00:00:03 | Xentac | slow down big spender... |
00:00:18 | scanline | those conductive pens are like $10! |
00:00:50 | Xentac | hehehe |
00:02:02 | carpman | oh wow |
00:02:12 | carpman | this is amazing |
00:02:13 | scanline | hi carpman |
00:02:29 | Xentac | carpman: are you high? |
00:02:38 | carpman | scanline: if I shelve a python object, all the references to other objects that that object contains are also shelved. |
00:02:48 | carpman | scanline: I can store a whole object tree in one line. |
00:03:21 | scanline | shelved? |
00:03:25 | scanline | doesn't know much python... |
00:03:47 | carpman | scanline: yeah, it stores python objects to the disk, and has an interface like a dictionary. |
00:04:18 | scanline | oh. serialization |
00:04:21 | carpman | yeah |
00:05:33 | scanline | spiffy |
00:14:58 | file | mmm |
00:17:10 | Xentac | hey kergoth_ |
00:17:12 | Xentac | guess what? |
00:17:27 | file | coughs |
00:18:44 | Xentac | kergoth? |
00:18:56 | kergoth | hey |
00:19:01 | Xentac | guess what |
00:19:45 | kergoth | what? |
00:20:25 | Xentac | I went to do an 'ipkg upgrade' this morning... and it stopped... so I debugged it... and it turned out that all of the upgrades were being installed to / instead of /mnt/ram or /mnt/card, etc |
00:20:30 | Xentac | so now my flash is all used up |
00:20:39 | Xentac | and I have to sift through ipkg droppings |
00:21:06 | kergoth | ipkg doesnt handle package installed to alternate dests |
00:21:14 | kergoth | use oipkg, tap on them one by one, to do it properly |
00:21:24 | Xentac | I realize that |
00:21:27 | Xentac | ok... |
00:21:29 | Xentac | but you said... |
00:21:30 | Xentac | last time... |
00:21:36 | Xentac | that ipkg upgrade would handle that |
00:21:41 | kergoth | i thought it would |
00:21:43 | kergoth | turns out it doesnt |
00:21:45 | Xentac | gah! |
00:21:45 | kergoth | as you found out |
00:21:48 | kergoth | hehe |
00:21:53 | Xentac | and now my flash is all used up |
00:21:56 | carpman | chuckels |
00:21:58 | Xentac | pokes kergoth... hard |
00:22:03 | carpman | err... chuckles |
00:22:07 | carpman | or something like that. |
00:22:17 | Xentac | chuckels... that's like... french... right? |
00:22:23 | Xentac | ;o) |
00:22:44 | carpman | whatever, I think my brain went off duty for the night. |
00:22:46 | Xentac | anyway... I have to go get beat up now... but when I get back I have to fix my Zaurus |
00:22:59 | kergoth | Xentac: training? |
00:23:15 | carpman | wonders why samba wants libpng and libjpeg |
00:23:28 | kergoth | carpman: hm, that seems odd |
00:23:41 | Xentac | kergoth: correct... |
00:23:43 | carpman | yeah. it does. |
00:23:54 | Xentac | carpman: hmmmm... so that is can render jpegs and pngs? |
00:24:01 | Xentac | s /is/it |
00:24:01 | carpman | thinks he will stop this buuld before it installs X or something... |
00:24:07 | Xentac | hehehe |
00:24:20 | Xentac | it's probably trying to install an X based samba client... |
00:25:45 | Xentac | kergoth: you see... when I say I'm being beaten up it means that I'm going to a class that night... |
00:25:57 | Xentac | if I say I'm beating people up, it means I'm teaching |
00:25:58 | Xentac | ;o) |
00:26:01 | kergoth | hah |
00:26:03 | Xentac | is away: being beat up |
00:26:03 | kergoth | makes sense |
00:26:06 | kergoth | have fun |
00:26:14 | Xentac | ta ta |
00:27:34 | scanline | hmm.. looks like I don't have any more homework due this week |
00:28:00 | scanline | verifies |
00:28:29 | scanline | yup |
00:28:36 | scanline | time for some hot textbox widget action! |
00:28:39 | scanline | after some coffee... |
00:29:07 | kergoth | nice |
00:29:14 | Xentac | comes back for a sec... |
00:29:25 | Xentac | scanline: woohoo hot action! that's my phrase! |
00:29:31 | scanline | hehehe |
00:29:34 | scanline | Xentac: you influenced me :) |
00:29:44 | scanline | anyway... maybe I'll have some hot openzaurus action too ;) |
00:29:56 | scanline | kergoth: does openzaurus support make -j yet? =) |
00:30:25 | kergoth | scanline: good question |
00:30:31 | scanline | I'll have to test that |
