IRC logs for #opendune on OFTC at 2009-11-18
⏴ go to previous day
08:27:20 <boekabart> gooooooooood morning الراقص!
08:28:50 <boekabart> ok... nice one irssi: "72:90 < boekabart> gooooooo..."
09:01:30 <DorpsGek> SVN: truebrain (r514) [JIT] -Fix: two more problems found by planetmaker
09:05:53 <DorpsGek> SVN: truebrain (r515) [JIT] -Fix (r514): don't forget to add new files
09:06:18 <DorpsGek> SVN: truebrain (r516) [JIT] -Fix (#11, #13): two more problems found by tneo
09:06:43 <TrueBrain> tneo: the problem with the MGV was that it could not deploy ;)
09:09:06 <DorpsGek> SVN: truebrain (r517) [JIT] -Fix (#12): 1.00 savegames can't be loaded; instead it restarts the campaign the savegame was in (tnx to SmatZ for the savegames)
09:10:34 <DorpsGek> SVN: truebrain (r518) -Update (r514, r515, r516, r517, #11, #12, #13): update decompiled code
09:10:43 <TrueBrain> Xaroth: Mantis script is still broken :(
09:13:37 <TrueBrain> k, that were all the crash logs and stuff :)
09:14:35 <Xaroth> boekabart: I think it's not liking the utf8 mess :P
09:14:40 <Xaroth> use a proper encoding, ASCII!
09:15:01 <TrueBrain> Xaroth: same as the one I told you before, when there are multiple 'r' and multiple '#', it doesn't do anything :p
09:15:07 <TrueBrain> any other combo seems to work
09:15:28 <planetmaker> wow... you're fast fixing things :-)
09:16:19 <planetmaker> btw, may I once again propose to implement an autosave functionality - as crude as it may get even - quite early on?
09:16:34 <planetmaker> It makes testing WAY more fun, if I don't have to remember to save every 2 minutes or so
09:16:37 <TrueBrain> if we would know how ;)
09:16:49 <planetmaker> uhm... but one can manually save, can't oneß
09:17:10 <Xaroth> yes, but that part of the code isn't decompiled yet :P
09:17:19 <TrueBrain> yesterday we discovered some key functions in the scripting engine .. but no sign yet of save/load stuff :p
09:17:21 <planetmaker> don't tell me it's hidden in the de-compiled code and you don't know where... ? :-O
09:17:21 <Xaroth> I can't think this morning
09:17:35 <TrueBrain> you do realise there are 100,000 lines of decompiled code, right? :p
09:17:37 <Xaroth> < planetmaker> don't tell me it's hidden in the de-compiled code and you don't know where... ? :-O << I won't then :)
09:17:54 <planetmaker> hehe :-) I didn't count, no :-)
09:17:59 <TrueBrain> we uncovered about 10,000 of them (still already a nice progress :p)
09:18:05 <TrueBrain> and rewrote 5,000 of them
09:18:41 <planetmaker> ok :-) I guess I have no real idea how it looks like :-) - even though I looked at it ;-)
09:18:48 <TrueBrain> and even if I would find the routine who takes care of it .. I have no real idea how to make it autosave :(
09:18:58 <planetmaker> and assumed you were WAY more apt at reading that stuff than me ;-)
09:19:12 <Xaroth> think I'll rewrite the mantis thing.
09:21:16 <TrueBrain> planetmaker: you should try rewriting a function yourself, it ispretty fun work :)
09:21:32 <TrueBrain> it shows nicely how deterministic a computer application is :p
09:22:41 <Xaroth> TrueBrain: did the other erronous one also have a [TAG] ?
09:22:52 <TrueBrain> 518 doesn't have a [TAG]
09:23:09 <TrueBrain> and I told you about it :p (forgot about it too :p)
09:24:10 <Xaroth> I'm tempted to just change the whole checkin script :P
09:24:31 <planetmaker> TrueBrain, I wish, I wish I had as much time as I wished to have ;-)
09:24:42 <TrueBrain> same here my friend, same here ..
09:25:43 <planetmaker> but I guess I should know assembler for that end, eh?
