#exult@irc.freenode.net logs for 28 Jul 2002 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage


[00:07:52] --- Darke|afk is now known as Darke
[00:08:16] <wjp> morning
[00:08:29] * Darke bows. Morning.
[00:09:00] <Darke> Wow. A rather... exciting night was it not? I can still see my |afk on the screen. *grin*
[00:09:40] <wjp> hm, I wasn't really here most of the time
[00:09:48] <wjp> installing windows *shudder*
[00:11:05] --> SB-X has joined #exult
[00:11:17] <SB-X> Greetings all.
[00:11:53] <Darke> Hi.
[00:12:12] <Darke> wjp: I'm not sure if I should say "My sympathies" or cackle evily. *grin*
[00:12:51] <wjp> how about cackling evily while offering your sympathies? :-)
[00:12:54] <wjp> hi sbx
[00:15:17] <Darke> I so very, very much *evilgiggle* feel your excrutiating *evillaugh* pain at such an arduous *evilcackle* task.
[00:21:59] <SB-X> my hard disk had an Uncorrectable Error when fsck ran on it today-is that a bad thing?
[00:22:31] <wjp> hm, sounds bad
[00:22:45] <SB-X> DataRequest Error
[00:22:53] <SB-X> but then the third time i did fsck it didnt get it
[00:23:01] <SB-X> and i couldnt get any bad blocks detected
[00:26:35] <wjp> hm
[00:27:25] <wjp> what filesystem?
[00:27:45] <SB-X> fsck
[00:27:47] <wjp> oh, wait, DataRequest error is a lower level error than that
[00:27:48] <SB-X> e2fs
[00:28:15] <SB-X> sounds like it to me :)
[00:28:41] <wjp> hm, are you sure it was a 'DataRequest Error' ?
[00:28:54] <wjp> not that the two were just separate words?
[00:29:15] <SB-X> they were seperate
[00:29:16] <SB-X> hda: read_intr: status=0x59 { DriveReady SeekComplete DataRequest Error }
[00:29:16] <SB-X> hda: read_intr: error=0x40 { UncorrectableError }, LBAsect=22282882, sector=664
[00:29:16] <SB-X> end_request: I/O error, dev 03:07 (hda), sector 664
[00:29:33] * wjp nods
[00:30:15] <wjp> you got it multiple times, right? all of them the same sector?
[00:30:36] <SB-X> i got it twice
[00:30:49] <SB-X> sector 662 then 664
[00:31:56] <wjp> hm, can a HD relocate sectors transparently?
[00:32:25] <SB-X> heh
[00:32:27] <SB-X> don't look at me
[00:32:37] * SB-X looks at Darke.
[00:32:40] <wjp> try 'smartctl -a /dev/hda' (or whatever device)
[00:33:26] <Darke> wjp: That's what that SMART drive feature of hdds is supposed to do.
[00:33:46] <wjp> yes, that's why I asked what 'smartctl' returns ;-)
[00:33:54] <SB-X> I have no such program.
[00:34:12] <SB-X> SMART drive?
[00:34:18] <wjp> hm, it's in the 'kernel-utils' package for me, btw (RH)
[00:34:28] <wjp> yeah, it's an acronym for something
[00:34:44] <wjp> some kind of self-diagnosis/repair thingie or something
[00:34:50] <Darke> Yep.
[00:35:22] <wjp> Self-Monitoring, Analysis and Reporting Technology
[00:35:27] <SB-X> i thought it was a DOS program :)
[00:35:30] <SB-X> oh that was EZ Drive
[00:35:34] <SB-X> okay
[00:35:51] <Darke> http://www.pc.ibm.com/us/infobrf/ibsmart.html
[00:36:29] <wjp> hm, my HD has apparently been up for 1115 hours, in 67 'runs'
[00:36:53] <wjp> sounds about right; I got it early june
[00:37:09] <SB-X> how did you check that?
