#exult@irc.freenode.net logs for 25 May 2002 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage

[00:13:44] --> cnj has joined #exult
[01:31:51] --- wjp|sw2 is now known as wjp
[01:32:03] * wjp is back, and goes straight to bed
[01:32:05] <wjp> night :-)
[01:32:10] <-- wjp has left IRC ("Zzzz...")
[01:32:53] <Dominus> he he
[02:18:12] <-- Dominus has left #exult ()
[02:38:48] <-- Darke has left IRC (calvino.openprojects.net irc.openprojects.net)
[02:38:48] <-- cnj has left IRC (calvino.openprojects.net irc.openprojects.net)
[02:38:48] <-- kefka has left IRC (calvino.openprojects.net irc.openprojects.net)
[02:38:48] <-- voyager_106 has left IRC (calvino.openprojects.net irc.openprojects.net)
[02:43:19] --> Darke has joined #exult
[02:43:19] --> cnj has joined #exult
[02:43:19] --> kefka has joined #exult
[02:43:19] --> voyager_106 has joined #exult
[02:44:45] --> Kirben has joined #exult
[02:44:45] --- ChanServ gives channel operator status to Kirben
[02:47:44] --> Suxeb has joined #exult
[02:47:50] <Suxeb> greetings
[02:47:59] <Darke> Hi.
[02:48:09] <Suxeb> hi!
[02:48:22] <Suxeb> how goes it?
[02:48:37] <Suxeb> log seemed interesting but i forgot what was i thought was interesting about it
[02:48:41] <Darke> Umm... 'normal' I suppose, or as normal as it gets around here. *grin*
[02:48:59] * Suxeb changes Darke's schedule to dance.
[02:49:16] * Suxeb gives Darke some rum.
[02:50:03] <Darke> Let's see System Shock 2, exult_studio, random webpages with 'exult' in them, and 3dfx. I think that's about everything that was a converational piece today. *grin*
[02:53:52] <Suxeb> :)
[02:53:57] * Suxeb recalls.
[02:54:20] <Suxeb> you know "very unique" sounds odd?
[02:55:02] * Darke doesn't think so, but knows why some might think so. *Grin*
[02:56:03] --- Suxeb is now known as sb-x
[02:56:48] <sb-x> How do you like these function names: client_stuff_queue, client_read_buf, client_send_queue, server_message?
[02:57:02] <sb-x> s/client_send_queue/client_write_queue/
[02:57:58] <Darke> What's the differece between send and write?
[02:59:46] <sb-x> ...
[03:00:07] <sb-x> one sends messages and the other writes them
[03:00:28] <sb-x> but actually stuff is what sends messages
[03:00:34] <sb-x> there is no send
[03:00:42] * Darke ahhs. Sorry. 'stuff' and 'write'? *grin*
[03:01:04] <Darke> Ok. So write just dumps the data into it, then it waits for a 'stuff' to send it?
[03:02:20] <sb-x> no it is the other way around
[03:02:27] <sb-x> at least i think thats how i want it to work
[03:02:28] <sb-x> brb
[03:04:21] * Darke considers you probably want to replace 'stuff' with 'write' and 'write' with 'send' then. 'stuff' is really do generic. *grin*
[03:11:13] <sb-x> ack
[03:11:17] <sb-x> i mean back
[03:11:18] <sb-x> and ack
[03:11:32] <sb-x> client_write_queue actually does call 'write'
[03:11:43] <sb-x> on the socket descriptor that the client owns
[03:12:13] <sb-x> what do you mean by stuff being generic?
[03:12:24] <sb-x> it is for generic data :-) (in a char* buffer)
[03:13:10] <Darke> It doesn't really tell me anything about what it does 'append_data' or 'add_data' tells me something 'stuff' doesn't. *grin*
[03:14:06] <sb-x> How do you think I should name it and still keep the 'queue' part?
[03:14:25] <sb-x> Because there is also 'client_stuff_buffer'
[03:14:42] <Darke> 'client_append_queue'?
