueBrain> I don't think it changes :P
15:18:02 <TrueBrain> the code that can flip that switch has been removed :)
15:18:45 <DorpsGek> SVN: truebrain (r2239) -Change: emu_flags.inf no longer has any effect, so call the MPU_Interrupt() directly
15:18:53 <glx> hanged in voices it seems, so yes not related
15:19:03 <TrueBrain> and now I am off for a BBQ
16:43:18 <DorpsGek> SVN: glx (r2240) -Fix (r1517): conversion error and some cleanup
17:01:18 <Alberth> *** glibc detected *** /home/hat/odune/blah/trunk/opendune: double free or corruption (top): 0x0000000000867080 ***
17:03:54 <DorpsGek> SVN: glx (r2241) -Add: removed some drivers related variables that are duplicates of game config variables
17:06:03 <Xaroth|Work> only 8 days to go \o/
17:08:24 <glx> I guess it's the second free in Driver_UnloadFile
17:11:55 <DorpsGek> SVN: alberth (r2242) -Add: Move selection state out of global.
17:15:02 <Alberth> your last commit fixed the problem
17:15:39 <Alberth> makes no sense, does it :)
17:16:29 <glx> anyway this double free can't happen
17:17:05 <Alberth> program crashed after fade out of the intro, and now it starts :p
17:17:55 <glx> yes random crashes still happen
17:18:15 <glx> with SDL still handling the mouse ;)
17:56:45 <DorpsGek> SVN: alberth (r2243) -Add: Move unit status update timer out of global.
17:57:40 <DorpsGek> SVN: alberth (r2244) -Add: Moved animation timers and variables out of global.
18:11:40 <DorpsGek> SVN: alberth (r2245) -Add: Removed 329E from global.
18:19:24 <Alberth> Assertion '!e->dead' failed at pulse/mainloop.c:216, function mainloop_io_enable(). Aborting. <-- don't close the window while sound is being played :p
18:26:26 <Xaroth|Work> release 1: 'note: you cannot close the game while sound is being played'
18:26:35 <Xaroth|Work> release 2: 'background music implemented'
18:31:35 <DorpsGek> SVN: alberth (r2246) -Add: Move XYPosition to opendune.h, move cursorHotSpots out of global.
18:43:24 <DorpsGek> SVN: alberth (r2247) -Add: Move object selection position and layout out of global.
18:56:14 <DorpsGek> SVN: alberth (r2248) -Add: Moved map direction table out of global.
19:00:07 <glx> Alberth: to be exact it's when voice is being played ;)
19:02:06 <Alberth> all sound is still slow as hell here, perhaps without emu, it gets better
19:04:11 <DorpsGek> SVN: alberth (r2249) -Add: Moved border indices and sizes array out of global.
19:04:47 <glx> sounds are fully without emu
19:12:07 <DorpsGek> SVN: alberth (r2250) -Add: Moved creditsAnimation and creditsAnimationOffset out of global.
19:13:07 <DorpsGek> SVN: glx (r2251) -Fix: a deadlock happening when closing the game while a voice is being played
19:15:42 <glx> now I can click the cross without problems at "the planet arrak..."
19:15:53 <glx> previously it always hanged
19:19:52 <Alberth> hmm, I also had hanged games while closing, but not sure whether it was playing sound.
19:21:27 <DorpsGek> SVN: alberth (r2252) -Add: Move stored displayed data out of global.
19:22:59 <Alberth> enough of the global movement game for today :)
20:40:53 <TrueBrain> so how are we doing? :D
20:53:37 <glx> your script detects REMOVED ?
20:54:44 <TrueBrain> that is why I spend so much time on it :)
20:54:54 <TrueBrain> btw, game hangs again if you have no sound with sound enabled
20:55:49 <glx> that hang should be easy to fix now
20:55:50 <TrueBrain> owh, it checks removed_ :p
20:56:48 <DorpsGek> SVN: truebrain (r2253) -Add: moved out var_2BAC (read-only static information)
21:01:24 <DorpsGek> SVN: truebrain (r2254) -Add: hello mister g_inputIgnore
21:01:39 <TrueBrain> slow going thisway :p
21:04:17 <DorpsGek> SVN: truebrain (r2255) -Add: g_global->variable_8DE3 is going away
21:04:58 <TrueBrain> sound/music + voice is forced loaded I think :p
21:06:25 <glx> for sound/music it should be ok
21:06:50 <TrueBrain> just also realised I forgot to remove the parameter
21:06:51 <glx> but the driver is still loaded I think
21:07:25 <glx> without voices intro plays better ;)
21:07:44 <TrueBrain> we need a better system for it :)
21:08:00 <TrueBrain> but first, libemu ;)
21:08:04 <glx> trying something for that
21:08:15 <TrueBrain> if you like, I guess you can add an if (index == 0) return or something
21:13:08 <DorpsGek> SVN: truebrain (r2256) -Add: one more variable out of g_global (these commits are boring)
21:14:47 <DorpsGek> SVN: truebrain (r2257) -Add: s_var_8420
21:15:08 <DorpsGek> SVN: truebrain (r2258) -Fix (r2257): out of boredom, forgot to rename the g_global entry
21:22:13 <TrueBrain> hmm ... 81E8 is readonly, but zero ..
