#pentagram@irc.freenode.net logs for 14 Jul 2003 (GMT)

Archive Today Yesterday Tomorrow
Pentagram homepage


[01:01:06] --> Kirben has joined #pentagram
[01:01:06] --- ChanServ gives channel operator status to Kirben
[02:36:12] --> ShadwChsr has joined #pentagram
[02:36:17] <-- ShadwChsr has left #pentagram ()
[05:04:22] --> Cashman has joined #pentagram
[05:20:46] <-- Cashman has left IRC ()
[07:18:26] --> wjp has joined #pentagram
[07:18:26] --- ChanServ gives channel operator status to wjp
[09:31:23] --> Cashman has joined #pentagram
[09:39:21] <-- Cashman has left IRC ()
[09:55:46] --> Dominus has joined #pentagram
[09:55:58] --- wjp is now known as wjp|work
[09:55:59] <wjp|work> hi
[09:56:18] <Dominus> ah, here we go
[09:56:20] <Dominus> couldn't connect to irc.freenode.net the whole last evenning
[09:56:25] <wjp|work> yeah, DNS issue
[09:56:37] <wjp|work> freenode.info or freenode.org was supposedly still working
[09:56:47] <DarkeZzz> -lilo- [Global Notice] Hi all. The freenode domains were renewed somewhat at the last minute, due to a lack of time to work on some procedural guidelines for the nonprofit. Domain service should repropagate within the next few days. Apologies for the inconvenience, and thank you for using the network.
[09:56:59] <Dominus> got to remember this, next time :-)
[09:57:10] <wjp|work> DarkeZzz: yes, that one :-)
[09:57:58] <Dominus> it happened the same time I set up a firewall program on my system and at first I wasn't sure if that prevented it somehow...
[09:58:08] <DarkeZzz> Not that it's *too* much of an embarrisment/hassle, after all, even Microsoft can forget such a trivial thing as renewal of domain names... *cough*
[09:58:19] <Dominus> he he
[10:08:40] --> Cashman has joined #pentagram
[10:09:11] <Cashman> hey wjp I got my new graphics card - I still only get 20 odd fps in pentagram, thats fine but I see renedering totally relies on cpu
[10:09:35] <Cashman> I kinda guessed that but I thought my fps might have upped just a littke
[10:09:42] <wjp|work> why? it's 2d rendering
[10:09:51] <wjp|work> modern graphics cards only improve 3d...
[10:10:02] <wjp|work> pentagram is memory bandwidth limited
[10:10:52] <Cashman> yeah I worked that out - just that I had an onboard card running before and my system performances has increased a little as expected
[10:11:14] <Cashman> but 2d rendering is the same, I get it now anyway
[10:11:23] <Cashman> hows work
[10:11:24] <Cashman> ?
[10:11:26] <Cashman> busy?
[10:12:33] <Cashman> yeah mem bandwidth is a prob - man DDR hit the market quick, my mobo only supports sdram, my new graphics card has ddr - guess I'll have to upgrade my mobo one of thease days
[10:18:45] <-- Cashman has left IRC ()
[10:30:36] --> Cashman has joined #pentagram
[10:53:17] <-- Dominus has left IRC ("enough for now")
[10:54:47] <-- wjp|work has left IRC (vinge.freenode.net irc.freenode.net)
[10:55:00] --> wjp|work has joined #pentagram
[10:55:01] --- ChanServ removes channel operator status from wjp|work
[10:55:37] <-- Cashman has left IRC ()
[11:13:47] --- wjp|work is now known as wjp
[11:13:48] --- ChanServ gives channel operator status to wjp
[11:13:54] --- wjp is now known as wjp|work
[11:35:58] --> Cashman has joined #pentagram
[12:13:24] <-- Cashman has left IRC ()
[12:14:40] <wjp|work> DarkeZzz: any experience with figuring out why ssh's publickey Doesn't Work(tm) for one account? :-)
[12:24:17] <wjp|work> nvm... got it :-)
[12:28:53] <wjp|work> (home directory was 775...)