[03:14:59] <sb-x> but i'm not sure if i'll need that one since it makes no sense to append data to the received data storage area :)
[03:15:06] <sb-x> oooh
[03:15:16] <Darke> If you're just 'appending' data to the end of the queue in preperation for a write or something.
[03:15:19] <sb-x> but doesn't that make you think i'm appending a queue?
[03:15:24] <sb-x> to the queue
[03:16:00] <sb-x> client_append_to_queue
[03:16:08] <Darke> That's better.
[03:16:17] <sb-x> ClientA2Q?
[03:16:19] <sb-x> hehe
[03:16:49] <sb-x> ucfunc_389 :P
[03:16:55] <sb-x> sorry
[03:16:56] <sb-x> thanks
[03:17:45] * Darke thwaps. Bad!
[03:19:00] * Darke just found out the problem he's spent probably the last three days trying to fix with his decompiler, is almost certainly a problem with gcc3.0.4. He's a little sick of usecode at the moment. *grin*
[03:21:00] <sb-x> A problem with GCC, or a problem with the programmer?
[03:21:06] <sb-x> You decide.
[03:21:09] * sb-x nods knowingly.
[03:22:03] * sb-x is really glad he hasn't attempted an upgrade.
[03:22:48] * Darke is pretty sure it's gcc. But he's just started to rewrite the relevant code from scratch today, so should have a good chance to find out himself. *grin*
[03:23:41] * sb-x types 'emerge' and nothing happens.
[03:23:47] <sb-x> :`-(
[03:23:52] <sb-x> :'-(
[03:24:56] * Darke is actually doing that at the moment. Creating a .tbz2 file for both gcc3.0.4 and gcc3.1, then going to delete all the include rememanants from gcc3.0.4 and then install gcc3.1 over the top. So, in theory, it'll work. *grin*
[03:25:46] <sb-x> I have never seen the tbz2 extension.
[03:25:52] <sb-x> just .tar.bz2
[03:26:15] <sb-x> and .bz2
[03:26:15] <Darke> Yeah. I don't like the shorthand easier, but it's what emerge creates if you tell it to create a 'precompiled' package for you. *grin*
[03:27:18] <sb-x> Gentoo uses BZIP2 packages?
[03:27:24] <sb-x> precompiled packages i mean
[03:27:27] * sb-x wants.
[03:28:14] <sb-x> i understand it is source mostly
[03:28:58] <Darke> Normally only on the base install CD (you can start with most of your basic system precompiled if you want, saves time). The only .tbz2 files available are the ones you create yourself. So you can comple the system on one machine of a particular cpu type, then just install the 'packages' on every other machine of the same machine type, rather then compiling it slowly on all.
[03:29:56] <Darke> Also useful if you, say, have a i586/600Mhz and a i586/150Mhz. Compile on the faster, then just copy and install to the slower. *grin*
[03:56:00] <sb-x> what are rememanants?
[03:59:39] <Darke> Just old header files from 3.0.4 which remain in /usr/include somewhere, whereas the 'new' 3.1 files are put in a subdirectory further into it. So the compiler finds the old (and bad) headers first, rather then the new (and good) headers. I'm not sure if it's 3.0.4 or 2.95.3 that's the problem, since I had both installed, I just uninstalled 2.95.3, so I'll find out soon no doubt. *grin*
[04:00:22] <sb-x> /usr/include/usr/include?
[04:00:32] <sb-x> /usr/include/3.1?
[04:01:26] <sb-x> did you use the -V switch?
[04:02:29] <Darke> No -V. Just ran gcc, which defaults to gcc3.1. I'm pretty sure it's just hitting an old header. It's just getting some serious errors when #including any of the C++ library headers.
[04:04:02] <sb-x> my C++ headers are in /usr/include/g++-3
[04:05:49] <Darke> Mine too, then there's a set sitting under i686-something-or-other/ for cpu specific ones.