21:25:20 <DorpsGek> SVN: truebrain (r2259) -Add: 3 more variables from g_global
21:25:54 <DorpsGek> SVN: truebrain (r2260) -Fix (r2259): forgot g_global AGAIN
21:26:12 <glx> how to increase commit counter ;)
21:26:54 <TrueBrain> still hangs on Pick House
21:27:42 <glx> so SDL audio init succeed
21:28:02 <TrueBrain> get like 10 messages :p
21:29:32 <DorpsGek> SVN: truebrain (r2261) -Add: moved one more out of g_global
21:29:38 <TrueBrain> why does it 'hang'?
21:29:51 <TrueBrain> can't we detect there something is wrong?
21:34:48 <DorpsGek> SVN: glx (r2262) -Cleanup: leftover declarations
21:41:33 <glx> do the same in DSP_Stop()
21:41:59 <TrueBrain> owh, my last commit did not commit, fuck
21:42:24 <DorpsGek> SVN: truebrain (r2263) -Add: a few more variables
21:43:52 <TrueBrain> DSP_SetTimeConst:64
21:44:41 <TrueBrain> so not there either
21:45:42 <glx> then the SDL_PauseAudio(0) in DSP_ProcessBlock
21:46:06 <TrueBrain> after my next patch :p
21:52:37 <DorpsGek> SVN: truebrain (r2264) -Add: more variables. Also, good luck finding a machine without mouse these days :D
21:54:20 <TrueBrain> nope, seems to work fine
21:54:43 <TrueBrain> of course I can just attach a debugger
21:55:09 <glx> they caused hang on exit for me :)
21:55:25 <TrueBrain> so why didn't you remove it :P
21:55:36 <TrueBrain> still hangs without btw
21:55:49 <TrueBrain> sleep in GUI_PickHouse
21:56:57 <DorpsGek> SVN: truebrain (r2265) -Fix (r2264): guess ... I forgot it again
21:57:28 <glx> so it is playing without playing
22:04:07 <glx> DSP_ProcessBlock() sets playing status, and it should unset it when called by DSP_Calback() once all data is passed
22:04:32 <glx> so there's a problem with the callback
22:12:41 <DorpsGek> SVN: truebrain (r2266) -Add: moar variables
22:13:38 <glx> ok now I need to use that in the init phase :)
22:14:13 <glx> because I guess you get a bunch of ALSA messages now
22:14:27 <TrueBrain> I am used to them :P
22:14:35 <TrueBrain> well, only 2 sets tbh
22:15:47 <TrueBrain> haha, yes on every voice, 10 lines :D
22:20:08 <DorpsGek> SVN: truebrain (r2267) -Add: we don't do mouse snapping
22:20:54 <glx> should disable voices at init
22:21:16 <glx> easy to check with the intro
22:21:55 <TrueBrain> and then it seems fine
22:22:22 <glx> you should have text in the intro
22:25:45 <DorpsGek> SVN: glx (r2268) -Add: detect voice playback problems at init
22:26:54 <glx> oh and drivers are disablable via dune.cfg ;)
22:27:06 <DorpsGek> SVN: truebrain (r2269) -Fix: the last easy g_global entry :(
22:29:00 <TrueBrain> like many many many many many many more shit :P
22:30:27 <glx> should I assume there's always a mouse ?
22:35:08 <DorpsGek> SVN: glx (r2270) -Fix: removed useless checks
22:37:30 <TrueBrain> doing the last g_globals ...
22:38:07 <glx> hmm are string_XXXX counted ?
22:38:20 <TrueBrain> else it would be much higher :p
22:38:40 <TrueBrain> I have 20 entries open ... getting there
22:38:54 <glx> I mean counted in removed
22:39:03 <TrueBrain> it is not in my list
22:39:40 <TrueBrain> owh, SaveLoad issues
22:39:54 <TrueBrain> owh, that was easy :D
22:39:56 <TrueBrain> nice job on the global stuff :D
22:40:39 <DorpsGek> SVN: truebrain (r2271) -Add: some more variables ... I lost count
22:42:15 <glx> I can do 4060 (even if I can't name it)
22:42:31 <TrueBrain> 1 out of the 21? Sure :p
22:42:39 <TrueBrain> I am at 3900 from the top
22:45:34 <glx> unknown are ignored too I guess
22:45:44 <TrueBrain> how else are there 21 left? :p
22:47:24 <DorpsGek> SVN: truebrain (r2272) -Add: and more
22:48:27 <glx> I'm doing those after 4060
22:48:47 <TrueBrain> I hate it when I get 37 hits back from search
22:49:25 <TrueBrain> longest is 10 hits for you, so yeah :P
22:55:32 <TrueBrain> oeh, the game starts and quits immediatly
22:55:37 <TrueBrain> I did something wrong ... somewhere ....
