IRC logs for #opendune on OFTC at 2011-08-06
⏴ go to previous day
10:50:32 *** Alberth has joined #openDune
10:50:33 *** ChanServ sets mode: +o Alberth
11:31:02 <TrueBrain> ugh, spend 3 hours to figure out my xorg was updated (and with that the ABI), but not the input drivers
11:31:09 <TrueBrain> which then went into some half-half mood
11:31:14 <TrueBrain> I could move my mouse, but not click
11:31:44 <Alberth> be glad you know how to use a keyboard for getting system info :p
11:32:26 <TrueBrain> then again, I use Gentoo
11:32:29 <TrueBrain> not much of a choice :D
11:33:32 <TrueBrain> right, starport issues ...
11:33:45 <TrueBrain> no scripting errors
11:34:12 <Alberth> oh, highly configurable linux :)
11:36:53 <TrueBrain> so once I had them deliver my shit to the wrong place
11:36:57 <TrueBrain> once I had graphical errors
11:37:06 <TrueBrain> but nothing turns up :(
11:39:09 <Alberth> weren't startports buggy as well?
11:39:20 <Alberth> (in the original game)
11:39:41 <TrueBrain> but I dont understand why no errors are given etc etc
11:39:49 <TrueBrain> I am guessing something in scripts go wrong
11:40:00 <TrueBrain> but under no way the screen should be affected
11:40:08 <TrueBrain> I can handle that it delivers at the wrong place :p
11:41:42 <Alberth> swap starport deliveries if you play at april 1st :)
11:45:50 <TrueBrain> but okay, the game looks really stable for the rest
11:45:55 <TrueBrain> guess we should release 0.6 tomorrow
11:48:18 <TrueBrain> order of CPU time spend (excluding video shit):
11:48:26 <TrueBrain> well, those I expect
11:48:36 <TrueBrain> lolz! we should inline those :P
11:48:42 <TrueBrain> GUI_Widget_HandleEvents .. well .. meh
11:48:54 <TrueBrain> Tools_RandomRange is SLOW as fuck :p
12:00:03 <DorpsGek> SVN: truebrain (r2354) -Fix: Editbox still had a while(true){} in it, without delay
12:01:23 <DorpsGek> SVN: truebrain (r2355) -Fix: region select screen went 100% too
12:04:17 <TrueBrain> I guess I should also experiment a bit with 1.07hs datafiles etc
12:17:04 <DorpsGek> SVN: truebrain (r2356) -Add: show a more nice error message when loading the first file fails (missing data-files is best guess)
12:17:51 <DorpsGek> SVN: truebrain (r2357) -Add: be a bit more to the point if dune.cfg is missing (would be odd)
12:18:55 <TrueBrain> intro runs with 1.07hs .. promising :p
12:19:28 <TrueBrain> there are 4: 1.00us, 1.07us, 1.07hs, 1.07eu
12:19:49 <TrueBrain> 1.07hs is 1.07eu in most ways I believe
12:21:24 <TrueBrain> works without problem, odd :p
12:22:18 <glx> data files are compatible :)
12:22:27 <TrueBrain> most pak files are of identical size
12:22:27 <glx> exe was a different story
12:22:31 <TrueBrain> I wonder where the differences are
12:25:43 <TrueBrain> slightly different (better?) fonts
12:25:51 <TrueBrain> different translations
12:26:21 <TrueBrain> planetmaker: you here?
12:26:38 <TrueBrain> verfolgt is german, right?
12:26:52 <TrueBrain> k, +1 for my german :P
12:27:18 <TrueBrain> I just wanted to make sure to not make a stupid comment which now follows:
12:27:25 <TrueBrain> hs german translation is significant worse
12:28:03 <TrueBrain> glx: do you know what is in mapmach.cps?