00:31:27 | Xentac | scanline: just wait till you are heard saying, "hot buttery goodness, from between your legs"... then you'll know you were really influenced by me |
00:31:35 | Xentac | notes there's only really one safe time to say that |
00:31:39 | scanline | yeep |
00:32:09 | Xentac | when someone's eating popcorn |
00:32:13 | Xentac | anyway... |
00:32:14 | Xentac | is back (gone 00:06:11) |
00:32:15 | carpman | Xentac: arent you supposed to be away beating people up? |
00:32:19 | Xentac | is away: getting beat up |
00:32:26 | Xentac | carpman: no no... tonight /I/ get beat up |
00:32:38 | carpman | looks at above butter comment |
00:32:42 | carpman | Xentac: even better |
00:32:51 | Xentac | doesn't have to leave till like 7:00... but likes to get there early |
00:32:59 | Xentac | carpman: you didn't get it? bah |
00:33:04 | Xentac | leafs. |
00:34:04 | carpman | http://slashdot.org/articles/02/09/11/230203.shtml?tid=137 |
00:34:07 | carpman | Wow, I want one! |
00:34:14 | carpman | It can be my lair! |
00:34:29 | file | O.o |
00:37:32 | file | mmm |
00:44:59 | scanline | https://navi.picogui.org/public.5479/media/images/art/phased_spacebug.jpeg |
00:45:16 | scanline | ^ motion blur composited with the original :) |
00:45:42 | file | pokes scanline a bit |
00:46:11 | file | must g |
00:46:13 | file | er go |
00:51:42 | carpman | you know, now that I learned how to install gentoo without destroying my home dir, I'm REALY tempted to install it on mach5 |
00:52:10 | scanline | hehe |
00:52:21 | scanline | I'm kinda tempted to install it on yoshi.. though I'm happy with debian |
00:52:40 | carpman | dreams of dumping in ALL of the optimization flags into make.conf |
00:55:20 | carpman | I have to say, lain runs alot faster now |
01:02:54 | carpman | well, good night. |
03:39:41 | Xentac | is back (gone 03:07:22) |
03:41:34 | scanline | feels like taking a break from textbox to make some eye candy |
03:42:30 | Xentac | gee... |
03:42:40 | Xentac | I can handle 1 set of 500 kicks... |
03:42:41 | Xentac | but two sets... in 2 nights... |
03:42:42 | Xentac | ugh |
03:45:52 | Xentac | anyone here to hear my suffering? |
03:47:20 | scanline | nods |
03:47:31 | hikke | nods too. |
03:47:32 | Xentac | oh good... at least someone is... |
03:47:38 | Xentac | suffer suffer... |
03:47:41 | Xentac | but yeah... |
03:47:50 | Xentac | we did 500 kicks again tonight... |
03:47:54 | Xentac | it's more like 400... but still... |
03:48:08 | scanline | who did you kick? :) |
03:48:27 | Xentac | no one... |
03:48:34 | Xentac | and then we did sparing... and kata... |
03:48:46 | Xentac | needs to do some of the (kick)boxing classes |
03:49:01 | Xentac | after all... if I'm not dancing friday nights... what better to do than get your ass kicked more? |
03:49:57 | scanline | hehe |
03:50:21 | Xentac | at least then I'll be able to take a hit a lot better |
04:01:15 | Xentac | resolves to have a working Zaurus before tomorrow. |
04:04:09 | Xentac | scanline: if you have a binary, or even better a tarball, of some Picogui stuff I'll put it on there too |
04:06:10 | scanline | there's that old tarball, and some not-quite-working ipks |
04:06:19 | scanline | needs fixing |
04:06:28 | Xentac | wanna make a special -xentac.tar.gz ? ;o |
04:06:31 | Xentac | er ;o) |
04:06:41 | scanline | not really :) |
04:06:48 | scanline | I'll fix the openzaurus stuff soon |
04:06:48 | Xentac | oh come on... it'd be cool... |
04:07:16 | Xentac | ok... |
04:10:49 | Xentac | and what does that mean to me |
04:10:50 | Xentac | ? |
04:11:10 | scanline | nothing? |
04:11:15 | Xentac | oh, ok... |
04:11:16 | scanline | i dunno |
04:11:20 | Xentac | so... when do I get a tarball...? |
04:11:21 | Xentac | hehehe |
04:12:15 | Xentac | almost has his Zaurus in tip-top working condition... |
04:17:07 | Xentac | what the hell? |
04:17:13 | Xentac | will have to have a little talking to ipkg... |
04:19:52 | Xentac | woah... it might have actually worked... |
04:27:47 | Xentac | yay... now to actually do the upgrading |