[04:09:55] --- Darke is now known as Darke|afk
[04:12:45] --- sb-x is now known as sbx|afk
[04:22:53] <-- Darke|afk has left IRC (calvino.openprojects.net irc.openprojects.net)
[04:23:44] --> Darke|afk has joined #exult
[04:23:44] --- ChanServ takes channel operator status from Darke|afk
[04:45:54] --- Darke|afk is now known as Darke
[04:46:51] <sbx|afk> wb
[04:46:54] <sbx|afk> oops
[04:46:56] --- sbx|afk is now known as sbx
[04:46:57] <sbx> wb
[04:48:27] <Darke> Thanks. 'Welcome back' yourself. *grin*
[04:49:19] * sbx is really on the telephone but technically is at the keyboard.
[04:59:46] * Darke got hit with a telephone call, just as he said 'welcome back'. You obviously jinxed him.
[05:01:20] * sbx cackles.
[05:01:38] <sbx> It is probably a sane time of day there for telephone calls. It is midnight here.
[05:21:37] * Darke got _another_ one about 5 minutes after you said that...
[05:34:58] <-- cnj has left IRC (Read error: 110 (Connection timed out))
[06:16:08] * sbx claims to have helped his friend get his cable modem working without doing anything.
[06:18:00] <Darke> BTDTGTTS. Although I tend to have a general 'electronics fix it' aura, rather then cable modem specific. *grin*
[06:19:57] * sbx looks up the acronym. "been there done that got the t-shirt"? :)
[06:20:45] <sbx> Well he said it hasn't been working all day and we talked about the software drivers and then about the modem and then about the software again and then he unplugged the modem and plugged it back in and Win98SE detected it.
[06:20:49] * sbx congratulates himself.
[06:21:34] <sbx> i think it overheated
[06:21:54] <Darke> That's the acronym. *grin*
[06:22:23] <Darke> Actually, it was probably just that the modem needed a reboot. *grin*
[06:23:54] <sbx> Win was getting "Unknown.Device.DECS%" from it
[06:24:04] * Darke 's modem nas a 'reset switch'. *sigh*
[06:24:09] <Darke> s/nas/has/
[06:24:34] <sbx> it did in fact need a reboot, but it was also hotter than it had ever been before
[06:24:49] <sbx> you don't think that would cause it?
[06:25:03] <sbx> i know internal hardware i have had acted that way, but i dont have usb things
[06:25:14] <Darke> Fair enough. *grin* I've never seen a problem with over heated cable modems, at least until they melt anyway, which a friend of mine's did.
[06:25:36] <sbx> :O
[06:26:08] * sbx actually has a USB mouse attached to the PS/2 port.
[06:26:16] <Darke> It got hot enough to melt the plastic casing and it dribbled down over the mounting board and electronics, I think it died just about then. It worked 'perfectly' until that event.
[06:26:25] <sbx> it was causing my floppy drive to act up somehow, and then it stopped
[06:26:42] <Darke> Weird.
[06:27:01] <sbx> I know of that happening with DSL
[06:27:54] <sbx> actually I heard it caught fire, but I wasn't there it is just secondhand knowledge :)
[06:37:05] * sbx creates the format for his server messages.
[06:37:51] * Darke sends a message to format sbx's server.
[06:38:22] * sbx responds with ERR "Invalid protocol in use." and BYE.
[06:38:44] * Darke pouts.
[06:39:22] * Darke clears up a little more hdd space and starts his gcc3.1 compile again. It seems to want about 700Meg as temp working space + compiler source.
[06:39:49] * sbx used to have a 640Meg hdd.
[06:40:06] * Darke has a 128Meg hdd in his firewall.
[06:41:06] * Darke remembers a time long gone when he had a 20Meg hdd the size of a 5.25" drive. A long, long, LONG time ago.
[06:41:10] * sbx asks Darke where he got a 128Meg hdd.
[06:41:36] * sbx doesn't think he has been alive that long.
[06:41:38] <Darke> Out of the i386 that happens to have another 128Meg hdd in it too. Nice little drives.