12:31:44 <TrueBrain> hmm .. I don't see enough difference in 1.07us data to be a problem
12:31:48 <TrueBrain> sure it misses french and german
12:31:59 <TrueBrain> (just had to do that :p)
12:32:09 <TrueBrain> and the subtitle is different :p
12:32:30 <glx> voices should differ too then
12:33:20 <TrueBrain> I should check language files indepth I guess
12:33:24 <TrueBrain> to make sure the order is the same too
12:33:35 <TrueBrain> diff is acting weird, it see a binary difference between two files, of which one does not exist
12:34:22 <TrueBrain> funny enough, the 1.07us version has french and german in his scenario pak :p
12:34:36 <TrueBrain> but yeah, some voices are renamed
12:34:56 <glx> hmm file list is inside the exe
12:35:20 <TrueBrain> only goes for 6 files
12:35:30 <TrueBrain> zreport[123] -> report[123]
12:35:56 <TrueBrain> and this is weird ... us doesn't have frigate.wsa
12:35:59 <TrueBrain> guess it is not used?
12:36:32 <TrueBrain> right, let me extract language files and diff again
12:40:09 <glx> if (voiceSet == 0xFFFF || g_variable_3E54[voice] != NULL) break;
12:40:09 <glx> switch (g_config.language) {
12:40:09 <glx> case LANGUAGE_FRENCH: i = 'F'; break;
12:40:10 <glx> case LANGUAGE_GERMAN: i = 'G'; break;
12:40:12 <glx> snprintf(filename, sizeof(filename), str + 1, i);
12:40:23 <TrueBrain> problem is a big word
12:40:38 <TrueBrain> hmm .. weird ... in the 'hs' version, stringID 339 is not Load Game
12:40:44 <TrueBrain> it is the end of the list in fact :p
12:42:00 <TrueBrain> same for US version
12:42:05 <TrueBrain> so we might want to find where that is used
12:43:05 <TrueBrain> the .eng sometimes differ binary
12:43:10 <TrueBrain> but after our read, they are identical
12:43:24 <TrueBrain> guess langread doesn't always work
12:43:29 <TrueBrain> what is in mentath.eng for example
12:43:34 <glx> I see no trace of 0x153 nor 339
12:43:48 <TrueBrain> k, guess it is unused then ;)
12:44:20 <TrueBrain> a diff with the 1.00 version scared me :p
12:44:47 <TrueBrain> lets not go there, lets stay with 1.07 :D
12:44:57 <glx> guess 1.00 doesn't have "Warning: Original saved games are incompatable with the new version. The battle will be restarted."
12:45:07 <TrueBrain> 1.00 has a lot not :p
12:45:20 <TrueBrain> mentath.eng comes back with 2 empty strings?
12:45:53 <glx> yes langfiles can contain cps/wsa
12:46:08 <TrueBrain> can we detect that, for pretty-ness?
12:46:49 <TrueBrain> where for example does it open mentath.eng? Any idea?
12:47:36 <glx> oh mentath.eng should have a NAME chunk in it
12:49:16 <TrueBrain> does appear to contain plain text
12:50:37 <glx> your langread should work if you pass it the NAME chunk data instead the whole file
12:50:39 <TrueBrain> but I can;t really find where it is used ...
12:50:51 <glx> GUI_Mentat_LoadHelpSubjects
12:51:13 <TrueBrain> why did my grep not show it :(
12:51:30 <glx> probably because it's "MENTAT%c"
12:54:05 <glx> hmm there's also DESC chunk :)
12:54:41 <glx> NAME contains the subject list
12:55:53 <TrueBrain> NAME chunk is not encoded
12:56:20 <TrueBrain> owh, that is on another level
12:56:20 <glx> it might be for french :)
12:56:34 <TrueBrain> what is in the DESC chunk btw?