04:28:43 | Xentac | learns a lot about ipkg and it's file formats... yay for him... |
04:31:38 | Xentac | geez... the symlinks take up more space than the actual programs (from what I can tell here...) |
04:32:12 | Xentac | doh |
04:32:14 | Xentac | nevermind... |
04:32:22 | Xentac | some things didn't disappear when they were supposed to... |
04:32:36 | Xentac | fun... time to hunt for files... |
04:32:56 | Xentac | actually... one chance... |
04:58:51 | Xentac | woohoo |
04:59:01 | Xentac | the vnc viewer works now :) |
04:59:05 | Xentac | tests it with his server... |
04:59:10 | Xentac | it's purdy |
05:01:41 | Xentac | hmmm... I seem to have a problem though... |
05:01:46 | Xentac | no right click ;o) |
05:24:15 | Xentac | damn... |
05:24:24 | Xentac | I got the OZ vncserver working... |
05:24:30 | Xentac | giggles like a schoolgirl |
05:32:35 | Xentac | except the keyboard... but I'm still working on it... |
05:32:41 | Xentac | I think this is a working state... |
05:32:45 | Xentac | I'm ready to go to bed soon |
05:35:27 | Xentac | woohoo! keyboard! |
05:39:53 | Xentac | sweet... and it'll work for picogui too... hehehe |
05:53:31 | Xentac | is away: about to go to sleep |
08:57:15 | scanline | is away: Sleep, that's where I'm an COBOL and Object Pascal programmer! |
12:26:44 | Xentac | is back (gone 06:33:12) |
13:58:23 | Xentac | is away: werk |
14:23:05 | gonkulator{work} | wb carpman |
14:34:24 | carpman | hi gonkulator |
14:36:16 | carpman | PicoBot: seen gonkulator{work} |
14:36:16 | PicoBot | gonkulator{work} was last seen on #picogui 13 minutes and 11 seconds ago, saying: wb carpman [Thu Sep 12 09:29:19 2002] |
16:03:39 | carpman | hi scan[ibook] |
16:03:47 | carpman | I looked at the app.confs |
16:04:04 | carpman | basePath=/bin/ is the problem I think |
16:04:25 | carpman | basePath needs to be the path to the application's dir. |
16:04:29 | scan[ibook] | hi carpman |
16:04:33 | scan[ibook] | which app.conf had that? |
16:04:53 | carpman | scan[ibook]: I know tpcal |
16:05:01 | carpman | and canvastiles |
16:05:06 | scan[ibook] | k |
16:05:29 | carpman | atomicnav is good |
16:06:24 | carpman | gridgame is b0rken |
16:07:14 | carpman | yeah, those 3 are it. |
16:12:32 | jimlay | you guys should advertise more. :) |
16:12:42 | jimlay | are you guys using C? or c++ |
16:12:48 | scan[ibook] | jimlay: yeah, we should :) |
16:12:54 | scan[ibook] | jimlay: so far, just C |
16:13:03 | scan[ibook] | jimlay: the server code is all in C, client code can be in any language |
16:13:05 | jimlay | :) <- josie is happy. |
16:13:45 | jimlay | of course, I'm interested. I looked around the site a bit, and notice the design document was in german (my german isn't perfect yet. :) but I didn't see any. |
16:14:06 | jimlay | Where can the source be dl'd? Is there an anonamous cvs server? |
16:14:44 | jimlay | only found out you exist because he has been haning out in #fresco a lot lately. |
16:14:56 | scan[ibook] | ohh... that thing that hunger sent me a link to |
16:15:01 | scan[ibook] | that's not mine, it just mentions picogui |
16:15:08 | scan[ibook] | try http://picogui.org |
16:15:27 | jimlay | ohhh... hmm interesting. that looks much more promising. |
16:15:42 | jimlay | wonders why hunger sent me to that page. |
16:16:21 | scan[ibook] | hehe |
16:16:37 | jimlay | nm.. just temporar loss of my sanity. Hunger sent me to the right place I just got confused. |
16:18:42 | jimlay | are you using binary sockets or text sockets for the protocol? |
16:19:02 | jimlay | nm.. I should just shut up |
16:19:13 | scan[ibook] | binary... |
16:19:20 | scan[ibook] | there's a document about the protocol up on picogui.org |
16:19:22 | jimlay | and read all your docs before I bombard you with questions right. :) |
16:19:27 | jimlay | I know I was reading it. |
16:19:29 | scan[ibook] | hehe |
16:19:35 | jimlay | lol that's why I said I should just shut up. |
16:28:23 | jimlay | do you have any openGL support. |
16:28:50 | scan[ibook] | yes |
16:29:01 | scan[ibook] | it's rather experimental, but it's there |
16:29:55 | hikke | sdl-apps? |