[00:37:14] <wjp> with smartctl :-)
[00:37:59] <SB-X> An operating system that allows S.M.A.R.T. commands to pass through, such as:
[00:37:59] <SB-X> Windows ® 3.1
[00:37:59] <SB-X> Windows® 95
[00:37:59] <SB-X> Windows® 98
[00:37:59] <SB-X> Windows NT® 4.0
[00:38:00] <SB-X> OS/2® (requires a bus master device driver with hooks to the S.M.A.R.T. information)
[00:44:21] <wjp> any luck in locating smartctl?
[00:45:08] <SB-X> this page says i need a SMART capable hard disk drive
[00:45:19] <wjp> most modern HDs are
[00:46:52] <wjp> hm, this WD 2.5Gb drive doesn't support it, but an IBM 6Gb disk does
[00:47:25] <wjp> hmm.. reallocated sector count is 11 on that IBM disk
[00:48:47] <wjp> I should be going
[00:48:48] <wjp> g'night
[00:48:50] <SB-X> thanks
[00:48:52] <SB-X> night
[00:48:56] <-- wjp has left IRC ("Zzzz...")
[00:49:23] <SB-X> Darke: is SMART just for IBM drives?
[00:50:00] <Darke> I don't think so. They were the first to install it in their HDDs though, and they market it clearly on them. *grin*
[00:57:06] * SB-X finds the source for smartsuite 2.1.
[01:00:13] <Darke> There's also 'ide-smart' here: http://lightside.eresmas.com/ I've never tried either though, I'm just emerging them. *grin*
[01:00:34] <SB-X> why was i supposed to run smartctl anyway? i forgot
[01:04:06] <SB-X> yay
[01:04:14] <SB-X> Device: WDC WD300AB-00BVA0 Supports ATA Version 5
[01:04:14] <SB-X> Drive supports S.M.A.R.T. and is enabled
[01:05:22] <SB-X> *grumble* there were 22 ATA errors
[01:05:29] <SB-X> i don't know what that means
[01:05:31] * SB-X shrugs.
[01:06:49] <SB-X> Darke: emerged yet?
[01:07:25] * Darke nods. He's been playing with it for quite a while. *grin*
[01:08:05] <Darke> I don't appear to have any ATA errors on my drives.
[01:17:57] * SB-X grrs at the moongates.
[01:19:00] <SB-X> ive tried this 6 times and cant get to new magincia
[01:20:27] <Darke> Why not go by ship?
[01:21:10] <SB-X> now i keep getting knocked back, spilling blood
[01:21:22] <SB-X> I thought this would be _faster_.
[01:21:23] <SB-X> :-)
[01:22:22] --> Kirben has joined #exult
[01:22:22] --- ChanServ gives channel operator status to Kirben
[02:57:05] --- SB-X is now known as sb-x|a-fk
[02:57:14] <sb-x|a-fk> bbl
[02:57:34] * Darke pawwaves.
[05:02:33] --> Colourless has joined #Exult
[05:02:33] --- ChanServ gives channel operator status to Colourless
[05:03:14] <Colourless> hi
[05:03:40] <Darke> Hi.
[05:03:51] <Colourless> someone really fucked up the pathfinding it seems :-)
[05:04:24] <Colourless> the non hash search didn't bother to check the 'z' coord between nodes
[05:12:01] * Darke snickers. That could cause just a _few_ problems. *grin*
[05:13:20] <Colourless> and I wondered why pathfinding was real slow when moving up and down stairs :-)
[06:05:14] --- sb-x|a-fk is now known as sb-x
[06:05:14] <-- Colourless has left IRC (Read error: 104 (Connection reset by peer))
[06:05:18] <sb-x> hello
[06:05:21] <sb-x> hey he left
[06:05:54] <Darke> Hello again. *grin*
[06:06:14] * sb-x bows.
[08:04:26] <-- sb-x has left IRC (Remote closed the connection)
[08:05:47] --> SB-X has joined #exult
[08:05:58] <Darke> Welcome back.
[08:06:39] <SB-X> re
[08:06:45] * SB-X thanks Darke.
[10:32:09] --> Fingolfin has joined #exult
[10:32:16] <Fingolfin> yo
[10:32:18] --- ChanServ gives channel operator status to Fingolfin
[10:32:20] <SB-X> yo
[10:32:45] <Darke> Hi.
[11:09:18] * SB-X goes to sleep.