13:01:16 <glx> hmm ok to display help it reads INFO chunk
13:01:53 <glx> which contains address in the file
13:04:17 <TrueBrain> the NAME section doesn't even contain an index thingy
13:04:57 <TrueBrain> just not 2 bytes, but 1
13:05:01 <glx> NAME section contains a list of subjects :)
13:06:10 <glx> with an offset, and some info about how to display it in the list
13:07:18 <glx> first byte in the string is its size
13:07:34 <glx> followed by uint32 offset (BE)
13:08:32 <TrueBrain> and this is only for mentat ... lol
13:08:36 <TrueBrain> makes comparing hard :(
13:10:53 <TrueBrain> uploaded a new langread.c
13:11:03 <TrueBrain> it bails when it finds a chunk file
13:11:54 <TrueBrain> owh, only mentat is not a language file
13:13:22 <TrueBrain> owh, still a few are empty
13:15:05 <TrueBrain> I did something wrong I guess
13:15:07 <glx> some for sure contains "buttons"
13:15:08 <TrueBrain> all files are empty now
13:17:38 <TrueBrain> langread.c is weird
13:19:14 <TrueBrain> okay, title.eng ... any idea when it is used, and what is in it?
13:20:28 <glx> and.eng is in the intro ;)
13:20:52 <glx> between westwood logo an virgin logo
13:21:41 <glx> Sprites_LoadImage(String_GenerateFilename("AND"), 2, g_palette_998A, GFX_Screen_GetSize_ByIndex(2));
13:22:10 <TrueBrain> I now made a list with changes between the pak files
13:22:16 <TrueBrain> let me see one by one if they matter
13:22:22 <TrueBrain> mapmach.cps for example ...
13:22:26 <TrueBrain> new6p.fnt, meh, don't care
13:22:30 <TrueBrain> the other 2 miss new6pg.fnt
13:22:35 <TrueBrain> guess we won't be playing in french
13:25:31 <glx> oh help subject list is very smart :)
13:25:39 <glx> but not human friendly ;)
13:26:22 <TrueBrain> german, fine :p No playing in german for hs and us :p
13:26:39 <TrueBrain> in german language file in the hs version: Name and Rang :D
13:28:02 <glx> let see if I can transform s_helpSubjects to a struct
13:28:24 <glx> because for now it's just a big string :)
13:29:08 <glx> very smart stuff, but well :)
13:29:44 <TrueBrain> okay .. 2 vocs changed in hs->eu
13:29:49 <TrueBrain> frepair and goverour
13:29:53 <TrueBrain> I am sure that will be fine
13:30:02 <TrueBrain> german also has: brrn, chaom and misc
13:30:06 <TrueBrain> I am sure they will be no sisue
13:30:15 <TrueBrain> and all changed mentat[aho]
13:30:24 <TrueBrain> I doubt anyone will play in french/german atm, so I don't care
13:30:42 <TrueBrain> which leaves the question of mapmach.cps
13:30:46 <TrueBrain> I guess we should finish a game to find out
13:31:01 <TrueBrain> for the us version, a few stand out
13:31:11 <TrueBrain> mentath .. well .. I can check that in a bit, but I am sure it is fine
13:31:16 <TrueBrain> title.eng .. no clue when it is used
13:31:20 <TrueBrain> intro.fnt, ugh, don't care
13:31:23 <TrueBrain> and missing frigate.wsa
13:31:37 <glx> title.eng should be the title animation
13:32:05 <Alberth> I get all kinds of graphic glitches during battle
13:32:17 <Alberth> perhaps the x2 algorithm?
13:32:35 <glx> you know the "dune bla bla" think animated in the begining of intro
13:32:39 <TrueBrain> Alberth: when/where?