09:25:56 <TrueBrain> no, just a basic understanding of program flow is enough
09:26:11 <TrueBrain> the i286 instruction set is simplified enough
09:30:15 <planetmaker> hm... looks quite cryptic to me. Though you're certainly right, it needs time to get used to. And time in this sense is more than just a couple of hours
09:31:25 <planetmaker> /* Unresolved jump */ emu_ip = 0x00DD; emu_last_cs = 0x01F7; emu_last_ip = 0x00DD; emu_last_length = 0x0008; emu_last_crc = 0xAD68; emu_call();
09:31:25 <planetmaker> <-- let me guess. A line like that will crash the game, if called?
09:34:04 <TrueBrain> but there are not that many left ;)
09:34:24 <TrueBrain> of which a few are after a Terminate
09:37:27 <TrueBrain> LOL! I am watching the latest NCIS ... you see a dead person on a table being cut in ... and you see her breath!! :p
09:46:17 <Xaroth> yeh, definately rewriting the checkin thing
09:46:36 <TrueBrain> just make sure it works ;)
09:48:49 <Xaroth> best of it all is, it most likely would fail if the commit message was >1024b
09:49:01 <Xaroth> since it reads 1024 byte, then checks if it can parse
09:49:19 <Xaroth> seeing the message is a variable length, it'd ignore the fact there was more to come
10:01:00 <Xaroth> argc/argv are much easier things to use than reading STDIN...
10:04:52 <TrueBrain> the latest NCIS is briliant :)
10:06:11 <Xaroth> with newline support, even
10:06:27 <Xaroth> history file will be a bit of a mess, but hey :P
10:12:25 *** TinoDidriksen has joined #openDune
10:17:00 <TrueBrain> hmm .. that can be annoying .. hmm ..
10:35:55 <TrueBrain> tss, cheating AI ..
10:36:01 <TrueBrain> it just places the building in direct-ready mode
10:38:58 <Xaroth> TrueBrain: -fix/stuff should be ended with a : right?
10:39:08 <Xaroth> as in -Fix: msg or -Fix (): msg
10:40:49 <Xaroth> just made the wierdest regex evar:
10:40:51 <Xaroth> -(?:fix|add|branch|cleanup|codechange|change|doc|documentation|feature|merge|move|prepare|update|regression|release|remove|revert|sync|backport)(?:\:)?
10:41:01 <Xaroth> ... now to clean that up to something more.. neat :P
10:41:14 <TrueBrain> why you want to scan for the exact keywords?
10:41:17 <TrueBrain> why not just skip to the first (?
10:41:26 <TrueBrain> (or :, what ever comes first)
10:42:31 <Xaroth> er, i want to filter out the keyword :)
10:46:46 <TrueBrain> hmmm .. an AI can only build 5 times at the same time .. or 5 buildings in the CY .. which is odd, as it happened it couldn't store the info in the 5 long array ... sometimes, things are WEIRD
10:50:36 <Xaroth> (?:\[([a-z]*)\] )?-([a-z]{3,})(?:\:)
10:51:01 <Xaroth> filters out the [jit] stuffs
10:54:14 <TrueBrain> lol, the AI really is fucked up :p
10:54:27 <TrueBrain> if you destroy 5+ buildings, it won't build a new one for a while :)
10:54:42 <TrueBrain> what it does on structure destroy, is save in a 5 array buffer where the structure was
10:54:47 <TrueBrain> it starts to make a new structure immediatly
10:54:54 <TrueBrain> when that structure is done, it checks that array where it should place it
10:55:04 <TrueBrain> if by then the structure is no longer in the buffer, it is removed again
10:55:16 <TrueBrain> in short, destroy 1 critical building and 5 non-essentials
10:55:20 <TrueBrain> and you will never see the ciritcal building again
10:55:58 <TrueBrain> it is worth a blog I guess
11:03:09 <TrueBrain> oh, and it appears AIs can run out of credits (didn't expect that)
11:04:16 <Xaroth> Logmessage : [JIT] -Fix (r1, #5123, 321, r879) BLAAT
11:04:16 <Xaroth> Parsed : (r1, #5123, 321, r879) BLAAT
11:04:16 <Xaroth> Fix List : r1, #5123, 321, r879
11:05:08 <Xaroth> that should do the trick :)
11:06:44 <Xaroth> it can even filter out the rest of the message
11:06:59 <Xaroth> so more 'pretty' log messages on mantis
11:19:28 <Xaroth> TrueBrain: do you want issues to be set to fixed when they appear in a svn commit
11:19:31 <Xaroth> or just a log message?