[11:09:20] <-- SB-X has left IRC ("Ultima! Ultima! Ultima!")
[11:45:57] --> wjp has joined #exult
[11:45:57] --- ChanServ gives channel operator status to wjp
[11:46:01] <wjp> hi
[11:46:07] <Darke> Hi.
[11:48:50] <wjp> hmm... interesting... tripwire picked up a changed libc-2.2.4.so on one of our machines
[11:54:19] * wjp wonders how to handle that
[11:56:36] <wjp> hmm... you know, I really don't like that 'forgot password' feature of phorum
[11:56:46] <wjp> it means anybody can reset my password
[11:58:36] <wjp> and that password is not-too-hard to predict too... it's the first 8 characters of the md5sum of the current date (in millisecs I think)
[12:19:30] <wjp> hm, only a single bit is different in that libc :-)
[12:27:26] * Darke snickers. That sounds odd, must be disk corruption or something.
[12:28:15] <wjp> yeah :/
[12:28:20] <wjp> time to get out smartctl again :-)
[12:29:21] * Darke grins. Timely.
[12:30:36] <wjp> hm, doesn't seem to be installed here; pity
[12:45:50] * Darke yawns and declares it 'nap time'. Night!
[12:46:03] --- Darke is now known as Darke|afk
[12:46:19] <wjp> night
[13:50:43] --> Cless has joined #Exult
[13:50:56] <Cless> hi
[13:51:10] <wjp> hi
[13:51:11] --- Cless is now known as Colourless
[13:51:13] --- ChanServ gives channel operator status to Colourless
[13:51:15] <wjp> nice catch on the pathfinding :-)
[13:52:26] <Colourless> it kind of became really obvious something was terribly wrong when i added lots and lots of debugging output to the pathfinder
[13:53:00] <Colourless> I noticed something strange going on, the pathfinder would say it's already been to a node, that I knew it hadn't already been to
[13:54:12] <Colourless> i always thought there was something screwy going on with the pathfinder and stairs
[13:54:54] <Colourless> countless times i've experienced unreasonable slow downs when going up or down stairs. i'm now guessing they were caused by the problem i found
[13:55:31] <Colourless> since it only effect people without hash_set, no one would have normally noticed. pretty much i was the only person to ever use that option :-)
[13:56:29] <Colourless> however since rc2 (which i made, without hash_set obviously), kirben has been compiling snapshots with mingw and gcc3.1, which afaik, also doesn't have hash_set. so, with the amount of users around, the problem showed itself real quick
[13:56:40] * wjp nods
[13:57:07] <wjp> sounds like this was it, then :-)
[13:57:17] <Colourless> i'm guessing so
[13:57:38] <Kirben> hash_set ? you mean HAVE_EXT_HASH_SET ?
[13:59:40] <Colourless> yeah
[14:00:12] <Colourless> ah, so you do have it
[14:00:48] <Kirben> yes used manual defines since I noticed it was available.
[14:01:35] <Colourless> the hash_set code had some slightly dubious code as well, which i changed. it appeared that when it was making the hashes it was doing hash = x << 16 | y << 8 | z; which didn't seem quite right. x and y both use 12 bits in exult, while the hash it would see, assumed they were only 8 bits
[14:01:59] <Kirben> not sure why it isn't auto detected under win32 though.
[14:02:01] <wjp> hmm
[14:02:14] <wjp> that could cause some interesting problems in some places
[14:03:49] <wjp> although, no... it would just make it slower
[14:04:11] <wjp> just the hash of course, not the key... silly me :-)
[14:05:01] <Colourless> the comparison class for hash_set had no problems.
[14:08:36] <wjp> Kirben: hm, so recent snapshots did use the hash_set?
[14:09:57] <Kirben> Yes, win32 makefiles have used -DHAVE_EXT_HASH_MAP and -DHAVE_EXT_HASH_SET since I switched to gcc 3.1.
[14:10:35] --> Cless has joined #Exult
[14:10:49] <-- Colourless has left IRC (Read error: 104 (Connection reset by peer))
[14:10:52] --- Cless is now known as Colourless
[14:10:58] --- ChanServ gives channel operator status to Colourless
[14:11:24] <Kirben> so last 6 weeks or so of snapshots.
