#pentagram@irc.freenode.net logs for 4 May 2003 (GMT)

Archive Today Yesterday Tomorrow
Pentagram homepage


[01:21:49] --> Kirben has joined #pentagram
[01:21:49] --- ChanServ gives channel operator status to Kirben
[05:17:21] --> sbx has joined #pentagram
[05:32:06] --- DarkeZzz is now known as Darke
[06:36:26] --> Colourless has joined #Pentagram
[06:36:31] --- ChanServ gives channel operator status to Colourless
[06:49:40] <-- Kirben has left IRC (Read error: 104 (Connection reset by peer))
[06:49:46] --> Kirben2 has joined #pentagram
[06:49:50] --- Kirben2 is now known as Kirben
[06:56:48] <Colourless> enterFastArea and leaveFastArea are going to be fun fun fun to implement... as in, i have no real idea of a nice method so far of doing it. Pretty they work as a proximity trigger, and as far as I can tell it was for 'about' on screen in the original.
[06:56:49] <Colourless> A logical way of doing it would be to call enterFastArea for any object that was rendered this frame, but not last frame, and call leaveFastArea for any object rendered last frame and not this frame...
[06:57:28] <Colourless> that is 'mostly' fine and isn't difficult to do
[06:57:47] <Colourless> however it's possible that on screen will be a too large or too small of an area
[06:57:58] <Colourless> (depends on resolution)
[06:58:59] <Colourless> and additionally it would mean that usecode would end up being called while we are painting which is bad, or 1 frame will be rendered before enterFastArea would be called, which is also bad since enterFastArea may do something significant to the object
[07:01:38] <Colourless> this simply will not do...
[07:02:11] <Darke> Makes sense.
[07:02:53] <Colourless> an alternative idea would be to link to the gross object culling that i will implement sometime (today perhaps)
[07:03:18] <Colourless> firstly list of items in the fast area will need to be kept so we can tell when they leave
[07:04:23] <Colourless> this list would be used to detect what object have left the fast area since last frame
[07:05:52] <Colourless> then when pulling in objects to render from the globs they are then set to be in the fast area. Any object in the old list that hasn't been pulled in has left the fast area, and any object that wasn't in the old list and has been pulled it has entered the fast area
[07:07:22] <Colourless> to detect when an object has just entered the list, a flag should probably be used.
[07:07:32] * Darke nods.
[07:07:38] * sbx smiles and nods.
[07:08:10] <Colourless> detecting when an object has left the fast area would involve iterating all the objects in the list, and seeing which ones no longer have the fast_area_flag set or similar
[07:08:23] <Darke> sbx: I see you have absolutely no idea what we're talking about. *grin*
[07:08:45] <sbx> I'm all over it man.
[07:08:52] <Colourless> thinking about it, having a last_in_fast_area integer which is set to the frame number of the last time the object was in the fast area might be better
[07:09:15] <Colourless> that way the flag will not need to be cleared
[07:09:33] <Darke> sbx: Just wait 'til I start rambling as to how the usecode-script language is going to look. *grin*
[07:10:05] <Colourless> so for everyone object something like this would happen
[07:10:37] <Colourless> if (item->last_in_fast_area != frame_num -1) item->enterFastArea();
[07:10:48] <Darke> Colourless: That would work. Is this list of things to be drawn being re-generated with each frame? Or are you just adding, or removing items with each frame?
[07:11:02] <Colourless> new list every frame
[07:11:17] <Darke> Ahh.
[07:11:17] <Colourless> probably have 2 lists that are exchanged and cleared for speed reasons
[07:11:22] <Colourless> new_fast_area_list->add(item);
[07:11:26] <Colourless> then
[07:11:39] * Darke nods.
[07:12:07] <Colourless> item->last_in_fast_area = frame_num;
[07:12:13] <Colourless> then
[07:12:27] <Colourless> for each item in old_fast_area_list
[07:12:50] <Colourless> if (item->last_in_fast_area != frame_num) item->leaveFastArea();
[07:14:24] <Colourless> it is likely to be unwise to call enter and leaveFastArea while iterating the objects, so the calls should be defered till next frame but that should be fine if the region is large enough
[07:14:50] <Colourless> so the processes are added to the kernel, but not executed
[07:20:42] <Darke> Yup.
[07:22:53] <Colourless> it is 'annoying' that the world isn't on the same grid as the screen.