16:29:56 | jimlay | ... if your curious. I am busy hacking X out of xfree86 so that's my relation fresco, and ggi and anyone else who might want a good set of hardware accelerated drivers to run a gui on. |
16:30:13 | scan[ibook] | oooh |
16:30:23 | scan[ibook] | OpenGL acceleration without X? =) |
16:30:25 | jimlay | you say that like it's a bad thing. |
16:30:33 | scan[ibook] | no, I'm happy |
16:30:42 | jimlay | well, it will still be xfree86, just there wont be X |
16:30:44 | jimlay | so yeah. |
16:31:00 | scan[ibook] | so something like DirectFB but more complete acceleration, and 3D support too? |
16:31:10 | jimlay | you'd have to do your own event handling and font rendering and stuff. But you'd have a hardware accelerated fb. |
16:31:12 | jimlay | that's the idea. |
16:31:21 | scan[ibook] | that's exactly what I want :) |
16:31:26 | jimlay | But I'm still trying to figure out the scope of it. |
16:31:31 | jimlay | It's what everybudy wants. |
16:31:40 | scan[ibook] | I haven't gone digging in the X source code much |
16:31:42 | jimlay | GGI, Fresco you, me lots of people. |
16:31:49 | jimlay | I have noticed. |
16:31:57 | jimlay | It's hard on your head. |
16:32:14 | scan[ibook] | actually.... maybe you could make a library that dlopen's XFree86 and uses it like a puppet :) |
16:33:22 | jimlay | 150mb of cross hatched undocumented global variables. located in 100 thousand differnet files with 100,000 + different types defined. |
16:33:28 | jimlay | you read my mind. :) |
16:33:42 | scan[ibook] | hehe |
16:33:51 | scan[ibook] | libXfrankenstein.so |
16:33:54 | jimlay | of course there will be a xfree86.so binary in /usr/bin |
16:34:01 | jimlay | LOL |
16:34:34 | jimlay | as near as I can tell, XFree86 should hav edone this anyways, it would have made there code better from there perspective too. |
16:34:44 | scan[ibook] | nods |
16:35:00 | jimlay | it would simply be a clean implementation if there dix (device independant layer was separated better from X than it is. |
16:36:05 | jimlay | (what I think is funy, is that to compile a binary hardware driver (like nvidia_drv.so) you have to have the X11 headers in the include path. |
16:36:29 | jimlay | I mean why are is the X protocol headers in the hardware driver? |
16:36:42 | scan[ibook] | hehe |
16:37:12 | jimlay | It's really not using htem though, it's just that they were to lazy to separate there include files well, so xf86.h includes the whole friggin world. |
16:37:49 | jimlay | is the actual design of your server documented somewhere? |
16:38:07 | jimlay | I'm looking at picogui_arch.png but all that shows me what's inscope where. |
16:38:34 | scan[ibook] | that diagram is pretty old |
16:38:36 | carpman | scanlines_brain.png |
16:38:38 | scan[ibook] | there's no real design doc |
16:38:40 | scan[ibook] | heh, yeah |
16:39:06 | jimlay | Like I'd assume you don't do poling on your input devices, but to not to that, you eather have to sleep on input from a particular device, multi thread, or do something I don't know about. |
16:39:32 | scan[ibook] | it's a select() loop |
16:39:46 | jimlay | but what is the function in the select? |
16:39:51 | jimlay | a sleep till input? |
16:40:02 | scan[ibook] | input drivers have functions that register themselves in the fd_set, and another function called after the select to determine if they were activated |
16:40:17 | jimlay | (btw that's exactly what xfree86 does.) select( WaitForSomething()) { |
16:40:49 | jimlay | so you do something like poll them once every 100microseconds? |
16:40:55 | scan[ibook] | all the drivers and all the sockets register themselves in select(), plus there's a timeout to handle things like blinking cursors and scrolling. So it just waits until network input, driver activity, or a timeout |
16:40:57 | jimlay | and just run a delay loop like that? |
16:41:08 | carpman | Not everything Xfree86 does is bad. Its just bloated, and needs major rewriting in the gfx department. |
16:41:21 | scan[ibook] | it's possible to either poll or wait for a fd in the input drivers |
16:41:24 | scan[ibook] | most drivers wait for a fd |
