#exult@irc.freenode.net logs for 2 Jun 2002 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage


[18:18:04] --> exultbot has joined #exult
[18:18:04] --- Topic for #exult is: Exult: an open-source engine for Ultima 7: http://exult.sf.net/. No humans or monkies allowed in this channel. *fluff*
[18:18:04] --- Topic for #exult set by Darke at Wed May 29 07:56:05 2002
[18:18:31] * wjp is going to play some Geneforge; bbl
[18:19:04] <-- wjp has left IRC ()
[18:22:04] --> Fingolfin has joined #exult
[18:22:55] <Fingolfin> yo
[18:22:57] --- ChanServ gives channel operator status to Fingolfin
[18:23:00] <Fingolfin> damn it's empty here
[18:26:35] --> Colourless has joined #Exult
[18:26:35] --- ChanServ gives channel operator status to Colourless
[19:11:14] --> Cless has joined #Exult
[19:11:33] <-- Colourless has left IRC (Killed (NickServ (Ghost: Cless!Cless@ppp1519.adelaide.on.net.au)))
[19:11:33] --- Cless is now known as Colourless
[19:11:35] --- ChanServ gives channel operator status to Colourless
[19:15:19] * Colourless scratches his head
[19:18:50] <Colourless> this win9x crash problem win mingw is just strange.
[19:21:22] <Colourless> doing the following doesn't work
[19:21:39] <Colourless> ofstream out;
[19:21:45] <Colourless> if (something_that_is_false)
[19:21:46] <Colourless> {
[19:21:55] <Colourless> code_that_never_executes();
[19:21:56] <Colourless> }
[19:22:11] <Colourless> U7open(out, GNEWGAMEVER);
[19:22:22] <Colourless> getVersionInfo(out);
[19:22:33] <Colourless> out.close();
[19:22:34] <Colourless> }
[19:22:54] <Colourless> but if i move "ofstream out;" to after the if, everthing works
[19:25:22] --> armav has joined #exult
[19:26:56] <-- armav has left IRC ()
[19:27:29] --- Darke|afk is now known as Darke
[19:29:18] <Darke> Hi.
[19:29:33] <Colourless> hi
[19:35:24] --> Dominus has joined #exult
[19:35:44] --- ChanServ gives channel operator status to Dominus
[19:35:51] <Dominus> hi hi
[19:35:57] <Colourless> hi
[19:36:51] <Colourless> Dominus: found and fixed the win9x exult studio bug. haven't committed it yet though
[19:36:59] <Dominus> what was it?
[19:37:06] <Dominus> great, btw
[19:37:23] <Colourless> trailing slashes on paths
[19:38:09] <Dominus> I had a most annoying mail from pertex in which he complained about Es not being compatible with Win9x
[19:38:33] <Dominus> sth like: "now do I have to purchase a new computer with XP to get it to work, or what?"
[19:38:37] <Dominus> most annoying
[19:40:39] <Dominus> so, taking from Jeff the new release will be 0.99.1 RC2
[19:41:37] <Colourless> yeah looks like it
[20:13:54] --> wjp has joined #exult
[20:13:54] --- ChanServ gives channel operator status to wjp
[20:13:59] <wjp> hi again
[20:14:06] <Dominus> hi
[20:14:08] <Colourless> hi
[20:14:25] * wjp is running around in the Main Research Labs.. yay :-)
[20:14:26] <wjp> (geneforge)
[20:14:48] <Dominus> he he
[20:15:50] * wjp looks at logs
[20:15:52] <wjp> that _is_ weird
[20:16:48] <wjp> Colourless: did you find the spaces problem?
[20:17:23] <Colourless> something is acting very screwey with the streams. if i remove my convert to short name code (so there is spaces), exult dies in Flex::is_flex(char *) when doing in.close() it seems
[20:17:44] <Colourless> of course i 'think' the problem is with spaces.
[20:18:00] <Colourless> the dir name that i'm using also has a & in it
[20:19:25] <Colourless> incidently only has a problem with mingw in win98. no issues in winxp, or when using msvc
[20:22:26] <wjp> hm
[20:22:34] <wjp> do you have the latest version of mingw?