[07:24:39] <Colourless> makes it hard to determing when something may or may not be on screen
[07:24:50] <Colourless> by only using it's 3d bounding box
[07:26:16] <Colourless> you can always use a cuple of dot products (which i do in some places) but that is messy
[07:29:46] <Colourless> it's not that much of a hardship, but it's annoying
[08:14:39] * Darke nods.
[08:15:57] * sbx smiles and nods.
[08:21:46] <-- Colourless has left IRC ("l8tr")
[08:35:34] --> wjp has joined #pentagram
[08:35:34] --- ChanServ gives channel operator status to wjp
[09:20:46] <-- sbx has left IRC ("X-Chat [1.6.4]")
[10:20:43] --> Dark-Star has joined #pentagram
[11:03:50] --> CashZzz has joined #pentagram
[11:04:18] --- CashZzz is now known as Cashman
[11:11:06] <Cashman> gee well I havn't compiled pentagram in about 2 weeks! I've just been getting my way through the irc logs
[11:11:15] <Cashman> what stage are you guys at!
[11:12:00] <Cashman> do we have some graphics on screen now? it appears with what I see in the logs - what I find interesting is everyone talking about there hardware configurations
[11:12:10] <Cashman> and what optimizations will finially be needed
[11:13:20] <Darke> Yep. There's a little bit of graphics now. Not that you can *do* anything with the graphics. Looks like a console window with an image as it's background, it does. *grin*
[11:14:15] <Cashman> ok nice! - well irc is certainly doing well
[11:14:35] <-- Kirben has left IRC (Read error: 104 (Connection reset by peer))
[11:15:56] <Cashman> is there a pic!
[11:16:07] <Cashman> http://www.users.on.net/triforce/usage.png I found this but its not a pic of the map
[11:20:18] --> Kirben has joined #pentagram
[11:20:18] --- ChanServ gives channel operator status to Kirben
[11:25:23] <Darke> You're probably looking for: http://www.users.on.net/triforce/pent_world.png
[11:26:10] <Cashman> thanks darke!
[11:27:36] <Darke> Colourless posted it in the forum. *Grin*
[11:28:02] <Cashman> thanx to Colourless!
[11:28:33] <Cashman> ?? any progress on grapping a p4 from uni or is that not going to happen!
[11:28:44] <Cashman> I mean if you have access to a p4 ur lucky enough
[11:29:48] <Cashman> heheh grapping! ohh so bad spelling error
[11:30:13] <Darke> Not really. The dual P3-1ghz I've co-opted as my desktop machine effortlessly blows away the P4 2ghz's floating around the place. *grin*
[11:31:01] <Cashman> heh ok!
[11:31:11] <Cashman> nice
[11:33:19] <Cashman> man I didnt know colourless made a duke port! cool
[11:33:21] * Darke tries to remember how to use GIMP, so he can try to make a nicer looking console font. *grin*
[11:33:25] <Cashman> hehe just been looking at his website
[11:33:34] <Cashman> yeah
[11:33:45] <Cashman> I dont use linux so I dont follow
[11:38:24] <Darke> Colourless (I presume) made a console font for pentagram, it's a bit difficult to read some of the letters, so I've been looking at tweaking them a bit.
[11:39:21] <wjp> isn't that just because of the graphical/colourfull background?
[11:40:37] <Darke> I don't think so. I think the main problem is just the fact the letters are so *tiny* on a 1600x1200 display. *grin*
[11:40:54] <Cashman> I really like the screeny!
[11:41:47] <wjp> kind of looks like U8, doesn't it? :-)
[11:42:21] <Cashman> yeah its great wjp
[11:42:48] <Darke> I wonder if there's any way to configure the height it'd accept as a font. My terminal font is only a little bigger, so it wouldn't be difficult... hmm... maybe if we had a xml specification file that describes the font, then the console gets the font details from that...
[11:43:28] <wjp> ...
[11:43:30] <wjp> :-)
[11:44:15] <Darke> Config's created before loading the font (obviously) so it wouldn't be too difficult.
[11:45:55] <wjp> I toyed with a 'theme' config file a bit a few months ago
[11:46:00] <Darke> Something like <font><path>"@data/DarkeConFont.tga"</><height>9</><width>8</></> Or something. Then have a <confont>"@data/DarkeConFont.cfg"</> section somewhere in pentagram.cfg?