[14:24:23] <-- Kirben has left IRC ("System Meltdown")
[16:09:48] <-- Fingolfin has left IRC ("bbl")
[16:24:35] <wjp> time for me to go home; I'll bbl
[16:24:38] <-- wjp has left IRC ("[x]chat")
[18:16:11] --> wjp has joined #exult
[18:16:11] --- ChanServ gives channel operator status to wjp
[20:00:32] --> Fingolfin has joined #exult
[20:00:40] <Fingolfin> yo
[20:00:42] --- ChanServ gives channel operator status to Fingolfin
[20:00:50] <Colourless> hi
[20:19:17] <-- Colourless has left IRC ("time for me to go")
[21:14:00] <-- Fingolfin has left IRC ("42")
[21:22:11] --> vividos has joined #exult
[21:22:20] <vividos> hi
[21:22:51] <wjp> hi
[21:25:39] <vividos> how's exult going to be lately? any new features?
[21:25:55] <wjp> hm, not really; mainly bugfixes
[21:26:13] <wjp> although, artaxerxes is porting exult to the Zaurus
[21:26:18] <wjp> (linux-based PDA)
[21:26:43] <vividos> I already saw it, it's marvelous!
[21:28:48] <wjp> I was trying to compile uwadv, btw
[21:29:00] <wjp> (but I got sidetracked with scummvm audio problems, in the meantime)
[21:29:18] <vividos> hehe
[21:29:31] <vividos> current uwadv cvs isn't linux friendly, btw :)
[21:29:37] <wjp> I noticed :-)
[21:30:15] <vividos> in the next days there is a new release, I then update the makefiles
[21:30:35] <vividos> Fingolfin was already successful to build it on MacOS X :)
[21:32:46] <vividos> what was the scummvm audio problems?
[21:32:59] <wjp> distorted speech in DOTT
[21:34:25] <vividos> and what was the reason?
[21:34:51] <wjp> don't know yet :-)
[21:36:09] <vividos> ah ok :)
[21:36:14] <vividos> using SDL_mixer?
[21:36:33] <wjp> hm, no, not AFAIK
[21:41:29] <wjp> *phew*.. ok, got all missing cpp files :-)
[21:42:34] <vividos> what project, uwadv?
[21:42:37] <wjp> yeah
[21:42:49] <vividos> it could even compile :)
[21:43:12] <wjp> I had to get rid of the std::vector.at() calls, and some creative makefile hacking :-)
[21:43:48] <wjp> ok, origin logo :-)
[21:44:16] <vividos> yay!
[21:44:22] <wjp> hm, getting 100 fp/s
[21:44:26] <wjp> fps, even
[21:44:30] <wjp> so, how do I move? :-)
[21:44:47] <vividos> a = turn left, d = turn right, w = move forward
[21:44:55] <wjp> any way to increase light level?
[21:45:03] <vividos> do you see the 3d view?
[21:45:07] <wjp> yeah
[21:45:12] <vividos> yes, it already should be brighter
[21:45:59] <vividos> in ingame_orig.cpp, search the call to glFogf(GL_FOG_DENSITY,0.2f). the lower the value, the brighter the level
[21:46:22] <vividos> or just get rid of all calls to glEnable(GL_FOG)
[21:47:21] <wjp> ok, much better :-)
[21:47:35] <wjp> looks very good :-)
[21:47:54] <vividos> how many frames per second?
[21:48:10] <wjp> about 100
[21:48:14] <vividos> ingame?
[21:48:16] <wjp> yeah
[21:48:18] <vividos> wow!
[21:48:26] <wjp> tnt2ultra
[21:48:40] <vividos> but a rather fast processor I guess
[21:48:45] <wjp> athlon 1800 xp
[21:49:19] <vividos> in debug mode, I rarely get 15 fps
[21:49:39] <wjp> how do I get into debug mode?
[21:50:01] <vividos> using msvc :) or simply omit the -O2 switch
[21:50:18] <wjp> -O2?