[20:24:59] <Colourless> no sure.
[20:25:04] <Colourless> s/no/not/
[20:33:54] <Colourless> i'm downloading the latest to see what happens. but I don't really have any more time to do anything tonight.
[20:34:05] <Colourless> download is 10 mb
[20:40:17] <-- Colourless has left IRC ("time for me to go")
[20:50:32] <-- Fingolfin has left IRC ("42")
[21:11:55] * Darke thinks we need an 'interesting but not particularly useful' documents directory for pentagram. *grin*
[21:12:53] <Dominus> he
[21:13:11] <wjp> what kind of documents do you have in mind, exactly?
[21:14:18] <Darke> Currently I've got two. One's a list of all the globals in u8, seperated into those 'used' (about 450 of them), and 'unused' (about 50 of them).
[21:14:43] <Darke> Which is a 'kind of neat, but what application might it have?' document. *grin*
[21:14:49] <wjp> maybe we should create a docs/usecode/ subdir?
[21:16:11] <Darke> And another is a list of urls, from google pointing to general u8/TGWDS information. Everything from Ely's posts on engine details, to the 'announcement' of TGWDS 2. *grin* 'Interesting' reading, but again not much use.
[21:17:04] <Darke> It might be worth it. Either that or a 'scratch' directory, where we dump half-formed docs and general information, then move them to a more applicable place when they're of some use, or delete them when they're no longer useful?
[21:18:13] <wjp> that would be ok too
[21:19:57] * Darke prefers the 'scratch' directory himself, just on the principal of you can just drop 'information' into there, without having to worry about cvs related worries of getting correct directories/filenames correct 'first time' and so on. *grin* It's less of a 'permanance' thing.
[21:21:46] <Darke> <cvs_gripe> Of course if the cvs history was dependant on the file itself, not the file or directory name, that really wouldn't be a problem. </cvs_gripe>
[21:22:10] <wjp> yeah, cvs is really annoying
[21:28:26] * Darke (on an completely different topic) would love to know how they handled 3 or 5 byte long global flags in u8u. Did they push them as 4 or 6 bytes? Is the byte padding at the 'start' or 'end'? As a literal 3 or 5 byte sequence, thus mis-aligning the stack?
[21:28:53] <wjp> I think I've seen examples of misaligned stack vars
[21:28:56] <Darke> Hmm... actually, the 1 byte ones are certainly pushed as 2 bytes (I'm looking at a `push global; push byte; cmp word` sequence atm.)
[21:29:25] <wjp> oh, no, that was a misaligned local var
[21:30:08] * Darke guesses from that, there's padding at the 'front' of the odd sized ones, just like the pushed bytes.
[21:31:02] <Darke> I see they removed the 10byte long 'lastPotPeriod' flag. *grin* They must have moved that into the engine or something, rather then handling it in usecode.
[21:40:47] * Darke will create a 'scratch' directory in the 'old' module, to dump the files into for now.
[21:41:52] <wjp> I'd make it a docs/scratch, dir, btw
[21:41:57] <wjp> s/,//
[21:42:05] * Darke considers that once we've moved most of the code into the 'pentagram' module, the old module will likely look like a 'scratch' module anyway. *grin*
[21:42:18] <Darke> *nod* Good idea. Will do.
[21:42:40] <wjp> I'm fine with a docs/scratch dir in the new module too, btw
[21:44:29] <Darke> It'd probably be best to put it in the new module then, since the docs module in old/ is depreciated, isn't it?
[21:44:39] <wjp> it's not there, even
[21:45:33] * Darke checks. Indeed it isn't. Shows you how often he looks at the docs, doesn't it? *grin*
[21:46:25] <wjp> :-)
[21:47:36] * Darke claims it's because there's no real usecode documentation, except in the code. Something he'll likely have to fix at some point in time. *grin*
[22:35:13] * wjp should go to bed
[22:35:16] <wjp> bye
[22:35:18] <-- wjp has left IRC ("Zzzz...")
[23:46:06] --> Kirben has joined #exult
[23:46:06] --- ChanServ gives channel operator status to Kirben