22:57:12 <DorpsGek> SVN: truebrain (r2273) -Add: more variables
22:58:46 <TrueBrain> it is r2272, but ...
22:59:55 <glx> hmm it closes after an bullet sound here
23:01:39 <glx> hmm it doesn't always close at the same place
23:02:11 <glx> something ends the game loop ?
23:03:49 <DorpsGek> SVN: truebrain (r2274) -Fix (r2272): initializing values can be important. black magic. it is fun.
23:08:21 <DorpsGek> SVN: truebrain (r2275) -Add: MOARRR
23:08:30 <TrueBrain> down to the last 10
23:10:05 <TrueBrain> k, one remaining table before I hit 4060, woei :)
23:10:54 <TrueBrain> make a printf to dump it, start app, copy paste, change it :p
23:13:17 <TrueBrain> hmm, we always assumed this thing was astruct
23:13:27 <TrueBrain> but it now seems more like a struct, now I dump the info
23:16:38 <DorpsGek> SVN: truebrain (r2276) -Add: hello mister array at 3A3E. And 3E52 variable.
23:17:15 <TrueBrain> you did 4060, 628C and 6294, right?
23:17:22 <TrueBrain> then I will start on the other end, 6C71
23:19:21 <DorpsGek> SVN: truebrain (r2277) -Add: 2 more, done
23:19:42 <TrueBrain> 6C6E and 6C6C left; you taking them, or should I?
23:21:47 <TrueBrain> well, I have their patch ready
23:21:55 <TrueBrain> I don't want to conflict with any of your possible work, so let me know ;)
23:22:44 <DorpsGek> SVN: truebrain (r2278) -Add: 6C6E and 6C6C done
23:23:10 <TrueBrain> well, that was fun :)
23:35:19 <DorpsGek> SVN: truebrain (r2279) -Fix: don't depend on other includes to include your includes (also remove global.h and libemu.h where possible)
23:36:47 <DorpsGek> SVN: glx (r2280) -Add: g_global can now go (last 3 done)
23:37:22 <DorpsGek> SVN: truebrain (r2281) -Fix: remove last few global.h includes
23:38:35 <DorpsGek> SVN: truebrain (r2282) -Add: global.h, thank you for your service. We salut you!
23:38:38 <TrueBrain> :D:D:D:D::DD:D:D:D:D:D:D:D:D:D:D
23:39:01 <glx> hmm libemu can go now, right ?
23:39:47 <DorpsGek> SVN: truebrain (r2283) -Fix: remove main.c
23:40:45 <DorpsGek> SVN: truebrain (r2284) -Fix (r2283): no longer call libemu IRQ handler
23:41:06 <DorpsGek> SVN: truebrain (r2285) -Add: we no longer depend on LibEMU. Remove it from Makefile (MSVC to follow, I hope; /me looks at glx with pretty eyes)
23:41:19 * TrueBrain takes glx and makes a dance
23:41:23 <TrueBrain> dancing dancing dancing
23:41:48 <DorpsGek> SVN: truebrain (r2286) -Add: remove now unneeded libemu include files
23:41:50 <TrueBrain> almost forgot those
23:42:12 <TrueBrain> for you the honours to fix MSVC project :D
23:42:23 <TrueBrain> I am so happy :D:D:D:D:D
23:45:08 <glx> hmm ok svn played with me :)
23:46:04 <Xaroth|Work> so long libemu \o/
23:46:08 <DorpsGek> SVN: truebrain (r2287) -Remove: also, the memory files are no longer needed (or useful)
23:46:49 <DorpsGek> SVN: glx (r2288) -Remove: libemu from MSVC project
23:51:31 <TrueBrain> 1 conversion and 1 _dup2 warning
23:51:35 <DorpsGek> SVN: glx (r2289) -Fix (r2283): some WIN32 includes disappeared in the move
23:52:39 <TrueBrain> when I start OpenDUNE I get error 1
23:53:06 <TrueBrain> I rebooted my widnows
23:53:09 <TrueBrain> but still no sounds here
23:53:37 <TrueBrain> that is not promising
23:54:10 <TrueBrain> what did I break and how ...
23:55:19 <TrueBrain> even as administator
23:56:36 <TrueBrain> weird, all this time it worked under windows
23:56:40 <TrueBrain> and now it just stopped
23:56:57 <TrueBrain> my sound system does indicate there is a stream of audio
23:58:08 <glx> yes the < fix is weird ;)
23:58:26 <TrueBrain> does voice etc work for you?
23:58:37 <DorpsGek> SVN: glx (r2290) -Fix: MSCV likes to warn
23:58:58 <TrueBrain> then why does it fail here?
23:59:49 <TrueBrain> other sound outputs make little difference
continue to next day ⏵