16:41:31 | scan[ibook] | some that can't, like sdlinput, have to poll |
16:41:40 | jimlay | xfree86 has the best hardware support out there if that's what you mean by gfx. |
16:42:01 | carpman | no, I mean the code that does the drawing, not the hardware department. |
16:42:01 | jimlay | (of any windowing system) but personally I think the protocol is way outdated. |
16:42:53 | jimlay | yeah... there are 2 directories in the main directory that havn't been touched in like 10 years, and they handle all the drawing. |
16:42:55 | jimlay | :) |
16:43:06 | scan[ibook] | hehe |
16:43:11 | jimlay | scanline: sorry you lost me, fd? |
16:43:17 | scan[ibook] | jimlay: file descriptor |
16:43:45 | scan[ibook] | jimlay: include/pgserver/input.h defines the interface, a struct inlib |
16:44:01 | jimlay | but if a driver does something like getc(/dev/mouse) the execution halts till it gets something. |
16:44:12 | scan[ibook] | right.. so that's not what the driver does :) |
16:44:32 | scan[ibook] | the driver has an fd_init function that adds its file descriptors and/or timeouts to the set of stuff to be select()'ed |
16:44:57 | scan[ibook] | then its fd_activate is called afterwards to detect if that driver's fd was activated, and if so it's safe for the driver to read without blocking |
16:45:02 | jimlay | ok. So then you have a way of polling. |
16:45:08 | scan[ibook] | it's not polling |
16:45:17 | scan[ibook] | pgserver will stay sleeping until some device has input |
16:45:30 | jimlay | "read without blocking"? |
16:45:59 | scan[ibook] | that's what select() does, it waits until a set of file descriptors has something available to read or write |
16:46:46 | jimlay | 's eye's pop out at man select. |
16:46:50 | scan[ibook] | hehe |
16:46:59 | jimlay | I thought you ment select as in switch. |
16:47:05 | scan[ibook] | ohh |
16:47:22 | jimlay | for somereason it didn't register that select was a syscall that would let you sit on multiple file handles. |
16:47:45 | jimlay | I was going to say you sound like your doing something like waitfor(getc(mouse) | getc(keyboard)) |
16:47:52 | scan[ibook] | hehe |
16:47:53 | jimlay | but that would require threading. |
16:48:02 | scan[ibook] | yeah. pgserver isn't threaded (for compatibility) |
16:48:07 | scan[ibook] | err, portability |
16:48:15 | jimlay | what's how I did my windowing kernel. |
16:48:26 | jimlay | was with pthreads. |
16:48:42 | scan[ibook] | threading is nice on big systems, but on small systems the overhead is too much |
16:48:51 | scan[ibook] | I might make pgserver optionally threaded at compile-time later |
16:49:14 | jimlay | to call it a windowing kernel is a bit of an overstatement, it's a kernel, w/ intended for use in a windowing systems some friends of mine want to put to gether for themselves. |
16:49:38 | jimlay | but what would you do with multiple threads? |
16:50:50 | scan[ibook] | generally makes the GUI more responsive |
16:51:01 | scan[ibook] | one big drawing operation can't block other things from happening |
16:51:13 | jimlay | I see. |
16:51:38 | scan[ibook] | not a high priority in picogui though.. small systems would be bogged down by threads, large ones you don't notice a problem |
16:51:39 | jimlay | do you know if there is a windowing system out theree that does backing store on windows as they move off screen? |
16:51:50 | scan[ibook] | MacOS X's Quartz does that |
16:51:57 | scan[ibook] | eats a ton of memory |
16:51:59 | jimlay | ? really? cool. |
16:52:16 | jimlay | well it shouldn't be too bad if only cache the applicatio for as long as it's in focus. |
16:52:41 | jimlay | so like if you pick a window up and slide it and then slide it back on, it's doesn't have to redraw. |
16:53:04 | jimlay | But it's fine if a program has to redraw after a decent period of no usage. |
16:53:08 | scan[ibook] | yeah. That's one of the things I wanted to achieve in picogui, but I did it without buffering everything :) |
16:53:40 | jimlay | so picogui does this too? |
16:53:58 | jimlay | how does it run at 1600x1200 w/sdl? |