11:22:26 <TrueBrain> hmmm ... this I really don't get ... when you press 'upgrade', variable_06 is set to & 0x2 ... but when you build something, variable_06 is the name of the structure you are building .. well, the StringID
11:22:41 <TrueBrain> one comes from the Info struct :p
11:31:19 <TrueBrain> haha, this is funny: if (.. && h->credits != 0) { if (s->hitpoints < si->hitpoints / 2 && h->credits != 0) ....
11:50:31 <TrueBrain> if (s->upgradeTimeLeft > 5) {
11:50:32 <TrueBrain> s->upgradeTimeLeft -= 5;
11:50:35 <TrueBrain> s->upgradeTimeLeft = 0;
11:50:38 <TrueBrain> if (s->upgradeTimeLeft == 0) {
11:51:57 <TrueBrain> did you know ...... that repairing a vehicle is highly more expensive if your repair bay is more damaged?
12:02:54 <TrueBrain> emu_incw(&emu_get_memory16(emu_ds, 0x00, 0x37A4));
12:02:55 <TrueBrain> emu_decw(&emu_get_memory16(emu_ds, 0x00, 0x37A4));
12:11:03 <TrueBrain> if a script fails to execute, it aborts all other structures too
12:11:16 <TrueBrain> I think it is a bug in for unrolling ..
12:24:28 <Xaroth> Fatal error: Maximum execution time of 30 seconds exceeded in /var/www/bugs.opendune.org/htdocs/core/config_api.php on line 88
12:25:24 <TrueBrain> hmm, I am suprised that starport is not handled in GameLoop_Structure
12:31:34 <TrueBrain> still a bit of work, but almost there
12:32:27 <TrueBrain> it looks like repair of vehicles was first handled by one part, later changed to another part :p
12:34:24 <Xaroth> apparently there's a new RC
12:36:56 <Xaroth> can't hurt to try, right :P
12:37:10 <TrueBrain> you should really be more clear in such lines :p
12:37:47 <Xaroth> doing too much at once
12:37:53 <Xaroth> so I don't really make sense, even to myself :P
12:44:28 <TrueBrain> make sure to make the 'r' links ;)
12:45:18 <TrueBrain> MrFlibble is as useful as always :) (no sarcasm)
12:45:41 <TrueBrain> Xaroth: no, but it would be nice if the 'r1' stuff links to svn.opendune.org
12:48:58 <Xaroth> right, it appears to work now
12:49:26 <Xaroth> wait, i forgot the revision!
12:49:41 <glx> lot of emu_String_GetString() ;)
12:50:49 <TrueBrain> yeah :) It is a bit like OpenTTD :p
12:50:59 <TrueBrain> check out the windtrap part :p
12:51:13 <glx> reading the diff for now :)
12:51:35 <TrueBrain> Xaroth: and please add a test with 2 bug-reports attached
12:54:09 <Xaroth> and another mixed #1, 1
12:55:07 <TrueBrain> now the 'r' are left ;)
12:56:04 <TrueBrain> I want linkylinks to svn.opendune.org!! :p
12:59:33 <Xaroth> revisions will be tricky
12:59:44 <Xaroth> seeing the whole converter for bugs is a huge mess
13:04:29 <TrueBrain> only in the line you put them behind eachother
13:04:32 <TrueBrain> that should be easy enough :p
13:04:41 <TrueBrain> Refers to: r1, r879. <- so those
13:07:24 <glx> +,...,...,...,...,...,.../* XXX -- EEeeuuuuuuuhhhhhh...... WHAT?! */ <-- lol :)
13:08:41 <TrueBrain> yeah, read the next 2 lines
13:08:45 <TrueBrain> it really doesn't make any sense to me
13:08:57 <TrueBrain> so if you like, please confirm I didn't make a weird mistake
13:08:58 <glx> I saw the lines after the comment yes
13:15:51 <glx> indeed there's no mistake
13:16:15 <TrueBrain> really ... what the fuck :)
13:16:29 <TrueBrain> but okay, it means I can just remove the lines :)
13:17:30 <TrueBrain> I realyl really don't get what that global variable variable_37A8 does ... or 37A2 for that matter ... every structure it is changed ...
