#exult@irc.freenode.net logs for 30 Jul 2002 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage

[05:05:12] <-- Darke|afk has left IRC (calvino.openprojects.net irc.openprojects.net)
[05:34:04] <-- Kirben has left IRC (calvino.openprojects.net irc.openprojects.net)
[05:34:57] --> Kirben has joined #exult
[06:37:26] <-- Kirben has left IRC (calvino.openprojects.net irc.openprojects.net)
[06:38:09] --> Kirben has joined #exult
[08:14:33] --> Darke|afk has joined #exult
[08:14:49] --- Darke|afk is now known as Darke
[08:20:26] --> wjp has joined #exult
[08:20:27] --- ChanServ gives channel operator status to wjp
[08:20:30] <wjp> hi
[08:23:28] <Darke> Hi.
[08:39:59] --> sbxult has joined #exult
[08:40:04] <sbxult> hello
[08:40:07] <wjp> hi
[08:43:47] <sbxult> wjp: i downloaded smartsuite, what do i do with smartctl?
[08:50:12] <wjp> run it?
[08:52:34] <sbxult> i don't know what to do with the information
[09:03:54] <wjp> hm, there should be a "relocated sector count" or something
[09:04:03] <wjp> (and some other fields which indicate errors)
[09:14:04] <sbxult> yeah
[09:14:08] <sbxult> under Vendor Specific SMART Attributes with Thresholds:
[09:14:14] <sbxult> ( 5)Reallocated Sector Ct 0x0032 200 200 112 0
[09:14:34] <sbxult> fields are Attribute Flag Value Worst Threshold Raw Value
[09:15:03] <sbxult> it also says "ATA Error Count: 22"
[10:15:58] * sbxult goes to sleep.
[10:16:00] <sbxult> see you later
[10:16:04] <-- sbxult has left IRC ("Ultima! Ultima! Ultima!")
[11:23:23] <wjp> http://escherdroste.math.leidenuniv.nl/ <-- interesting site if you like Escher's art
[11:33:49] * Darke ooohs. Nice!
[11:34:12] * Darke likes Escher's 'grid', it's almost worthy of hanging on a wall by itself.
[11:34:52] <wjp> site is getting a couple of hundred hits per minute at the moment
[11:35:27] * wjp feels sorry for our webserver
[11:35:56] <wjp> (it got coverage from the NYtimes, btw)
[11:36:40] * Darke ahhs. That explains it. Wait until the slashdot effect hits it. *grin*
[11:36:54] <wjp> well, somebody here _did_ submit it to /.
[11:37:16] <wjp> I can't wait until people in the US actually wake up :-)
[11:37:40] * Darke snickers. It'll probably be hammered.
[11:38:12] <wjp> ah well, at least this'll put our gigabit uplink to good use for a change :-)
[11:38:56] * Darke snickers. That's one 'benefit' then. *grin*
[11:40:36] * Darke yays! They've got mpegs! Woo!
[11:41:16] * Darke patpats the webserver and decides not to be nasty and download them now. He'll wait 'til it cools off a bit in a day or so.
[11:42:20] <wjp> go ahead and download them now if you want
[11:42:52] <Darke> Nah. I need to sleep, not be mesmerised looking at infinate looping escher images. *grin*
[11:42:58] <wjp> hehe :-)
[11:57:06] * Darke yawns and curls up. Sleeptime. Night!
[11:57:18] <wjp> night
[11:57:21] --- Darke is now known as Darke|afk
[12:18:25] <wjp> time for me to go home
[12:18:53] <wjp> LOL... it got accepted on /.!