16:54:01 | scan[ibook] | sort of... it buffers everything at the primitive level, instead of as a pixmap |
16:54:04 | jimlay | it can't be all that smooth. |
16:54:17 | jimlay | so it does buffer _everything_ |
16:54:25 | scan[ibook] | it isn't smooth at high resolutions yet, unless you use OpenGL or something. there's no support for 2D acceleration |
16:55:02 | scan[ibook] | jimlay: it's not really like a window buffer though.. it just keeps a list of the primitives to draw for each divnode so the server can very quickly redraw areas of the screen |
16:55:11 | jimlay | do you mean there is no support for 2d acceloration? or none of your console/targets support hw accel? |
16:55:44 | scan[ibook] | picogui has great acceleration support in theory, but there are only two accelerated drivers (x11 and opengl) and they're both pretty experimental |
16:55:58 | scan[ibook] | I want to add a DirectFB driver, and this X without X thing would be great too :) |
16:56:11 | jimlay | I see. |
16:56:14 | scan[ibook] | picogui's video layers are pretty flexible |
16:56:27 | scan[ibook] | there's a very general driver that you layer more specific drivers on top of |
16:56:33 | scan[ibook] | so you end up with a chain of drivers, like: |
16:56:54 | jimlay | are they all dynamically loaded? |
16:56:57 | scan[ibook] | general driver -> 16-bit framebuffer -> linux framebuffer -> 90 degree rotation |
16:57:13 | scan[ibook] | they are dynamically loaded, but not dynamically linked (yet) |
16:57:40 | jimlay | hasn't quite got wich one is which strait yet. |
16:57:52 | jimlay | ld loaded? or dl loaded? |
16:57:59 | scan[ibook] | you can load and unload drivers at runtime, but they have to be compiled in |
16:58:04 | jimlay | ok. |
16:58:07 | jimlay | so no dlopen. |
16:58:16 | scan[ibook] | yeah. but it shouldn't be hard to add later |
16:58:28 | scan[ibook] | we're also planning on making widgets loadable with dlopen |
16:58:49 | jimlay | the kernel I wrote, has dlopen only, with a dynamic api, and all modules have to be hot swapable. |
16:59:04 | jimlay | but you don't have a dynamic api do you? |
16:59:31 | scan[ibook] | what do you mean by dynamic api? |
16:59:39 | jimlay | like if I wanted to write a module for picogui that was loaded in and extended the api, could I do this? with out having to recompile picogui? |
16:59:53 | scan[ibook] | not yet |
17:00:13 | jimlay | I mean the api (set of functions that are available) grows and shrinks dynamically. |
17:00:44 | jimlay | is your window manager part of the kernel? or a separate program? |
17:00:58 | scan[ibook] | it's part of the serve |
17:01:00 | scan[ibook] | err, server |
17:01:16 | scan[ibook] | it's planned to make that dynamic eventually |
17:01:36 | scan[ibook] | pgserver is kinda like the linux kernel before module support right now :) |
17:01:57 | jimlay | I See. because I like the idea of customizable windowmanagers, it's just that they should have to plugins for the server, not clients. |
17:02:08 | scan[ibook] | well, class is over.. be back soon |
17:02:17 | jimlay | ok, nice talking to you. |
17:02:24 | jimlay | class is over? are you in class? |
17:11:13 | scan[ibook] | back |
17:12:43 | jimlay | are you in school right now? |
17:12:53 | scan[ibook] | yes |
17:12:57 | jimlay | what school? |
17:13:06 | scan[ibook] | second year at the University of Colorado at Boulder |
17:13:29 | jimlay | do most people cary a laptop to class with them and sit on IRC ignoring the teacher? |
17:13:33 | jimlay | why do you go to class? |
17:13:44 | scan[ibook] | hehe.. I don't ignore the teacher :P |
17:48:59 | scan[ibook] | jimlay: keep me posted on your X without X project :) |
17:49:13 | jimlay | sure |
17:49:23 | jimlay | can you drag windows in pgui? |
17:49:30 | scan[ibook] | yeah |
17:49:34 | jimlay | I can't seem to move them. Is it supposed to be that way? |
17:49:41 | scan[ibook] | drag the title |
17:49:55 | jimlay | i might be slow but I'm not that stupid. |
17:49:58 | scan[ibook] | :) |
17:50:08 | jimlay | I think I'm using hte wrong applications, or something. |