[12:50:57] <DarkeZzz> Nice to see ssh has a sanity check or two. *grin*
[13:01:34] <wjp|work> well, in this particular case the directory is supposed to be 775 :-)
[13:01:49] <wjp|work> (it's a special group with a select few users)
[13:23:34] --> Colourless has joined #Pentagram
[13:23:34] --- ChanServ gives channel operator status to Colourless
[13:24:01] <Colourless> hi
[13:26:18] <Colourless> fun fun fun....
[13:26:33] <Colourless> GravityProcess moves an item...
[13:26:46] <Colourless> moving an item can cause it to leave the fast area...
[13:27:03] <Colourless> if the item has flag 0x80 set, it is destroyed when leaving the fast area...
[13:27:13] <wjp|work> hi
[13:27:17] <Colourless> GravityProcess has no idea the item has been destroyed
[13:27:18] <wjp|work> hmm.. fun indeed
[13:29:56] <wjp|work> does it terminate itself if the item is gone?
[13:30:07] <wjp|work> (when getItem(item_num) == 0, that is)
[13:30:13] <Colourless> yes
[13:30:41] <wjp|work> http://www.gamespot.com/gamespot/features/all/gamespotting/071103minusworld/1.html :-)
[13:31:00] <Colourless> hmm, this probably isn't a real problem
[13:31:11] <Colourless> just looking at the gravity process code
[13:32:32] <Colourless> might just be a bug in my code
[13:33:20] <Colourless> hehe, real life :-)
[13:33:33] <wjp|work> "For example, even if your character specializes in dentistry, that doesn't mean you can perform a root-canal operation on yourself."
[13:38:59] <Colourless> :-0
[13:39:05] <Colourless> s/0/)/
[13:52:52] <Colourless> this is just icky
[13:54:03] <Colourless> item->destroy() will kill the list iterators when closing containers and when a chunk leaves the fast area
[13:55:57] <wjp|work> not much we can do about that
[13:56:20] <wjp|work> although it doesn't _have_ to happen in a list
[13:57:19] <Colourless> well, we could defer the actual destroying of the items till a time when we know nothing is being iterated (i.e. in App::run())
[13:58:30] <Colourless> would probably put it into the object manager
[14:00:03] <Colourless> or mayby world
[14:00:36] <wjp|work> how about the 'old trick' of moving the iterator forward before destroying the item?
[14:00:55] <Colourless> i'm already doing that in one place
[14:02:01] <Colourless> i guess i can just update 'known' problems with that trick
[14:02:21] <wjp|work> iterating object lists shouldn't happen in many places
[14:02:51] <Colourless> no it doesn't
[14:03:15] <Colourless> and i think the only 2 problem places are actually going to be ContainerGump::Close and CurrentMap::unsetChunkFast
[14:09:19] <Colourless> yay, it didn't crash this time :-)
[14:09:51] <wjp|work> whee :-)
[14:10:50] <Colourless> something fun that you'd like to know. some of the glob eggs in the east road have the fast area flag set in the map file, which stops them from being expanded :-)
[14:11:16] <Colourless> so, i have to unset the fast area flag when loading maps
[14:11:55] <Colourless> oh yeah, i think i'll mention that the skull of quakes now works :-)
[14:12:10] <wjp|work> nice :-)
[14:12:26] <Colourless> sort of... that avatar doesn't get up
[14:12:33] <Colourless> and i know why too
[14:12:43] <wjp|work> that doesn't happen with current cvs either
[14:12:55] <Colourless> avatar->enterFastArea() is so the avatar will get up after skull of quakes :-)
[14:13:03] <wjp|work> ah :-)
[14:13:21] <wjp|work> nice hack :-)
[14:13:34] <Colourless> of course i can't think of a nice way of getting it to work