[06:42:15] <Darke> Bit on the noisy side, but great for a smoothwall installation though.
[06:44:05] <sbx> what is a smoothwall?
[06:50:08] <Darke> A rather nifty firewall, among other things: http://www.smoothwall.org
[06:59:53] --> cnj has joined #exult
[07:30:07] <-- sbx has left IRC (Read error: 113 (No route to host))
[09:28:27] --- Darke is now known as Darke|afk
[10:56:40] <-- Kirben has left IRC ("System Meltdown")
[12:13:51] --> Kirben has joined #exult
[12:13:51] --- ChanServ gives channel operator status to Kirben
[12:46:53] <-- cnj has left IRC (Read error: 110 (Connection timed out))
[12:59:26] --> dmiles has joined #exult
[13:02:31] --> cnj has joined #exult
[14:01:21] --> bj0ern has joined #exult
[14:01:24] <bj0ern> hiho
[14:20:08] --> Fingolfin has joined #exult
[14:20:37] <Fingolfin> hi
[14:20:39] --- ChanServ gives channel operator status to Fingolfin
[14:48:47] <-- bj0ern has left IRC ("[BX] Chester Cheeta uses BitchX. Ayeuhayueuhayueuh!")
[15:58:47] <-- Kirben has left IRC ("System Meltdown")
[16:04:11] --> wjp has joined #exult
[16:04:11] --- ChanServ gives channel operator status to wjp
[16:04:14] <wjp> hi
[16:04:49] <Fingolfin> yo
[16:11:48] --> Colourless has joined #Exult
[16:11:48] --- ChanServ gives channel operator status to Colourless
[16:12:01] <Colourless> hi
[16:12:05] <wjp> hi
[17:24:11] --> bj0ern has joined #exult
[17:24:19] <bj0ern> re
[17:26:16] <wjp> hi
[17:28:34] <Colourless> hello
[17:43:36] --> Dark-Star has joined #exult
[17:53:01] <-- Dark-Star has left IRC (Read error: 104 (Connection reset by peer))
[18:23:18] <bj0ern> oki, just compiled and installed exult on my PowerBook G4 with debian PPC linux.. seems to run well
[18:32:18] <-- Colourless has left IRC ("gone")
[18:50:31] --> robbe has joined #exult
[18:52:08] --- Darke|afk is now known as Darke
[18:54:02] <wjp> bj0ern: cool
[18:54:05] <wjp> robbe: hi
[18:54:10] <wjp> Darke: morning :-)
[18:57:56] <Darke> Hi. *grin*
[19:00:07] <robbe> compiling with g++-3.1 right now ... *yawn*
[19:00:19] * Darke yays! His gcc3.1 installation worked! He just needed to remove gcc3.0.4's /usr/include/g++-v3/ directory.
[19:01:16] <Darke> Weird. Pentagram looks to be compiling quicker with gcc3.1 then gcc3.0.4.
[19:03:04] --> omeros has joined #exult
[19:03:25] <-- omeros has left IRC (SendQ exceeded)
[19:03:28] --> omeros has joined #exult
[19:03:41] <robbe> Anyway, exult is much more g++-3.x friendly than it used to be.
[19:04:02] <wjp> Darke has been keeping it up-to-date :-)
[19:04:39] <omeros> Have any of yuo had any problems sharing yuor work? (ie, your patches work on your own installation of exult, but when you gzip them and send them to another machine -- esp a different platform), exult crashes when it tries to load the game) ?
[19:05:17] <omeros> I've been using rc1-2 from debian unstable to create the game.
[19:06:06] <wjp> hm, I never tried, but the patches I downloaded from CVS all worked ok here
[19:06:46] <robbe> which platforms are involved?
[19:07:05] <omeros> Have people released their own patches into the CVS tree? Perhaps you mean "code patches"...I mean "game patches" to superimpose my own game over u7.