17:50:32 | scan[ibook] | how did you run it? |
17:50:51 | jimlay | as just pgserver the first time, and this time as pgserver -nc skin/vr3.conf |
17:51:02 | jimlay | and then I go to a different terminal and start apps. |
17:51:23 | scan[ibook] | are you running it under X, with sdlfb? |
17:51:46 | jimlay | but like I run bouncyball and it pops up and bounces and everything works but I can't movei t. |
17:51:54 | scan[ibook] | oh |
17:51:57 | jimlay | I'd assume I'm under X with sdlfb |
17:52:00 | scan[ibook] | you can't drag popup boxes yet |
17:52:07 | jimlay | I'm in x and I'm using sdl |
17:52:15 | scan[ibook] | nods |
17:52:23 | jimlay | and I think it's sdlfb, but I don't know much about sdl yet at all. |
17:52:39 | jimlay | going to read about that today or tomarrow. Been trying to get it for a fewdays now. |
17:52:45 | scan[ibook] | most applications are on nonoverlapping panels, that you can drag to resize. popups are modal and not movable |
17:53:10 | scan[ibook] | sdlfb is a picogui driver that emulates various types of framebuffers on top of SDL |
17:53:23 | jimlay | can you drag side to side? |
17:53:27 | jimlay | or just up and down? |
17:53:49 | scan[ibook] | in the current window management, you can drag in one dimension and rotate the panel |
17:53:55 | scan[ibook] | so you can rotate it, then drag side to side... |
17:54:07 | scan[ibook] | some people like the current window management, but I'd like to have other styles as well |
17:54:15 | scan[ibook] | something like ion but more intuitive, for handhelds |
17:54:44 | scan[ibook] | maybe traditional overlapping windows for drivers that support that |
17:54:52 | jimlay | so you can't get the bottom of the window off the bottom of the screen? (which ever side that happens to be via rotation) |
17:55:02 | scan[ibook] | right |
17:55:20 | scan[ibook] | you can click the title to shade it |
17:56:03 | jimlay | I must say that of all the hand held windowing systems I've seen this is by far the coolest, and most usefull. |
17:56:15 | jimlay | But see I have a desktop. :) not a handheld. |
17:56:48 | scan[ibook] | yeah, right now picogui isn't suited for desktops, but that may change |
17:57:19 | scan[ibook] | once it has good acceleration support, multiple window management choices including overlapping, and a way to run X or Gtk/Qt apps in an emulation layer |
17:57:26 | scan[ibook] | then maybe it will be good for desktops |
17:58:24 | scan[ibook] | I'd also like to make picogui able to run rootless within some other GUI |
17:58:35 | scan[ibook] | that will be the most useful for making apps portable between handhelds and the desktop |
17:58:52 | jimlay | I don't know, if you do it right and have a way for gecko to link to the windowing system, so that it draws it's widgets directly to the video memory. It would be as fast as windows. This could be a basis for a lot of stuff, and with that alone you wouldn't need the X compatiblity, people would want native apps bad enough to write htem. |
17:59:18 | scan[ibook] | well, once we have loadable widgets we could have a gecko widget :) |
17:59:34 | jimlay | wegets... plural here. |
17:59:50 | jimlay | but yeah, it wouldn't suck like gecko does now on linux, |
18:00:26 | jimlay | I mean gecko is awsome, for rending stuff, but the proglem is X can't handle puting what it renders on the screen well enough. Wether that's geckos fault or X's I wont point fingers. |
18:00:53 | scan[ibook] | nods |
18:01:29 | scan[ibook] | I'd love to have a picogui system with good hardware acceleration, that can run emacs and gecko :) |
18:01:57 | scan[ibook] | man.. I'm hungry |
18:02:01 | jimlay | grumble.. emacs the people wrote the emacs gui.. they almost shouldn't have bothered. |
18:02:27 | jimlay | I actually never use it in a gui. It's slower and less convienient than it is in a terminal. |
18:02:36 | scan[ibook] | picogui's terminal emulator is still missing enough features that emacs won't run correctly |
18:02:45 | jimlay | If they hadn't bothered i wouldn't have to type emacs -nw every single time I start emacs. |