[11:46:07] * Darke was just about to suggest the same thing. *grin*
[11:46:37] <Darke> You'd have a 'general' theme, then just override based on the game you're loading.
[11:47:32] <Cashman> so you got past the drawing order for globs! and shpes etc. - is it based on ur origional method
[11:47:48] <Cashman> hmm I dont know who I'm addressing this question to
[11:52:20] <wjp> Colourless probably used the same ordering as in the 'old' map viewer
[11:54:22] <Cashman> oh ok thought it was colourless - oh thats right! yeah I talked to darke once about it
[11:55:54] <Cashman> just wondering because I havnt downloaded the new cvs and complined it yet but does the animation start before the usecode dies or no animation working at all yet
[11:57:38] <wjp> no animation yet
[11:57:44] <Cashman> ok
[12:01:05] <Cashman> anyone tried pulling up part of a crusader map
[12:02:37] <Darke> Of course not! It's unsupported remember. *grin*
[12:03:46] <Cashman> oh I knew that! just thought someone might try anyways!
[12:04:34] <Darke> Any incriminating evidence at http://pentagram.sf.net is a figment of your imagination. *nodnod*
[12:04:38] <Cashman> so whats currently being worked on? I saw logs of colourless jabbering on about fast and slow areas
[12:05:18] <Darke> Umm... I'm supposed to be doing something in relation to usecode scripts, but I got a little distracted messing with this console font stuff.
[12:05:21] <wjp> um, no, about FastArea's :-)
[12:06:44] <Cashman> ok, yeah Darke the current font doesnt look to bad but it would be good to work on it
[12:06:51] <Darke> A 'FastArea' in pagan/TGWDS comprises the currently viewable area of the screen, plus a little around the outside.
[12:07:31] <Darke> Either about a 2x2 area with the screen being 1x1 in the center, or a 3x3 area with the screen being 1x1 in the center. We're really not sure which, and those are just guesstimates anyway. *grin*
[12:07:45] <Darke> Basically it's the area where 'stuff happens'.
[12:07:47] <Cashman> oh ok! - reference to what you are viewing and the map as a hole I see!
[12:07:57] <Cashman> ok
[12:08:22] <wjp> could be something like all 'globs' which are partly or fully on-screen
[12:08:35] <Darke> Anything out side the area should 'stop' doing whatever it does when it gets outside it (like a 'forcefield' or a fireplace animating), and anything that enters the area should start.
[12:09:04] <Darke> wjp: Wouldn't explain being shot at from way off screen in tgwds. *grin*
[12:09:22] <wjp> a fixed number of globs around it too, then? :-)
[12:09:30] <Cashman> oh ok - switching in and out
[12:09:55] <Cashman> is full screen getting put in soon! faster? always isnt it
[12:10:01] <Darke> Which was one of *the* most frustrating bits about it, you could be being shot at almost a screen away and not be able to shoot at the attacker, since you couldn't see it to target it! *grin*
[12:10:03] <Cashman> sdl, directx, opengl - what ever
[12:10:14] <Cashman> hhe
[12:10:40] <Cashman> I see how important the design of that is now
[12:11:17] <Darke> wjp: It also just 'might' be the fact that once you've actiated (enterFastArea) an attacking turret, it doesn't deactivate immediately when you leaveFastArea.
[12:12:02] <Darke> Rather schedule itself for deactivation some time in the future... Hmm... I think I've seen that multiple times. Might have to test it out once I get a win98 box working again.
[12:12:18] <Cashman> hehe thats a good point!, even though your addressing that to wjp! I havn't played u8 in ages - I think I just might have to!!
[12:12:40] * Darke has never played u8.
[12:12:54] <Darke> Don't ask what I'm doing here, even *I* don't know. *grin*
[12:13:51] <Cashman> Darke! well I see yeah pentagram is in a window but what about a full screen mode option! I mean it will be faster - even without any code optimization??
[12:14:05] <Cashman> right!
[12:14:22] <wjp> probably won't be much faster in windows
[12:15:31] <Cashman> ? yeah I dont know jack about linux so I can't talk on that level but as far as ms windows is concerned! programs running in window mode or full screen - full screen is generally always faster and much faster!
[12:16:03] <Cashman> even a simple opengl, directx or sdl graphics demo
[12:17:07] <wjp> hm, I thought you could use accelerated directdraw in a window too?