[19:07:16] <wjp> yeah, I mean those too
[19:07:27] <wjp> (check the content/ dir in cvs)
[19:07:29] <omeros> game creted using exult-studio debian unstable (april 15th release I think)
[19:07:44] <robbe> There's the "SourceForge of Virtue Island" patch in CVS
[19:07:51] <omeros> trying to play game using exult on another machine (debian unstable) and also win98.
[19:08:04] <wjp> hm, you're using an exult-studio from april 15?
[19:08:07] <Darke> (gcc3.x) Tristan just updated everything to be gcc3.1 compliant about 3 days ago, so that's more likely to be the reason it's up to scratch, them my sporadic gripes about such trivialities as proper namespaces and casts. *grin*
[19:08:28] <omeros> I think...it's the exult-studio rc1 (0.98) ...
[19:09:17] <omeros> is there any special way to "export the games?
[19:09:46] <wjp> dunno, I haven't really looked into exult studio
[19:09:46] <robbe> omeros: so byte-sex difference can be ruled out
[19:10:00] <wjp> robbe: lol :-)
[19:10:08] <omeros> I just save the map and zip the dir, then move it to another machin, unzip, set new path and path dirs, and try to run it.
[19:10:26] <robbe> omeros: one thing you could try is test the patch from CVS's content dir on your different boxes
[19:10:33] <wjp> omeros: I'd upgrade to a recent exult-studio and exult snapshot
[19:10:51] <omeros> good idea. I haven't tried building the source from cvs yet.
[19:10:52] <bj0ern> there is a g++ 3.1? since when is it out?
[19:10:59] <wjp> it might not necessarily fix things, but exult studio has improved a lot over the last month
[19:11:08] <wjp> bj0ern: around a week or two I think
[19:11:13] <bj0ern> cool
[19:11:24] <robbe> omeros: I could provide .debs for CVS from yesterday, if yer lazy.
[19:11:26] <wjp> still no PCH though :-(
[19:11:36] <bj0ern> we should use that at work, ill have to ask if my colleague already upgraded
[19:13:37] * Darke suggests 3.0.4 for 'real' work unless you're keeping an eye on the gcc bug list. *grin* There's quite a few bugs popping up when optimisations are turned on, that gcc3.0.4 doesn't have.
[19:13:59] <omeros> I am lazzzy...been working on the game more than wanting to compile...I do too much compiling at work... :)
[19:14:12] <omeros> robbe: That would be great.
[19:14:23] <robbe> omeros: give me a few mins
[19:19:24] <omeros> roobe: if you could e-mail me a URL, I would be most appreciative...for i must run for the moment...
[19:19:33] <omeros> robbe, i mean...
[19:19:41] <omeros> thanks for your help today all...
[19:19:46] <-- omeros has left IRC ("BitchX: more nutritious than a six-pack.")
[19:20:02] --> omeros has joined #exult
[19:20:17] <omeros> I just realized I didn't leave a url...omeros@haushilderbrand.com
[19:20:25] <-- omeros has left IRC (Client Quit)
[19:37:44] <-- cnj has left IRC (Read error: 110 (Connection timed out))
[19:46:27] <-- Darke has left IRC (Remote closed the connection)
[19:48:47] <-- bj0ern has left IRC ("changing servers")
[19:51:10] --> bj0ern has joined #exult
[19:57:09] --> Darke has joined #exult
[19:57:09] --- ChanServ gives channel operator status to Darke
[19:57:59] * Darke bows. 'ello again.
[19:58:18] <wjp> wb
[20:15:44] <bj0ern> =)
[20:33:35] --> Dark-Star has joined #exult
[20:36:59] <wjp> hi
[20:37:24] <Dark-Star> hi
[20:38:05] <Dark-Star> how's progress on pentagram?
[20:38:22] <wjp> not much work on it the last week
[20:39:35] <-- robbe has left IRC (Remote closed the connection)
[20:40:12] * Darke has a decompiler for u8usecode that doesn't work, and doesn't work badly. So he's rewriting it. *grin* Spending 8 hours to code something then immediately spending 3 times that many trying to debug it, kind of makes him think that he should write it 'properly' to begin with now he knows what he's doing.