[14:14:04] <Colourless> perhaps by making any item that goes to 0,0 forced to go out of the fast area, even if it's the camera object
[14:15:24] <Colourless> ah, what the hell, i'll add the hack
[14:18:41] <wjp|work> do you think the length of the mouse cursor should depend on the absolute distance from the "center" or the relative distance? (compared to screen size)
[14:19:13] <Colourless> relative would be best i think
[14:19:23] <Colourless> but that is 'my' opinion
[14:19:28] <wjp|work> I was tending to that too
[14:20:22] <wjp|work> as for determining which frame to draw... I don't want to make things too complicated, so I was thinking to just use the 'normal' cursors unless setMouseCursor was called
[14:20:43] <wjp|work> the mouse cursors could maybe be a stack
[14:21:02] <wjp|work> (which would simplify some things and complicate others)
[14:21:30] <wjp|work> IIRC, the only things that change the mouse cursor are modal "things" (gumps/target)
[14:21:46] <wjp|work> (and saving/loading used a pentagram cursor, I think)
[14:22:55] <wjp|work> so I was thinking of having a modal gump's "open" call setMouseCursor, and the "close" resetMouseCursor
[14:22:56] <Colourless> and on moves
[14:23:02] <Colourless> moving items that is
[14:23:05] <wjp|work> ah, yes
[14:23:10] <wjp|work> which is modal too
[14:23:16] <Colourless> aye it is
[14:23:35] <wjp|work> speaking of which... I really need to get the kernel paused while dragging :-)
[14:23:37] <Colourless> or 'will' be :-)
[14:23:41] <wjp|work> :-)
[14:23:52] <wjp|work> might do that tonight
[14:25:25] <wjp|work> btw... to refresh my memory:
[14:25:31] <wjp|work> skeletal hand was for the diary
[14:25:33] <wjp|work> quill for saving
[14:25:39] <wjp|work> a blueish 'orb' for loading
[14:25:42] <wjp|work> ?
[14:25:56] <Colourless> there was a pentagram while saving and loading and changing maps
[14:26:00] <wjp|work> there were also a red and a blue dot
[14:26:21] <wjp|work> (frames 24 and 33 IIRC)
[14:26:27] <wjp|work> couldn't quite place those
[14:26:39] <wjp|work> maybe when you had the cursor exactly on the center?
[14:26:52] <wjp|work> (although I seem to remember you just got the shortest cursor there)
[14:27:38] <Colourless> might have been some editor related stuff
[14:30:40] <Colourless> uh, it was AVATAR::cacheIn that makes the avatar stand up
[14:30:45] <Colourless> not enterFastArea
[14:31:35] <wjp|work> hmm
[14:44:52] <Colourless> it crashed
[14:44:54] <Colourless> :-)
[14:44:58] <wjp|work> uh oh :-)
[14:48:03] <Colourless> well, the reason why it actually crashed, isn't the only problem :-)
[14:49:03] <Colourless> it crashed while attempting to delete the avatar
[14:49:10] <wjp|work> hm, again? :-)
[14:49:21] <Colourless> yes
[14:49:34] <wjp|work> wasn't that the problem that got us into this whole chunk-redesign in the first place? :-)
[14:49:34] <Colourless> when moving from the teleporter room to the catacombs
[14:50:15] <wjp|work> should the avatar be moved _after_ the fastarea has been updated by teleport?
[14:50:42] <wjp|work> if it's teleported into a chunk that'll be immediately deleted...
[14:50:50] <wjp|work> s/it/he/...
[14:53:42] <Colourless> it might be a bug in our code somewhere
[14:54:02] <wjp|work> assert on deleting the avatar?
[14:54:11] <wjp|work> although it's probably quite obvious where he's being deleted
[14:54:22] <Colourless> GLOBSWAP
[14:54:27] <Colourless> ip = 0x114 :-)
[14:54:47] <wjp|work> no surprises there
[14:59:43] * wjp|work counts down to 17:00...
[14:59:51] <wjp|work> 10...
[14:59:57] <wjp|work> 5...