[12:17:22] <wjp> ah well, I don't really know much about directx :-)
[12:17:25] <Darke> Not a clue. I wouldn't think window/fullscreen would make much of a difference. I don't think we're using any features of a 3dcard that opengl/directx would accelerate, so I don't think you're looking at any real speedups.
[12:17:51] <Darke> Last time I checked exult ran at about the same speed in a window as in fullscreen.
[12:17:54] <wjp> well, isn't it something like dga in linux?
[12:18:51] <Cashman> oh ok! well if you've done similar things with exult then you've probably answered my question! but I have seen a lot of code where it does make a difference
[12:19:14] <wjp> anyway, fullscreen will definitely be supported sometime
[12:19:22] <Darke> For instance Diablo2 run in a window vs. full screen doesn't make much of a difference unless you turn on full 3d support.
[12:19:27] <Cashman> as for accelereated directdraw in a window! I cant answer that becuase I've never see in used, if its possible?!
[12:19:28] <wjp> it's just that we don't really want to run it fullscreen now
[12:20:45] <Cashman> I understand your point and see where your comming from - there really isnt much point in putting that kinda of thing in at the moment anyways! nor is optimization important until the final stages - when ever that is decided, which might seam like never
[12:20:59] <Cashman> its heard I surpose when to decide about such things aye!
[12:22:11] <Darke> Something like that. *grin*
[12:22:39] <Cashman> oh well I better download cvs and see whats been happening for my own eyes
[12:22:50] <Cashman> hehe yeah sorry Darke! I'm babbling on now!
[12:24:20] <Cashman> hehe get this I'm watching tv - well its the formula 1 racing,
[12:24:55] <Cashman> one of the commitators was babbling on - how he wished he knew how much fuel the cars were currently using
[12:25:14] <Cashman> he said "we need thease teeeange hackers to hack into the mainframe" bla bla bla hhehe!
[12:26:03] <Cashman> man with racing especially formula 1 - fuel is everything when it comes to speed and weight
[12:26:33] <Cashman> anyways I've talked enough for one night/morning!
[12:26:39] <Darke> A bit pointless. All you need is someone to look over the guy in the pit sitting in front of a laptop screen. Most of those cars have so much computer equipment inside them, they can dump all relevant and irrelevant details out to the console.
[12:26:44] <Cashman> as its 12:24am here
[12:27:04] <Cashman> true true!
[12:27:15] <Cashman> yeah just get some little kid to do the runnings!
[12:27:30] <Darke> I remember there being a story recently about how since most of the teams use a form of wifi to transmit the data from the cars, you'd have to be worried about data interceptions by rival teams. No doubt they'd encrypt the stream though, but you'd have to wonder.
[12:27:31] <Cashman> job=runner
[12:28:00] <wjp> Darke: "no doubt"? :-)
[12:28:30] <Cashman> yeah thats true!! what about the cockpit cams! thats all wireless obviously
[12:29:11] <Cashman> theres been trouble with wireless cams in someplace hehe no encryption and eager phreaking kids who want to play with stuff
[12:30:21] <Darke> I think the cockpit cams are old tech, sattelite transmissions and stuff.
[12:30:40] <Cashman> yeah
[12:31:43] <Cashman> ouch some guys just spun out
[12:32:24] <Cashman> he spun out into the sand and of course with those tires you just get stuck, beaches
[12:32:26] <Cashman> beached
[12:32:53] <Cashman> 1 guy from a team of 2 brothers
[12:34:32] <Cashman> hmm sleepy
[12:34:37] <Cashman> sleepy
[12:34:40] --- Cashman is now known as CashAlmostZzz
[12:36:47] <CashAlmostZzz> m schumacher back in first!
[12:36:56] <CashAlmostZzz> yes
[12:45:49] <Darke> Yayfun. You can now override the default confont, by pointing it to a config file for the font, with the appropriate path in it. Of course you still can't setup new width/hight sizes but that just needs me to work out how to set them. *grin*
[12:46:47] <CashAlmostZzz> nice! - got to be a small but useful feature
[12:47:45] <Darke> FileSystem and Configuration seem to work very nicely together, that was almost too easy. *grin*
[12:47:48] --> Colourless has joined #Pentagram
[12:47:49] --- ChanServ gives channel operator status to Colourless
[12:48:22] <CashAlmostZzz> should be about the same under ms win?!