[12:19:59] <wjp> three hundred hits per second
[12:20:00] <wjp> yay :=_
[12:20:02] <wjp> :-)
[12:22:41] <wjp> LOL
[12:22:46] <wjp> this is great :-)
[12:22:53] <wjp> poor webserver :-)
[12:23:12] <wjp> 4000 connections to our webserver currently :-)
[12:39:21] <wjp> k, time to go now
[12:39:22] <-- wjp has left IRC ("[x]chat")
[13:06:44] --> wjp has joined #exult
[13:06:44] --- ChanServ gives channel operator status to wjp
[13:33:46] --> Colourless has joined #Exult
[13:33:46] --- ChanServ gives channel operator status to Colourless
[13:33:54] <wjp> hi
[13:33:56] <Colourless> ih
[13:33:58] <Colourless> hi
[13:34:06] <wjp> my uni's webserver is being slashdotted! yay! :-)
[13:35:01] <wjp> log file is growing at around a megabyte per minute
[13:35:05] <Colourless> why?
[13:35:05] <wjp> >300 hits/second
[13:35:18] <wjp> second story on /.
[13:35:37] <wjp> load is at 60, but the machine is still alive :-)
[13:36:01] <Colourless> 60! what sort of machine is it?
[13:36:09] <wjp> medium P2 :-)
[13:36:21] * wjp checks
[13:36:51] <Colourless> i see.... :-)
[13:37:08] <wjp> oh, high-end P2, even. (450)
[13:37:56] <wjp> only 128Mb RAM, though
[13:38:09] <Colourless> ow
[13:38:10] <wjp> luckily most of the pages requested fit in there, it seems
[13:38:42] <Colourless> that computer is lower than what's currently called low end
[13:39:57] <Colourless> i'm guessing now is not the time to check exultbot logs
[13:40:12] <wjp> nah, doesn't matter
[13:40:17] <wjp> different server
[14:28:35] <-- Kirben has left IRC ("System Meltdown")
[16:23:11] --> artaxerxes has joined #exult
[16:23:11] --- ChanServ gives channel operator status to artaxerxes
[16:23:20] <artaxerxes> hi all
[16:23:25] <Colourless> hi
[16:24:10] <artaxerxes> I've made a patch to clear the event queue just before the SI or BG menu appears.. should I commit it?
[16:24:56] <artaxerxes> it was an itch that I scratched... :)
[16:26:04] <Colourless> um, i don't know
[16:27:33] <artaxerxes> don't you find it annoying to see your mouse cursor going crazy on those screens if you happen to move your mouse during loading?
[16:28:08] <Colourless> can't say i've ever had a problem
[16:28:28] <artaxerxes> either you don't move or you don't use your mouse during loading then.
[16:29:58] <Colourless> i move the mouse, i get nothing.
[16:32:05] <artaxerxes> between the time you clic on "ULTIMA VII PART 2 SERPENT ISLE" and the actual Serpent Isle menu, move your mouse around. You'll see that as soon as the menu appears, the cursor follow wherever your mouse has been.
[16:32:37] <Colourless> nope, the mouse appears where it's supposed to be for me if i do that
[16:34:35] <artaxerxes> did you already modified the source?
[16:34:37] <artaxerxes> :)
[16:35:19] <Colourless> no
[16:35:56] <artaxerxes> weird
[16:36:25] <Colourless> unless the menuing system, or sdl even, is doing something stupid, each mouse event should be handled before the redrawing is done
[16:36:43] <Colourless> if only a single event was handled, i would get 'mouse lag' there too
[16:38:36] <Colourless> but of course exult is doing exactly that
[16:38:57] <Colourless> your change will not fix the true problem it will just hide it :-)
[16:39:36] <artaxerxes> it seems that when the BG or SI menu appear, the event queue is processed. However, it is filled with all the events that happened since you clicked on "ULTIMA VII PART x bla bla bla"
[16:39:47] <Colourless> hmm, then again, maybe exult is doing it right...
[16:40:04] <artaxerxes> which is wrong.
[16:40:50] <Colourless> no, exult is definately doing it wrong
[16:41:04] <artaxerxes> that's what I am saying! :)
[16:41:30] <artaxerxes> ignore events until the menu has fully appeared.