13:32:53 <glx> and the thing shown on top of the intro menu
13:32:54 <TrueBrain> and no, x2 algorithm makes no graphical glitches (as far as I know)
13:33:03 <TrueBrain> glx: ah! Yes, that changed :)
13:33:13 <Alberth> while fighting with ordos, with missiles and bullets flying all over the place
13:33:35 <TrueBrain> Alberth: the only issues I know about are with the sandworm, leaving residu
13:33:39 <TrueBrain> moving the map fixes that
13:33:47 <TrueBrain> and when using starport, sometimes you have weird effects
13:33:53 <TrueBrain> anything else you can reproduce, please make bug report :)
13:34:23 <TrueBrain> glx: Load Game _is_ used
13:34:27 <TrueBrain> I have a blank spot in the menu :D
13:34:49 <glx> x2 algo can't make glitches, it always uses original screen
13:35:12 <glx> so glitches should be present before transformation
13:35:28 <TrueBrain> k, frigate.wsa and mentath checks out fine
13:35:37 <TrueBrain> that leaves mapmach.cps to check
13:36:51 <glx> TrueBrain: "select your next conquest" screen I guess
13:37:03 <TrueBrain> finishing a game now
13:38:11 <glx> switch (g_config.language) {
13:38:11 <glx> GUI_Screen_Copy(1, 120, 1, 0, 38, 24, 4, 4);
13:38:12 <glx> GUI_Screen_Copy(1, 96, 1, 0, 38, 24, 4, 4);
13:38:24 <glx> translated text is somewhere in the image ;)
13:40:07 <TrueBrain> meh; the game hangs sometimes ..
13:41:05 <TrueBrain> ah, yes I had this hang before
13:41:14 <TrueBrain> it spins infnitely in GUI_HallOfFame_DrawBackground
13:41:58 <glx> oh that makes me think I should check if it's still crash when hall of fame is full
13:46:37 <TrueBrain> I don't really understand why this now becomes an infinite loop, but okay :p
13:51:45 <DorpsGek> SVN: truebrain (r2358) -Fix: strcpy over yourself is like memcpy over yourself, not always a good idea. Make the strcpy a memmove, to fix the crashes
13:51:53 <TrueBrain> right, and 1.07us works fine
13:56:07 <DorpsGek> SVN: truebrain (r2359) -Fix: to make OpenDUNE compatible with 1.07hs and 1.07us, replace "Load Game" string with "Load a game" string in intro menu
14:04:36 <DorpsGek> SVN: truebrain (r2360) -Fix (r2359): I didn't know they used stringID in switch cases ..
14:04:41 <TrueBrain> although Alberth did tell me they did
14:06:01 <DorpsGek> SVN: truebrain (r2361) -Fix: quick-and-dirty: in 1.07us data files some VOC files are named differently. Compensate for that.
14:06:48 <Alberth> what did I do this time? :p
14:07:26 <TrueBrain> you told me they used stringIDs to cmpare actions
14:08:10 <DorpsGek> SVN: truebrain (r2362) -Add: make note that we now support 1.07hs and 1.07us datafiles, and they should work without any problems
14:08:20 <Alberth> TrueBrain: no worries, I forgot that I said that too :p
14:08:27 <TrueBrain> Alberth: yeah, I have seen that before too
14:08:32 <TrueBrain> with some explosion types it happens
14:08:39 <TrueBrain> can't really reproduce, so meh
14:08:50 <glx> and it's not caused by x2 ;)
14:09:00 <Alberth> ah,that could be the reason
14:09:12 <Alberth> there are many explosion types
14:11:00 <TrueBrain> no, x2 can't be the cause of those things :)
14:11:05 <TrueBrain> as it does the whole screen, always
14:11:14 <TrueBrain> so it is 100% a representation of the screenbuffer :)
14:11:38 <Alberth> oh, it is top of smoke
14:11:56 <TrueBrain> I dunno if it is a conversion error, or that it always was this way
14:12:08 <Alberth> I have never seen it before
14:14:54 <TrueBrain> if you can reproduce it, bisect which revision please :D
14:15:01 <TrueBrain> glx: did you check Hall Of Fame, when full?