[12:48:31] * Darke blames Colourless for the ease of use of FileSystem and Configuration! He's an *evil* dragon! Now he's forcing me to shuffle *everything* into config files!
[12:48:37] <Colourless> uh, hi??!
[12:48:57] <Colourless> i've gotten sorting down to about 4 ms now :-)
[12:49:28] <CashAlmostZzz> thats under ms win colourless? you use msvc 7?
[12:49:36] <Darke> So it's faster then painting? *grin*
[12:49:39] <Colourless> msvc 6
[12:49:48] <Colourless> yeah, faster to sort than paint now :-)
[12:50:24] <CashAlmostZzz> ok - I await authorization in icq from colourless! hehe just that I use Icq and found ur website - well to get those nice pentagram screenshots of urs
[12:50:48] <CashAlmostZzz> nice to see you worked on a build engine port - I'm a fan of duke 3d and the build engine
[12:51:06] <wjp> hi Ryan :-)
[12:51:29] <wjp> Colourless: by using the per-glob object lists?
[12:51:29] <CashAlmostZzz> thanks
[12:51:31] <Colourless> the duke3d port is all of about 2 weeks old :-)
[12:51:32] <CashAlmostZzz> anyone else here use icq?
[12:51:35] <Colourless> wjp yeah
[12:52:52] <CashAlmostZzz> nice!! new and good
[12:53:18] <CashAlmostZzz> i've seen others and they lack sound support of anykind
[12:54:53] <Colourless> most of the work to port duke3d and build wasn't done by me.
[12:55:16] <Darke> Colourless: What fps do you get now? We can't let it get *too* high, or we might have to put in artificial limitors or something. *grin* We only need 10fps or something don't we?
[12:55:21] <CashAlmostZzz> i'm reading the website now
[12:55:30] <Colourless> we need 20 i think
[12:55:44] <CashAlmostZzz> fps for what?
[12:55:58] <Colourless> i get 40 fps windowed at the moment :-)
[12:56:20] <Colourless> i'm sure executing usecode will slow things down a bit
[12:56:23] <Darke> Colourless: Fair enough. To properly handle mousie clicks? I thought we'd concluded that the original ran at about 10, or I may just be confused. *grin*
[12:56:51] <Colourless> well, my tests are running animations suggested that 10 was too slow
[12:56:57] <Colourless> s/are/at/
[12:57:04] <CashAlmostZzz> I know the limits of my graphics card in 650x480x16 bit - 75 fps, so when I get around to compiling i wonder what I'll get, well seeing as there isnt any optimization probably around 40 (same as the dragon)
[12:57:23] <Colourless> pentagram old runs at 20 fps or 50 ms per frame
[12:57:56] <Colourless> it the interpolates inbetween frames for smooth movement
[12:58:11] <Colourless> under opengl pentagram old runs excessively smooth
[12:58:14] <Colourless> (for me)
[12:59:08] * Darke doesn't remember having any jerkyness with it either.
[12:59:17] <CashAlmostZzz> who did the opengl port
[12:59:21] <Colourless> i did
[12:59:37] <CashAlmostZzz> ok
[12:59:39] <wjp> who else :-)
[13:00:05] <CashAlmostZzz> fair enough
[13:00:11] * wjp wonders if he tried the opengl 'old'
[13:01:12] * Darke adds 'documenting the more commonly used internal classes' to his list of things to do with pentagram.
[13:01:48] <CashAlmostZzz> I'm still a fan of opengl - although so called gamers who think they know everything - eh what ever! seem to think it sucks now and its going out of phase hehe
[13:02:08] <CashAlmostZzz> hey everything has its purposes
[13:06:31] <Colourless> committed the new code
[13:07:03] <Darke> Since I'm going to age a fixedfont.cfg file for that font, shall we shuffle the fixedfont.tga into a /fonts/ dir, and drop the .cfg file in there with it? Or are we still up in the air as to how the data directory is going to work? *grin*
[13:08:45] <Colourless> uh, no idea
[13:10:38] <Darke> Umm... anyone got a coin? *grin*
[13:12:25] * Darke thinks he'll probably stay with the current way of doing things, at least for the moment anyway. *grin*
[13:15:34] --- CashAlmostZzz is now known as CashZzz
[13:16:56] <Darke> Of course now I'll be able to blame Colourless for his indecicivness when 6 months from now we've got an incredibly cluttered /data/ and have to rearrange it. *wizenod* This is what's known as 'delegation of responsibility'. *grin*
[13:17:55] <Colourless> i enabled fullscreen mode in pentagram and it runs at 75 fps here :-)
[13:18:34] <Darke> Yay!