[16:41:34] <artaxerxes> that should be ok
[16:42:23] <Colourless> actually, i am unsure if it's doing it right or wrong. part of me says wrong, different part say right
[16:42:48] <Colourless> unconditionally ignoring events isn't always a wise idea. thing can occur that you may want to know about
[16:43:06] <artaxerxes> my change empties the event queue, but you are right, the real fix would be to ignore events until the menu has appeared entirely (Except if playing intro)
[16:43:38] <artaxerxes> you are right
[16:43:43] <artaxerxes> like WM events
[16:44:01] <Colourless> yeah
[16:44:02] <artaxerxes> then just ignoring mouse events
[16:45:27] <Colourless> even then, it may not be good. if the mouse has been moved, but yet you don't move it again after the menu has come up, the cursor on screen will not actually be where the cursor actually is
[16:45:57] <Colourless> pressing a button, without moving the mouse could have unexpected results
[16:46:00] <Colourless> that = bad
[16:46:57] <artaxerxes> and indeed, the cursor gives the feeling of jumping. Adding a warp maybe?
[16:49:46] <Colourless> no, not reallyy. the best option would be to change MenuList::handle_events in menulist.cc to handle all mouse events, without causing a screen repaint. why it would be causing a screen repatint, i am yet to determine
[16:51:37] <artaxerxes> there is a mouse->blitdirty() in it
[16:51:53] <Colourless> yeah, that's the problem
[16:52:00] <Colourless> just checked it out myself
[16:52:25] <Colourless> what you want to do, is remove the blitdirty() call to outside the eventloop
[16:52:40] <Colourless> that also has certain problems
[16:53:08] <Colourless> the eventloop itself, only leaves once it's encounted a key press
[16:53:49] <Colourless> what you want to do it, only update the mouse once outside the loop, or when there are no more events
[16:55:38] <Colourless> you'll then want to change it from using SDL_WaitEvent to SDL_PollEvent and maybe a SDL_Delay
[17:02:35] <artaxerxes> could you rephrase what you said ?
[17:04:34] <Colourless> well basically you'd want to change the SDL_WaitEvent(&event); line to something like
[17:05:21] <Colourless> while(!SDL_PollEvent(&event)) { if (mouse_needs_blit) mouse->blit_dirty(); SDL_Delay(10); }
[17:05:52] <artaxerxes> but is what is it going to prevent the mouse from following old commands?
[17:05:54] <Colourless> then instead of blitting the mouse in event.type==SDL_MOUSEMOTION, you'd set a bool mouse_needs_blit to true, and so on
[17:08:14] <Colourless> note that you'll need set mouse_needs_blit to false after it's been blitted. you'll also need to add a if (mouse_needs_blit) mouse->blit_dirty(); to just before the mouse->hide() at the end of the function..
[17:08:36] <Colourless> mouse_needs_blit would obviously be set to false by default...
[17:08:53] <Colourless> as far as I can tell, that 'should' fix your problem
[17:09:19] <Colourless> i might be mistaken though
[17:09:38] <Colourless> the code above that, which repaints everything, may also cause a problem
[17:10:10] <artaxerxes> :)
[17:10:49] <artaxerxes> the thing is that the mouse will still 'jump' as soon as you will move.
[17:11:06] <Colourless> it shouldn't
[17:11:10] <artaxerxes> from where the graph is to where the mouse is recorded to be
[17:12:01] <Colourless> is mouse motion inside exult itself a problem?
[17:12:11] <Colourless> it's the way the events are handled in game
[17:12:57] <artaxerxes> inside exult, you mean during a game? I think things are generally ok
[17:13:10] <Colourless> yes
[17:13:28] <Colourless> all of the pending events are handled before repainting the screen
[17:13:46] <artaxerxes> ahhhh
[17:13:48] <artaxerxes> I understand
[17:15:04] <artaxerxes> say you actually move the mouse during loading.... the engine know the mouse has moved but does not show it on the screen (blit). But when the mouse becomes visible, it is where your cursor moves would have put it.