13:18:07 <TrueBrain> but it seems to be a 'write' only variable
13:18:20 <TrueBrain> for units and airunits the same thing happens, and no other usages ..
13:18:39 <TrueBrain> same for 37A2 and 37A4 ..
13:18:42 <TrueBrain> so useless, if you ask me :p
13:19:26 <TrueBrain> euh ... 379C is never written to
13:19:31 <TrueBrain> so how did it ever became non-zero ...
13:20:55 <TrueBrain> oh, it was never called :)
13:21:08 <TrueBrain> so lets see what happens when I enable it :p
13:22:32 <TrueBrain> cool, it allows you to control the AI :)
13:23:14 <TrueBrain> so yes, a debug thingy :)
13:25:52 <glx> strange sometimes 0x379C != 0 is decompiled, sometimes it's not
13:26:02 <TrueBrain> yes, when there are no jumps, it is decompiled
13:26:13 <TrueBrain> decompiling aborts when there is a non-conditional jump
13:26:30 <glx> even if it doesn't execute?
13:28:59 <glx> but there's always a jump else it's not an if
13:29:13 <TrueBrain> conditional jumps ;)
13:30:27 <Xaroth> TrueBrain: I can make the script input the links to the revisions rather than a list of r<numbers>
13:30:48 <TrueBrain> that is what I was refering to ;)
13:31:32 <TrueBrain> the 379C also shows a grid on the map
13:33:28 <TrueBrain> if 379E is set, you win the mission
13:35:00 <TrueBrain> the ',' is also part of the url :p
13:35:01 <Xaroth> yeh, i messed up the link
13:36:24 <DorpsGek> SVN: truebrain (r519) [JIT] -Add: found a debugAI switch, which allows you to control AI units
13:37:23 <DorpsGek> SVN: truebrain (r520) -Update (r519): update decompiled code
13:38:02 <TrueBrain> else I wouldn't have done this :p
13:38:18 <TrueBrain> glx: despite that, 37A2, 37A4 and 37A8 seem unused :)
13:38:23 <TrueBrain> I guess they are part of some script debugger
13:38:40 <glx> like the empty functions ;)
13:39:54 <TrueBrain> what is the last log line?
13:40:12 <TrueBrain> and I like it that it puts those things below eachother, can you do the same for 'Affects'?
13:40:47 <Xaroth> last log line is a link to svn.opendune.org with the revision comitted
13:40:47 <TrueBrain> oh, I understand what you tried to do with the last line, but the begin is a bit misleading :)
13:40:56 <TrueBrain> I would name it 'SVN url' or something
13:41:05 <TrueBrain> now it looks part of the log message :p
13:41:23 <TrueBrain> I like this debugAI functionality :) Hihi :)
13:41:31 <TrueBrain> you can move approaching enemies back :p
13:41:36 <Xaroth> and affects: is multi-line now as well
13:41:38 <TrueBrain> stop their harvesters :p
13:42:02 <Xaroth> debugAI gives full field of view?
13:42:36 <glx> there's another debug variable for that I guess :)
13:42:46 <TrueBrain> yes, but it makes the game not run :p
13:43:00 <Xaroth> right, you can continue to bugfix again, TrueBrain :P
13:43:02 <TrueBrain> hmm .. I named this var 'debugAI', but it also shows a grid and some minor other things it seems
13:43:07 <TrueBrain> Xaroth: thank you!!
13:43:22 <TrueBrain> debugGame is more representing I guess
13:43:33 <Xaroth> I'm sure there's a bug or three still in it, but I can't fix those until I see what goes wrong :P
13:44:28 <TrueBrain> bah ... sometimes I want to add comments like this: /* Reload the script with the structure type */
13:44:37 <TrueBrain> the code reads: Script_Reset(); Script_Load(s->type);
13:44:40 <TrueBrain> useless comment ...