[13:18:38] <CashZzz> ahahh thats what I was waiting for!
[13:18:43] <CashZzz> I was kinda right
[13:19:03] <Darke> Ah well, you'd best disable it now. Can't have actual performance happening in a pre-alpha engine can we? It'd set a precidence.
[13:19:07] <CashZzz> Colourless do you have a 32 mb second generation 3d card
[13:19:07] <CashZzz> like a 3d pro savage 4
[13:19:27] <Colourless> i have a 128 MB ATI Radeon 9700 Pro :-)
[13:19:35] <CashZzz> ohh
[13:19:50] <Colourless> i had actually already written the code for fullscreen just you couldn't actually enable it
[13:19:58] <CashZzz> man have you spent money
[13:20:04] <CashZzz> hmm 75 fps - what res
[13:20:08] * Darke has it's compatriate 128Meg GF4, but that's in his games machine. *grin*
[13:20:36] <Colourless> 640x480x32 probably
[13:20:57] <CashZzz> ok.
[13:22:19] <CashZzz> when full screen is added as offical I should get close to 75 fps 640x480 on my comp - so that would be interesting to see oneday, optimization wise
[13:22:45] <CashZzz> I only got a pro savage 4 (s3)(
[13:22:53] <Colourless> get about 80 FPS in 16 bit
[13:22:55] <CashZzz> everything good here apart from the graphics card
[13:23:06] <CashZzz> nice!
[13:23:07] <wjp> 55-60 fps windowed here
[13:23:13] <Darke> Hmm... I wonder what performance is like at 1280x1024. You'd need around that if you wanted to play tgwds with a 2x scaler.
[13:23:17] <wjp> sort 3, paint 5, time 17
[13:23:42] <wjp> why would you want a 2x scaler?
[13:23:54] <CashZzz> man yeah the tgwds e.g. crusader - lot lots lots more animation
[13:24:06] <Darke> wjp: "Because." *grin*
[13:24:46] <Darke> wjp: Admittedly though, I'd probably play them with an increased area, rather then high scaler. Especially with tgwds since you really need it. *grin*
[13:27:38] <wjp> running fullscreen keeps it at the same framerate
[13:28:33] <Colourless> that's because it's lacking an optimization
[13:29:37] <Colourless> which i will commit in a minute
[13:30:32] <wjp> cool :-)
[13:30:47] <wjp> anyway, 60fps should already be sufficient :-)
[13:31:06] <wjp> (of course game logic should make it quite a lot slower)
[13:31:16] <CashZzz> yeah 60-75 is good enough for non 3d engine
[13:31:36] <CashZzz> hmm well depends how you define 3d, non quake style engine games
[13:31:53] <wjp> 60-75? 25 is enough :-)
[13:32:02] <wjp> for some games 10 or even 1 :-)
[13:32:33] <CashZzz> ok sorry! ur quite right! -
[13:33:30] <CashZzz> well 60 if you were in a real high resolution hehe, anyways no bodys gonna want to use pentagram higher than 800x600, maybe not even higher than 640x480
[13:33:51] <CashZzz> seeing as though the shapes were designed with 320x200 in mind
[13:34:03] <Colourless> committed
[13:36:28] <CashZzz> tgwds - crusader series!! that was 640x480 wasnt it?
[13:37:22] <Colourless> 640x400
[13:39:01] <CashZzz> ok
[13:39:24] <CashZzz> yeah it didnt quite look 64x480 just from looking at it
[13:39:47] <CashZzz> but I thought it was
[13:42:20] * Darke gets about 40fps in windowed.
[13:42:58] <Colourless> wasn't likely to be much difference windowed
[13:43:06] <Colourless> fullscreen should be faster but quite a bit
[13:45:24] <wjp> they're both about 55-60 fps for me
[13:45:59] <Colourless> heh, i would have thought fullscreen would have been faster than that.
[13:46:18] <Colourless> i changed it from software single buffered to hardware double buffered
[13:49:05] <Darke> No difference here either. Not even hitting the refresh rate of 60hz, still only getting abour 40fps.
[13:50:07] <Colourless> my guess would be you aren't getting a hardware surface being created
[13:50:54] <Darke> Makes sense. Any indication of which in the console logs?