[17:15:23] <Colourless> yes
[17:15:36] <artaxerxes> instead of some predefined place that is not reflecting where the mouse actually us
[17:15:38] <artaxerxes> s/us/is
[17:49:10] <Colourless> time for me to go
[17:49:11] <Colourless> cya
[17:49:12] <-- Colourless has left IRC ("no comment")
[17:56:01] <artaxerxes> bye colourless
[18:15:28] <wjp> hi artaxerxes
[18:28:31] <wjp> 8:28pm up 117 days, 1:52, 1 user, load average: 91.63, 84.97, 80.97
[18:28:36] <wjp> 91 load, yay :-)
[18:58:24] <artaxerxes> hi wjp
[18:58:38] <artaxerxes> How in the world did you get a 91 load ? :)
[19:00:37] <artaxerxes> average 91 processes are on hold before getting a time slice of work
[19:00:57] <artaxerxes> I don't even think I have 91 processes! :)
[19:13:37] <wjp> slashdotted webserver :-)
[19:14:38] <artaxerxes> ah yes, about the escher guy?
[19:14:53] <artaxerxes> I'm guilty of charge then... :)
[19:28:08] <artaxerxes> did you follow the conversation with Colourless about the mouse issue?
[19:32:59] <wjp> no
[19:33:40] <artaxerxes> basically, menu does a blit_dirty() where it shouldn't
[19:34:10] <artaxerxes> as a result, the mouse acts erratically if you move it between the exult menu and the BG or SI menu
[20:01:45] <wjp> hm, yes, although I can't seem to reproduce it, the code does look like that
[20:03:46] <artaxerxes> for some reason, the exult menu does not have this problem
[21:14:51] <Darke|afk> wjp: Nice to see people are stress testing that webserver or yours. *grin*
[21:15:43] <artaxerxes> gotta leave
[21:15:50] <artaxerxes> see ya all
[21:15:51] <-- artaxerxes has left IRC ("using sirc version 2.211+KSIRC/1.1")
[21:16:51] <wjp> Darke|afk: yes, nice, isn't it? :-)
[21:17:04] <wjp> load reached 100 an hour ago :-)
[21:17:24] <Darke|afk> Ouch.
[21:17:47] <wjp> no signs of it slowing down anytime soon, either
[21:18:30] * Darke|afk boggles. >300 hits/second and it's not keeling over. Wow.
[21:19:46] <wjp> web log is 238Mb by now
[21:20:05] <wjp> at an approximate 3 lines (or 240 bytes) per hit, that's a million hits
[21:21:16] <wjp> hm, that's only 30 hits/second
[21:21:27] <wjp> I wonder if I made a miscalculation earlier
[21:24:32] <wjp> ah well, I guess the initial 300/s didn't last that long
[21:24:35] * Darke|afk hmms... it's halfway down the /. page, so it'll be gone 'soon'. *grin*
[21:25:36] * wjp crosses fingers :-)
[21:28:44] <Darke|afk> What's the problem? The webserver's holding up. *grin* Or are you feeling the drain in bandwidth over that gigabit link?
[21:28:51] <wjp> hm, does cvs auto-ignore .o files?
[21:29:02] <wjp> nah, bandwidth should be ok
[21:29:23] <Darke|afk> Nope. You need to have it in the .cvsignore.
[21:29:39] <wjp> but uwadv doesn't have it there, and it doesn't list them as "?"
[21:29:52] <Darke|afk> Weird.
[21:31:39] <wjp> also, all files seemed to have DOS-style endlines
[21:31:44] <wjp> s/seemed/seem/
[21:31:53] <wjp> but they aren't checked in as binary
[21:31:57] <Darke|afk> Hmm... `cvs import` ignores .o files.
[21:32:12] <Darke|afk> Odd.