[15:00:02] <wjp|work> right, I'm off ;-)
[15:00:04] <wjp|work> bbl :-)
[15:00:05] <Colourless> :-)
[15:00:09] <Colourless> cya
[15:00:14] <-- wjp|work has left IRC ("bbl")
[15:18:40] --> wjp has joined #pentagram
[15:18:40] --- ChanServ gives channel operator status to wjp
[15:19:38] <wjp> too... hot... *melts*
[15:19:46] <Colourless> wb
[15:25:00] <wjp> thx
[16:09:27] --> Dark-Star has joined #pentagram
[16:10:23] <wjp> let's see... what "states" can pentagram be in...
[16:10:26] <wjp> running normally
[16:10:31] <wjp> paused due to dragging
[16:10:37] <wjp> (does it pause when dragging gumps too, btw?)
[16:10:37] <Colourless> running :-)
[16:10:43] <Colourless> that's all :-)
[16:10:51] <wjp> paused due to a modal gump
[16:11:06] <wjp> let's go one level below "running" ;-)
[16:11:08] <Colourless> sigh, looks like i have to go 'test' something now :-)
[16:11:53] <Colourless> being that only pentagram team member that can actually go run the original game 'easily' can be a drag :-)
[16:12:23] * wjp tactfully doesn't mention that his parents' computer also runs U8
[16:12:43] <wjp> *oops* ;-)
[16:12:44] <Colourless> i said easily :-)
[16:14:14] <wjp> it's paused while dragging gumps
[16:14:21] <wjp> fire stops animating, anyway
[16:14:25] <Colourless> yes
[16:14:36] <Colourless> and the cursor is the normal cursor
[16:14:45] <Colourless> normal being move
[16:14:47] <wjp> (PC was already on and my parents weren't on it :-) )
[16:16:25] <Colourless> not about combat movement and cursors
[16:16:31] <Colourless> s/not/note/
[16:16:51] <Colourless> the cusror is alwasy the red arrow no matter where the cusror is on screen
[16:17:07] <wjp> yes, I thought old/ did that wrong
[16:17:33] <Colourless> however IF the cursor is in the 'run' zone and you right click, then you temporarily go normal movement state
[16:17:37] <wjp> I also thought it was rather annoying in the original that there was no indication of when you'd start running
[16:17:58] <Colourless> the cursor goes blue, and acts like the normal blue cursor
[16:18:27] <wjp> when exactly do you switch back to combat stance? as soon as you stop or as soon as you stop running?
[16:18:33] <Colourless> so when you are in 'running' mode you can walk and take single steps.
[16:18:40] <Colourless> when you let go of right
[16:19:12] <wjp> ok
[16:19:56] <wjp> http://games.slashdot.org/games/03/07/14/0454247.shtml?tid=127&tid=186&tid=206&tid=209
[16:21:15] <Colourless> blue magnifying glass was used for the read game 'menu'
[16:21:41] <wjp> oh, that was a magnifying glass? :-)
[16:21:47] <wjp> I guess I missed the handle
[16:22:40] <Colourless> fallout3... possibly the worst kept secret in gaming :-)
[16:22:47] <wjp> quite possibly, yes :-)
[16:41:29] <Colourless> tag skill won't have % signs any more according to the unofficial faq!
[16:59:33] <wjp> no skills will have one
[16:59:58] <Colourless> uh, that's what i meant :-)
[17:02:31] --> Fingolfin has joined #pentagram
[17:02:32] --- ChanServ gives channel operator status to Fingolfin
[17:02:39] <wjp> hi Max
[17:03:28] <Fingolfin> hi
[18:10:28] <Colourless> i still don't see why we have ucmachine execute the usecode code, when ucprocess could be doing that. it's not really a huge difference, but with the amount of code in UCMachine::execProcess that have p-> in them, it would seem logical to be doing things the 'other' way
[18:11:33] <Colourless> it would probably also mean that UCMachine wouldn't have to be a friend to UCProcess either, since UCMachine wouldn't need to know the internals of UCProcess
[18:15:53] <Colourless> ah well, i'm just rambling here
[18:17:10] <wjp> you have somewhat of a point
[18:17:31] <wjp> although I'm still not really sure about which way is "best"
[18:18:11] <Colourless> in the end it doesn't really matter, as long as it works
[18:20:44] <wjp> did you figure anything out about the deleting-Avatar crash, btw?