[13:51:06] <CashZzz> somthing to do with the os or maybe even a graphics card fps limiter
[13:51:24] <Colourless> no indication of it
[13:51:56] <CashZzz> I get 640x480x16 bit 75 fps with a program for example, and a friend of mine only gets 40 fps - hes got a better graphics card and same enviroment!
[13:52:18] <CashZzz> someone told me that some newer 3d accel's might have a hardware fps limiter
[13:52:31] <CashZzz> for certain things
[13:53:13] <CashZzz> guiess for winamp if anyone wanted to know - directx3 it was writte for
[13:53:23] <CashZzz> we were even running the same ver of directx! he even has a better motherboard and cpu than me
[13:53:54] <CashZzz> so that's kinda odd really
[13:53:58] <Colourless> shouldn't. but vsync can slow things down for no obvious reason. example if you are running at a refresh rate of 85 hz, but you can't render that fast and have vsync enabled, you'll only be able to render at 42.5 fps
[13:54:09] <CashZzz> oh well I have to go
[13:55:38] <Colourless> ah ha. enabling vsync in my ATI control panal decreased my fps to... just over 40 fps :-)
[13:55:47] <CashZzz> hmmm so that would be the vsync directx option in windows and not the actual prog written by the programmer - cause there is no control for vsync in guisss
[13:56:00] <CashZzz> ok I get ya! interesting and makes perfect sence!
[13:56:21] * Colourless turns vsync back off
[13:57:17] <CashZzz> thanks for point that out!
[13:57:24] <CashZzz> ow I really have to go
[13:57:27] <-- CashZzz has left IRC ()
[14:02:29] <Darke> Hmm... so your options are 8x8 or 16x16 if you want a fast blitting font.
[14:02:43] <Colourless> yes
[14:03:11] <Colourless> but it 'may' not work as expected yet
[14:03:25] * Darke ponders how much slower a non-8/16 font might be. Might be worth checking out for pedantic's sake. *grin*
[14:03:39] <Darke> What part of it 'may' not work? *grin*
[14:03:42] <Colourless> console assumes 8 pixel font i think
[14:04:17] <Darke> Yeah. DrawConsole has a constant of '8' in a couple of places.
[14:04:28] <Colourless> and >> of 3
[14:04:39] <Colourless> and << 3 too
[14:05:01] <Colourless> and in Console::CheckResize() too
[14:05:25] <Darke> Hmm... those variables really should be passed to PrintCharFixed anyway.
[14:06:07] <Darke> Either that or we just do away with the entire thing and use freetype with truetype fonts. *flee!*
[14:06:38] <Colourless> well, it's not exactly hard to change
[14:07:15] <Colourless> make all the << 3 into * font_size, all the >> 3 into / font_size, and all the various constant 8 to font_size
[14:07:32] <Colourless> then add font_size to the console class which matches the size of the font
[14:07:49] <wjp> hm... yes... that would explain why I'm getting almost exactly 42-43 fps in fullscreen dga mode
[14:07:57] <wjp> (vsync)
[14:08:41] <Darke> Colourless: *nod*
[14:09:08] <Colourless> of if you wanted to, have char_width and char_height
[14:09:50] <Darke> Was what I was considering doing, yes. *grin*
[14:09:59] <Colourless> you are correct in char_width and char_height should be passed to the PrintTextFixed and PaintCharFixed functions
[14:10:35] <Colourless> or there should be a TextureFont class that has that info into
[14:11:14] <Colourless> there is also going to be Shape based Fonts too
[14:11:41] <Colourless> eventually :-)
[14:13:32] * Darke was considering making a TextureFont class, but decided against it on the basis that he's way to tired to get it right. *grin* Will just commit the additional font.cfg stuff and fall asleep. (AKA, sit on irc and ramble incoherently for a couple of hours *then* fall asleep.)
[14:15:54] <Colourless> :_)
[14:15:56] <Colourless> :-)
[14:16:24] * Darke needs a larger supply of duct tape if Colourless is going to break his nose so much.
[14:17:28] <Colourless> yeah i know
[14:17:45] <Colourless> i keep pulling it off too, just terrible :-)
[14:18:56] * Darke tisktisks.
[14:20:43] * Darke duct tapes Colourless' horns together, then duct tapes a small fishing rod between them, hanging out above his muzzle, then attaches a carrot to the end of the fishing line. Vola! Instant perpetual motion dragon!