[21:32:56] * wjp is tempted to do a 'dos2unix `find -name "*.cpp" -or -name "*.hpp"`'
[21:33:21] * Darke|afk snickers. Sounds good.
[21:34:13] <Darke|afk> It's not as if anyone should notice, anyway, it should be autoconverted for the windows guys.
[21:34:31] <wjp> well, but shouldn't it have autoconverted it when committing too?
[21:37:43] <Darke|afk> IIRC if you commit windows EOLs from a unix box it doesn't.
[21:39:50] <wjp> any idea if MSVC (or any other compiler) treat enums as unsigned?
[21:40:23] <wjp> I'm getting dozens of signed/unsigned warnings wrt enums here
[21:41:05] <Darke|afk> http://www.cvshome.org/docs/manual/cvs.html Here's the docs anyway to cvs. *grin* Feel free to find out.
[21:41:46] <Darke|afk> No idea. IMO, it's more likely they don't compile with the somewhat more pedantic warnings on, so aren't getting that.
[21:49:42] <wjp> oh, vividos released a new version of uwadv today
[21:51:11] <wjp> I love those release names :-)
[21:51:16] <wjp> 'cheesecake' :-)
[21:54:46] * Darke|afk snickers.
[21:57:14] * Darke|afk likes cheesecake. *grin*
[22:10:49] * wjp hmms
[22:11:10] <wjp> in plain make, there's this nice $(VAR:PATTERN=REPLACEMENT) function
[22:11:15] * Darke|afk really afks now. Work. *grin* Bye.
[22:11:20] <wjp> awww... :-)
[22:11:36] <wjp> ah well, I'll have to finish that sentence tomorrow :-)
[22:11:40] <wjp> bye
[22:11:56] * Darke|afk y'know, has a university labby job, he doesn't get a chance to sit around on computers all day. *grin*
[22:12:12] <wjp> eh?
[22:12:19] <wjp> what kind of lab?
[22:12:24] <wjp> computer lab? :-)
[22:12:31] <Darke|afk> IT lab. Or rather dozens of IT labs. *grin*
[22:12:43] --> matto|wookin has joined #exult
[22:12:49] <wjp> what kind of a job do you have there?
[22:12:54] <wjp> tutoring? user support?
[22:13:03] <wjp> hi matto
[22:13:18] <wjp> sysop?
[22:13:23] <Darke|afk> User support, making sure the computer labs work, odd jobs. Is pretty much the entire job description
[22:13:29] <wjp> :-)
[22:13:35] <wjp> sounds like what I do :-)
[22:13:40] <matto|wookin> wjp!
[22:13:54] <matto|wookin> has anyone read the current "Joke of the Week" on libsdl.org ?
[22:14:01] <Darke|afk> There's 5 of us, covering 12+ computer labs, with 2000+ students, plus 100 staff and other assorted randoms.
[22:14:03] <wjp> except that these 3 weeks it seems to involve general sysadmin work, since both sysops are away
[22:14:03] <matto|wookin> I found it pretty interesting
[22:14:19] <wjp> heh, cool :-)
[22:14:59] <Darke|afk> wjp: There's a few people who handle the unix servers and building the windows/linux REMBO desktop images. Though, but we handle most of the
[22:15:05] <Darke|afk> 'human interface' stuff. *grin*
[22:15:09] <wjp> :-)
[22:15:14] * Darke|afk really must go now. *grin* Bye!
[22:15:19] <wjp> bye :-)
[22:15:23] * wjp heads over to libsdl.org
[22:15:24] <matto|wookin> bye *fluff*
[22:15:36] <matto|wookin> wjp: I'd be curious if any of them can stump you :)
[22:16:47] <wjp> hm, no, I know all the usages of those words, except the verb in 21
[22:17:13] <matto|wookin> I believe that is synonymous with "imply"
[22:17:31] <matto|wookin> but I admit I have only inferred the definition by hearing it used in sentences repeatedly :)
[22:17:33] <wjp> to make known especially publicly or formally, to communicate delicately and indirectly
[22:17:54] <wjp> (m-w.com)
[22:17:54] <matto|wookin> I think #6 is my favorite
[22:18:03] <wjp> :-)
[22:18:35] <matto|wookin> yeah. to communicate delicately and indirectly is the definition I was getting at ... "to imply" ..