13:44:53 <TrueBrain> glx: 37A8 is only read just before written to
13:47:22 <TrueBrain> it indeed seems you can resume the debugScenario in-game
13:47:49 <TrueBrain> if (g_global->debugScenario) return; <- again I would like to add a comment like: /* When in debugScenario, do not run structure tickers */
13:47:51 <TrueBrain> but I mean ... DAH!
13:58:51 <TrueBrain> hmmmm... structures cost more than they should :p
13:59:00 <TrueBrain> and my CY just degraded ..
13:59:23 <TrueBrain> which should be impossible, as in the first level degrade stuff is deactivated :s
14:05:12 <TrueBrain> affected areas: repair structure, repair of units, building structures, upgrading structures, Palace weapon, degrading of structures (from campaign 2 and on), scripts of structures (animation, ...)
14:09:22 <TrueBrain> k, gone for the rest of the day; let me know any problems with the above patch :) Tnx!
18:55:10 <tneo> deploy bug isn't correctly fixed :-P
18:55:19 <tneo> now the vehicle disappears
18:55:42 <tneo> want me to make a new bug report?
21:33:01 <TrueBrain> tneo: there is no such thing as 'not correctly fixed' ;)
21:33:05 <TrueBrain> Xaroth: did you test my patch?
21:33:55 <glx> I just read it, looks good
21:34:26 <glx> I'm trying to force jump resolving in emu_GameLoop_House
21:37:44 <glx> hey money cheating is not immediate :)
21:42:10 <TrueBrain> tnx for reading, but I also need someone to test it :)
21:42:15 <TrueBrain> see if it still works the same ;)
22:00:25 <DorpsGek> SVN: truebrain (r521) 2) If one structure gives a script error, all structures after it are not handled in the GameLoop
22:00:25 <DorpsGek> SVN: truebrain (r521) 1) Repair cost is calculated unfair (Palace repair is for free, and you always under-pay)
22:00:25 <DorpsGek> SVN: truebrain (r521) -Add: C-ified GameLoop_Structure(), figured out a lot of variables and names along the way. Also added 2 enhancements:
22:00:30 <TrueBrain> now trunk is the test-bed ;)
22:02:48 <TrueBrain> tneo: please make a new bug report about #11, as it had nothing to do with the crash-log, but a bug in Dune2 itself
22:04:21 <TrueBrain> Xaroth: wtf did you do? I can no longer download files from Mantis without being logged in (which is VERY annoying to retrieve crash-bins)
22:04:39 <TrueBrain> please remove such silly 'protections' .. we have plenty of bandwidth .. lets keep things simple and easy ;)
22:05:28 <Xaroth> TrueBrain: Didn't test your patch, was at missus
22:05:40 <TrueBrain> please do :) This is ANNOYING :)
22:06:04 <Xaroth> and when I'm logged in as anonymous I can download crash.bin on it
22:06:13 <TrueBrain> but you have to be logged in
22:06:28 <Xaroth> that's how mantis works
22:06:48 <TrueBrain> yesterday, and all days before that, I downloaded lots of crash-bins with wget
22:06:51 <TrueBrain> I did that all the time :p
22:06:53 <TrueBrain> (like we had that many :))
22:07:29 <TrueBrain> that was why I was like: what the fuck? :)
22:09:09 <DorpsGek> SVN: truebrain (r522) [JIT] -Fix (#14): one more problem found by tneo
22:09:20 <DorpsGek> SVN: truebrain (r523) -Update (r522, #14): update decompiled code
22:09:36 * TrueBrain hugs Xaroth, your new svn->mantis bridge works nicely
22:09:55 <TrueBrain> maybe omit 'refers to' when there are no 'refers to'
22:12:50 <TrueBrain> k ... it has been a LONGGGGgggggggg day
22:12:54 <TrueBrain> so I wish you all a very good night :)
22:12:57 <Xaroth> * Wiki Integration Enabled?
22:12:57 <Xaroth> * @global int $g_wiki_enable
22:13:06 <TrueBrain> glx: good luck on GameLoop_House ;)
22:13:25 <glx> it's not doable for now (missing jumps)
23:54:39 <Xaroth> TrueBrain: I checked what file_download calls
23:54:53 <Xaroth> and the difference to the previous installed version (made a backup)
23:55:29 <Xaroth> at least nothing that could change the result
continue to next day ⏵