[21:50:23] * wjp doesn't see that
[21:50:57] <vividos> should be somewhere in the makefile
[21:51:06] <wjp> I mean in the compile lines
[21:51:20] * Darke|afk wonders why that'd have any effect. He thought that only altered the optimiastion level.
[21:51:42] <wjp> Darke: optimization level affects speed, you know ;-)
[21:52:10] <vividos> most 2d user interface elements are converted from indexed to truecolor at render time
[21:52:53] * wjp recompiles with -O2
[21:53:46] <wjp> 330 fps in the intro
[21:53:54] <wjp> 200 in the main menu
[21:53:55] <Darke|afk> wjp: Well, yes. But unless the only 'feature' of the debug mode is going at 15fps, then I would guess that the removal of -O2 would have no effect on turning 'on' or 'off' it. *grin*
[21:54:15] <wjp> between 110 and 140 ingame
[21:54:57] <wjp> mostly around 110, though
[21:55:14] <vividos> not bad :) if just everyone had such a system :)
[21:55:17] <Darke|afk> Try -O3 and see how badly things break. *grin*
[21:55:25] <vividos> the introduction is also working
[21:55:28] <wjp> not with gcc-2.96, thanks :-)
[21:56:23] <wjp> hm, crash when I try to run the intro
[21:56:25] * Darke|afk resists the urge to laugh evilly.
[21:56:43] <vividos> were any messages on the console?
[21:56:50] <vividos> or in the stdout.txt?
[21:56:53] <wjp> oh, the uadata path needs a trailing slash
[21:56:54] <wjp> works now
[21:58:04] <wjp> should there be speech?
[22:00:07] <vividos> yes :) just add a line to uwadv.cfg that contains "cutscene_narration both"
[22:00:19] <vividos> the file wasn't updated yet
[22:00:49] <wjp> yay, nice :-)
[22:00:52] <wjp> works :-)
[22:01:27] * vividos shoves some bundles of money to wjp
[22:01:35] <wjp> huh?
[22:02:11] <vividos> :)
[22:04:09] <wjp> btw, apart from adding lots of files to the Makefile, you need to add a "CFLAGS = -Iresource/zziplib $(SDL_CFLAGS)" and '-lSDL_mixer -lz' to the LDADD
[22:04:28] * Darke|afk really afks now. He's off to work. *grin*
[22:04:33] <wjp> bye
[22:04:35] <vividos> yes, thanks
[22:04:37] <vividos> bye Darke
[22:05:38] <wjp> and unfortunately the at() calls cause problems here :-(
[22:06:09] <wjp> oh, and audio.cpp includes sdl_mixer.h, instead of SDL_mixer.h
[22:06:19] <wjp> (I even submitted a bug report for that last one :-) )
[22:06:47] <vividos> ok, I remove the at() calls.
[22:38:44] <wjp> yay, found the audio problem in scummvm :-)
[22:39:39] <vividos> what was the reason?
[22:39:57] <wjp> one part of the code thought the volume scale was 0-100, another part thought it was 0-255 :-)
[22:40:13] <wjp> this caused some rather serious clipping
[22:43:00] <vividos> :) do you use gdb to debug all that?
[22:43:15] <wjp> yeah, I used gdb a bit
[22:44:06] <vividos> are there some visual tools that can be used to debug? a frontend to gdb?
[22:44:16] <wjp> hm, yes, they exist
[22:44:24] <wjp> haven't used any, though
[22:52:35] <vividos> I bet the exult inventory was hard to do
[22:53:01] <wjp> in what way?
[22:53:18] <vividos> as there are no real "slots" where items could be dropped
[22:53:38] <vividos> and determining which object the user wants to take
[22:53:55] <wjp> oh, that...
[22:54:10] <wjp> yes, and it's still not flawless :-)
[22:54:56] <vividos> I guess the logic for that has to be reinvented, yes?
[22:55:06] <wjp> huge lookup-tables, mostly :-)
[22:55:35] <vividos> uw1's slotted inventory is nicer, though
[23:16:08] <wjp> I should be going
[23:16:08] <wjp> g'night
[23:16:23] <-- wjp has left IRC ("Zzzz...")
[23:18:52] <vividos> should be going, too. night!
[23:19:02] <-- vividos has left IRC ("Leaving")