14:17:55 <TrueBrain> all set for release as far as I know, minus the gfx glitches like sandworm and like Alberth above
14:18:01 <TrueBrain> but tbh, I can live with that :)
14:18:17 <TrueBrain> I suggest we give it another 24h, and then release 0.6 :D
14:18:21 <Alberth> sandworm was also in original dune2
14:18:34 <TrueBrain> it was? Never noticed ..
14:18:37 <Alberth> although less, perhaps
14:20:13 <Alberth> but I agree 0.6 should be released, too much progress not to :)
14:20:34 <TrueBrain> so much more stable :)
14:21:19 <TrueBrain> we came such a long way :)
18:12:23 <TrueBrain> how do you kill those artreides flying units?!
18:12:26 <TrueBrain> I can't get them to kill
18:12:30 <TrueBrain> and more and more get to me :(
18:12:49 <TrueBrain> I died 3 times now to them :p
18:13:02 <TrueBrain> 2 rocket turrets are not killing them, not really (they seem to miss?)
18:13:08 <TrueBrain> missile units don't shoot?
18:15:39 <Alberth> I have had such a problem too
18:15:48 <Alberth> but thought it was me :(
18:15:58 <TrueBrain> they seem undefeatable tbh
18:16:05 <Alberth> do you have missile launchers?
18:20:27 <Alberth> the graphics glitch seems quite old
18:20:48 <Alberth> but at 320x200 it is really hard to see
18:21:30 <TrueBrain> I just upgraded my air factory, but there is no new item to build
18:22:57 <Alberth> perhaps you can build a new item at the heavy factory or so?
18:24:26 <TrueBrain> I can never build those air units
18:24:46 <TrueBrain> neither IX, neither palace .. wtf?
18:25:26 <TrueBrain> I can order them via starport
18:25:37 <TrueBrain> it is just in that list
18:25:59 <TrueBrain> ah, had to build IX, which you get after starport
18:28:17 <TrueBrain> okay, the AI can shoot them down just fine
18:29:54 <TrueBrain> so it might be just me :)
18:31:10 <TrueBrain> but I really thought missile launchers also foughts them off
18:47:45 <Alberth> lol, an old version has some bugs in the sprites, and buildings appear in the sand :p
19:52:25 <planetmaker> TrueBrain: you need an air factory for planes
19:52:33 <planetmaker> it's a separate building
19:56:13 <TrueBrain> lol; I know that planetmaker :P
19:56:29 <planetmaker> :-) I wondered and didn't read back :-P
19:56:41 <TrueBrain> but you can upgrade it before you can build the second thingy
19:56:46 <planetmaker> judging from myself I sometimes have memory holes, though ;-)
20:31:34 <Alberth> \o/ I think I found the problem
20:39:01 <Alberth> ie if the condition at line 10 in fpaste fails, the loc06 should get set
20:40:56 <Alberth> you agree with that conclusion?
20:41:59 <Alberth> but I wonder why I considered reading such code ever 'normal' :p
20:44:39 <TrueBrain> and yeah, you seem to be right
20:44:46 <TrueBrain> but I am not putting my hands in any fire over these statements
20:45:03 <Alberth> too much that can go wrong :p
20:45:08 <TrueBrain> too many times it has :)
20:46:39 <Alberth> I first did a bisect, even though absence of the bug is very hard to see
20:46:58 <Alberth> and I ended up with a 'unit' conversion, so that was promising :)
20:47:06 <TrueBrain> does it also fix the sandworm effect? :P
20:47:23 <Alberth> when does a sandworm smoke? :p
20:47:57 <Alberth> well could be, the variable 0040 can mean anything
20:48:10 <Alberth> I only tested for smoke
20:55:30 <DorpsGek> SVN: alberth (r2363) -Fix (r1258): A failing bit test suffices for setting loc06 (fixes glitches of smoking vehicles driving to the south).
21:39:50 *** planetmaker has joined #openDune
21:39:50 *** ChanServ sets mode: +v planetmaker
continue to next day ⏵