[20:44:22] <Dark-Star> Darke: what do you mean it doesn't work? It did the last time I tried it...
[20:45:48] <Darke> The disassembler for u8usecode works, the decompiler I'm writing inside it doesn't. *grin*
[20:47:51] <Dark-Star> oh I misunderstood you then ... thought you mean the disassembler :)
[20:48:16] * Darke is not sure if his problem is that he's either trying to do something that has an 'undefined' effect according to the C++ standard, or that it's a gcc bug. There's a report on the gcc-bugs list that suggests the latter, but he still thinks it's just probably his fault. *grin*
[20:48:59] <Darke> No problem. *grin*
[20:49:24] * Darke regularly says 'disassembler' rather then 'decompiler' anyway confusing himself. *grin*
[20:49:34] * Dark-Star wonders why Darke would do something that has an "undefined" effect...
[20:50:35] * Dark-Star thinks that would be rather unproductive / illogical
[20:51:24] * Darke was calling an overloaded pure virtual function in a subclass, and this function was calling a function in the baseclass, which then called a function in the same derived class. It's exceedingly clean, but he's not sure if it's legal due to the fact he was doing it all inside the subclasses constructor, so the subclass might not be completely constructed...
[20:52:01] <Darke> Pefectly logic, since _I_ think it's legal, although the standard might say otherwise. *grin*
[20:52:08] <Darke> s/logic/logical/
[20:53:08] <Dark-Star> yes, I think that *should* work... but then, I'm no C++ expert myself
[20:55:21] <Darke> Bingo. *grin* My exact problem. It should work, but I've had a number of C++ texts in the past tell me that calling back and foward between virtual-overridden and non-virtual-baseclass functions in a constructor is something that should be avoided because they had apparently encountered compilers that don't like it. *grin*
[20:56:19] <Darke> In any event, I'm just stripping 90% of the code out of my original one and starting from almost scratch designing it to avoid that.
[21:04:09] * wjp is going to play some Geneforge
[21:04:13] <wjp> bbl
[21:04:14] --> cnj has joined #exult
[21:04:19] <-- wjp has left IRC ()
[21:20:48] <Fingolfin> Dark-Star: well using virtual functions in the constructor may not work at all like you'd expect it to
[21:21:05] <Fingolfin> Darke: even if you compiler likes it (it should like it if it wants to comply to the C++ standard)
[21:21:27] <Fingolfin> If A->B->C (i.e. A is parent of B is parent of B), and all provide the virtual method foo()
[21:22:00] <Fingolfin> then if you call foo() from an object of class C, it will always call teh "right" one, i.e. the one implemented from class C, even in the method bar() that is only implemented in A, and inherited by B and C
[21:22:31] <Fingolfin> but if the constructor of B is calling foo(), it will *always* call the foo() as defined in B, never the one of C, even if the object being constcuted is of class C
[21:29:19] * Darke ahhs. That looks like the problem, his 'B' class constructor was trying to call a pure virtual function and getting something completely different.
[21:30:13] <Darke> What was a pure virtual function in 'B' anyway, I'd overridden it in 'C'.
[21:49:11] --> sbxtreme has joined #exult
[21:49:13] --- sbxtreme is now known as sbx
[21:49:14] <sbx> hi
[21:53:53] <bj0ern> hi
[21:53:57] <Darke> Hi.
[22:18:40] * Darke hops off to handle a SIGREALITY.
[22:18:44] --- Darke is now known as Darke|afk
[22:19:23] <-- Dark-Star has left #exult ()
[22:21:37] --> es_bee-ex has joined #exult
[22:22:02] <-- sbx has left IRC (Killed (NickServ (Ghost: es_bee-ex!~sbx@
[22:46:56] <-- es_bee-ex has left IRC (Read error: 110 (Connection timed out))
[23:10:14] --> wjp has joined #exult
[23:10:14] --- ChanServ gives channel operator status to wjp
[23:10:18] <wjp> hi again