[22:19:00] <wjp> getting a sow to sow sounds interesting :-)
[22:19:03] <matto|wookin> haha
[22:19:27] <matto|wookin> #12 I don't understand ....
[22:19:33] <matto|wookin> a "row" among the oarsmen?
[22:19:39] <matto|wookin> oh wait ... I get it
[22:19:50] <wjp> :-)
[22:19:56] <matto|wookin> like oarsmen all sitting on the same bench
[22:20:02] <matto|wookin> I Was thinking "row" as in "argument" or "dispute"
[22:20:12] <wjp> yeah, that's the meaning they mean
[22:20:17] <matto|wookin> #14 is very clever :)
[22:20:28] <wjp> they were arguing about how to row
[22:21:07] <wjp> I guess it's a joke that buck = dollar, and 'doe' sounds like dough(sp?) = money?
[22:21:21] <matto|wookin> no
[22:21:32] <matto|wookin> they used the word "does" twice
[22:21:32] <wjp> (and the animal ones, of course)
[22:21:40] <wjp> oh, right
[22:21:40] <matto|wookin> once as a verb, once as a noun
[22:22:00] * wjp read straight over that first 'does' :-)
[22:22:06] <matto|wookin> I have never heard of "row" meaning an argument
[22:22:07] <wjp> too common a word to notice in that context :-)
[22:22:23] <wjp> no? hm, I have ;-P
[22:22:36] <wjp> you don't pronounce it the same as the other 'row', though
[22:22:43] <matto|wookin> 1. A boisterous disturbance or quarrel; a brawl. See Synonyms at brawl.
[22:22:51] * wjp nods
[22:22:59] <wjp> brawl, fight, same thing :-)
[22:22:59] <matto|wookin> row rhymes with cow?
[22:23:03] <wjp> yeah
[22:23:09] <matto|wookin> never heard that before
[22:23:13] <matto|wookin> or if I did.. don't remember
[22:23:54] <matto|wookin> I noticed the "does" because I got confused when reading the sentence :)
[22:23:58] <wjp> http://cougar.eb.com/sound/r/row00003.wav
[22:24:05] <matto|wookin> and had to go back and read it again
[22:24:32] <matto|wookin> too bad I don't have any sound on this machine :(
[22:24:37] <matto|wookin> hehe I like #9 also ... the dove dove
[22:24:38] <wjp> ah :/
[22:24:41] <wjp> yeah :-)
[22:25:00] <wjp> english is weird that you don't pronounce things the same way even when the spelling's the same
[22:25:31] <matto|wookin> hehe #18 .. haha ... I got confused reading that quickly too.. had to go read it slowly
[22:25:49] <matto|wookin> yeah... whoever designed English should be put in jail
[22:27:04] <wjp> is the 'b' still silent in 'number' (the 'more numb' meaning)
[22:27:05] <wjp> ?
[22:27:15] <matto|wookin> yes
[22:27:19] <matto|wookin> nummer
[22:27:20] <matto|wookin> hehe
[22:27:31] * wjp nods; k, wasn't too sure :-)
[22:27:43] <wjp> not really a word you hear every day :-)
[22:27:44] <matto|wookin> I am not even sure that "number" is a real word
[22:28:20] <matto|wookin> I guess it is according to dictionary.com
[22:36:43] <wjp> I should be going; have to get up in 6:30 hours :/
[22:36:46] <wjp> bye
[22:36:58] <-- wjp has left IRC ("Zzzz...")