[18:21:36] <Colourless> as far as I can tell, the original just wouldn't destroy the avatar if Item::destroy was called on the avatar
[18:21:47] <wjp> convenient :-)
[18:22:18] <Colourless> i would commit the changes now, but the ask gump is broken
[18:22:45] <Colourless> and i need to fix that first
[18:22:57] <wjp> how did that happen? :-)
[18:23:29] <Colourless> beats me
[18:33:12] <-- Kirben has left IRC (Read error: 54 (Connection reset by peer))
[19:24:03] <Colourless> i'm off
[19:24:11] <Colourless> cya
[19:24:14] <-- Colourless has left IRC ("casts invisibility")
[19:43:47] --> DraX has joined #pentagram
[19:48:18] <-- DraX has left IRC (Client Quit)
[21:22:42] --> Dominus has joined #pentagram
[21:23:36] <Dominus> hi
[21:24:03] * Dominus just wants to add that he is also able to run the original quite easily :-)
[21:24:24] <wjp> great! say.. there's this thing I wanted to know... ;-)
[21:25:22] <Dominus> as long as it is not about sfx :-) (that is the only thing not running correctly on XP with this game)
[21:27:23] * wjp tries to come up with a horribly annoying series of carefully timed actions (that don't involve any SFX, of course)
[21:27:35] <Dominus> he he
[21:27:51] <wjp> ah, thought of one :-)
[21:28:01] <wjp> can you pick up an item in mid-air after you threw it? :-)
[21:28:45] <wjp> (which, incidentally, really is something I was wondering about recently)
[21:28:47] <Dominus> say, what I'm really wondering about, and am quite sure there is an easy and logical explanation for, why are the animations of the execution scene working correctly but the guards make the moonwalk ?
[21:28:58] <wjp> guards use pathfinding
[21:29:19] <wjp> guess what I didn't implement yet ;-)
[21:29:22] <Dominus> ok, the other is usecode controlled, right?
[21:29:49] <wjp> usecode runs individual animations, yes, if that's what you meant
[21:29:58] <wjp> in case of the guards usecode runs pathfinding
[21:30:00] <Dominus> uhm, quite everything is usecode controlled in u8...
[21:30:44] <wjp> yes :-)
[21:30:49] <Dominus> scratch that sentecne "ok, the other is usecode controlled, right? " didn't think :-)
[21:31:16] <wjp> but for the execution scene pathfinding isn't used
[21:31:32] <wjp> it just makes NPCs take a fixed number of steps in a fixed direction
[21:31:46] <Dominus> k
[21:32:39] <Dominus> as for picking an item in mid-air, I doubt you could manage to test that :-)
[21:33:18] <wjp> you don't say :-)
[21:33:41] --> servus has joined #pentagram
[21:34:01] <Dominus> well, if you could slow down the animations while having the mouse speed run normal...
[21:35:06] <Dominus> my favourite playtesting in pentagram: throwing things into the water
[21:46:38] <Fingolfin> hehe
[21:46:56] <wjp> time for bed... 'night
[21:46:59] <Fingolfin> a pity that my patch that implements dropping items on the proper spot is gone together with my iBook <sigh>
[21:47:09] <Fingolfin> night willem
[21:47:15] <-- wjp has left IRC ("Zzzz...")
[22:35:43] <-- Dominus has left IRC ("enough for now")
[22:50:05] <-- Dark-Star has left IRC ()
[23:03:53] --> SB-X has joined #pentagram
[23:24:49] <-- SB-X has left IRC ("X-Chat")
[23:33:22] <-- servus has left IRC (Read error: 60 (Operation timed out))