[14:20:59] <Darke> Hmm... at least it will be if said dragon likes carrots anyway.
[14:21:45] <Colourless> hmm
[14:22:30] <Darke> Y'see, *this* is the sort of quote from me you should stick in the quotes file. Rather then quoting me out of context and making me sound perverted. *nod*
[14:22:51] <Colourless> but this isn't actually funny
[14:25:07] <Darke> Hey! I think it's funny! Though apparently you thought the other one was funny too. *grin*
[14:40:02] <Darke> Besides, all these quotes make me look perverted. And as everyone knows I'm a *complete* pervert. *wizenod*
[14:40:10] <Darke> Eep! Innocent! I mean *complete* innocent!
[14:40:27] <Colourless> proof at last
[14:41:37] * Darke grumbles. Obviously he should have been already asleep if he's made that large a mistake.
[14:43:07] * Darke denies any and all perversions, like enjoying working on pentagram! *grin*
[15:00:47] * Darke wonders if adding a flag called --enable-random-optimisations to the configure script would elect more confidence in our coding abilities?
[15:01:20] <Colourless> hmm... and what are these random optimisations?
[15:01:57] <Darke> Just amazingly complicated stuff, like y'know, adding an -O3 to the gcc compiling command. *grin*
[15:01:59] <wjp> only draw each item with a 50% chance? :-)
[15:02:27] <Darke> Ooooh! That would be good! It'd be fun to try and pick up an item though. *grin*
[15:03:51] <Colourless> he never said anything about dropping items from the trace
[15:07:29] * Darke thinks 50% chance of correctly executing usecode functions would be fun too.
[15:08:13] <Colourless> i think that should also include spawn opcodes too
[15:10:47] <Darke> I think we can probably remove the: "this can take awhile, so sit tight..." comment from ./bootstrap. It takes longer to do a 'make clean' then it does a bootstrap. *grin*
[15:12:10] <wjp> really? make clean is nearly instantaneous here
[15:12:19] <wjp> bootstrap takes a couple of seconds
[15:12:58] <Darke> Curious. Both take about the same time here.
[15:14:20] <Dark-Star> aargh, MSVC complains about the line 'void GameMapGump::Paint(RenderSurface */*surf*/)' (only a warning, though...)
[15:14:55] <Colourless> you can ignore that... it doesn't do that here
[15:15:07] <Darke> Eh? What's it complaining about?
[15:15:17] <Dark-Star> yes, I know, I was just wondering why MSVC is so picky about some details ...
[15:15:28] <Dark-Star> it says "*/ found outside a comment"
[15:15:32] <Dark-Star> *g*
[15:15:35] <Colourless> it's a bug as far as I can tell in msvc
[15:15:40] <wjp> lol
[15:15:50] <wjp> insert a space if you have to :-)
[15:16:08] <Colourless> that is a space in the */*
[15:16:12] <Colourless> make it * /*
[15:16:36] <Darke> Heh. I actually added those comments to remove warnings about 'unused variables'. What version of MSVC are you using?
[15:16:44] <Colourless> i'm using 6
[15:17:11] <Colourless> you shouldn't get unused variable warnings on unused args
[15:17:34] <Dark-Star> I'm using MSVC7... let me try and remove that comment...
[15:18:11] <Dark-Star> MSVC7 doesn't give me an "unused variable" warning
[15:18:29] * Darke thinks you should, but he's weird. *grin* gcc agrees with me anyway.
[15:18:30] <Colourless> doesn't give me an unused variable warning here either
[15:21:01] <Dark-Star> has the game window always been resizable or is that a new 'feature'? It doesn't seem to be supported though (i.e. garbage if you make it bigger)
[15:21:41] <Colourless> hmm, yes that would be unsupported
[15:34:34] <-- Kirben has left IRC (Read error: 104 (Connection reset by peer))
[15:34:36] <Colourless> i've got to go
[15:34:39] <Colourless> cya
[15:34:41] <-- Colourless has left IRC ("casts invisibility")
[15:34:46] <Darke> Bye!
[17:20:48] * Darke yawns and finally slips off to sleep. Night!
[17:20:51] --- Darke is now known as DarkeZzz
[19:51:07] --> Dominus has joined #pentagram
[21:44:01] <-- Dominus has left #pentagram ()
[21:54:31] <-- wjp has left IRC ("Zzzz...")