18:03:08 | jimlay | A really nice front end for emacs would be cool. |
18:03:36 | jimlay | Like if it was done in gecko, .. the way gecko works in ms windows. That would be a cool looking text editor. |
18:03:51 | jimlay | then you'd have the kitchen sink in a staneglass window. :) |
18:04:02 | scan[ibook] | haha |
18:04:07 | jimlay | is the nifty comm a real device? |
18:04:12 | scan[ibook] | no |
18:04:18 | jimlay | ok good |
18:04:28 | scan[ibook] | I've looked at the mozilla source a little, looks fairly easy to port |
18:04:53 | scan[ibook] | problem though is that most apps expect to be able to just draw stuff and have it appear onscreen |
18:04:57 | jimlay | I have not. I have looked for a spec on what needs to be done to port it, but I didn't find much. |
18:05:06 | scan[ibook] | that's possible in picogui, but then you lose most of picogui's benefits |
18:05:35 | jimlay | Is it all in the source? Considering there intentsions, I expected it to be fairly easy to port. |
18:06:00 | jimlay | Well, if gecko was built into the windowing system (interpret module) it could just "draw" on the screen right? |
18:06:13 | scan[ibook] | Yeah, there are just directories with classes to interface it to different GUIs and OSes.. they just implement drawing functions, font metrics, et al. |
18:06:38 | jimlay | how do they handle pixel formats? |
18:06:45 | scan[ibook] | not sure, I didn't get that far |
18:06:45 | jimlay | better question, how do you handle pixel formats? |
18:07:20 | scan[ibook] | outside pgserver, everything uses a "pgcolor" format which is just 24-bit RGB with some optional flags in the high bits |
18:07:37 | scan[ibook] | inside pgserver, everything's converted to a video-driver-specific format |
18:07:53 | scan[ibook] | so all the dithering, quantization, and color space conversion is done only at load-time |
18:08:04 | jimlay | (if I ask anyqyestions that are documented someplace, please send me to the docs. I hate it when people ask me questiosn that there are docs for (even more so if I've written them) |
18:08:13 | scan[ibook] | hehe |
18:08:23 | scan[ibook] | pgserver is mostly undocumented now, unfortunately |
18:08:53 | scan[ibook] | the protocol is pretty well documented, and there are docs for the C client lib, the python client lib, themes, and some other random stuff |
18:08:59 | jimlay | hmm... that seems like it's a good system interms of space, because it is not efficient for the binary to be prepaired to handle all the pixel formats. |
18:09:17 | jimlay | The problem is that it takes cpu time. |
18:09:26 | scan[ibook] | why? |
18:09:35 | jimlay | are your drawing functions drawing in the native pixel format? |
18:09:40 | scan[ibook] | yes |
18:09:42 | jimlay | like fill region stuff. |
18:09:45 | scan[ibook] | for example: |
18:09:57 | jimlay | ok so the only thing you have to blight are bitmaps then. |
18:10:05 | scan[ibook] | the client stores a rectangle primitive in a canvas widget, it gives the color as 0xFF0000 (red) |
18:10:17 | scan[ibook] | so the canvas widget immediately converts that into whatever the native pixel format is |
18:10:28 | scan[ibook] | then at render-time, there's no extra work to do |
18:10:42 | scan[ibook] | bitmaps are always stored in the native pixel format |
18:11:24 | jimlay | but for example, mozilla, would want to render a bitmap of the page evertytime, but that would be entirly in the pg pixel format, and you would have to blight it onto the screen. |
18:11:33 | jimlay | And that would be expensive for every frame. |
18:11:37 | scan[ibook] | oh |
18:11:44 | scan[ibook] | for stuff like that, there's shared memory |
18:12:04 | scan[ibook] | you can ask pgserver to convert a bitmap to a shared memory segment, and it will return information about the pixel format for the client to use |
18:13:21 | scan[ibook] | normally all the rendering is done server-side, so the client doesn't have to care what format the driver uses. this allows picogui to support weird formats, like ASCII or strange hardware specific formats |
18:13:57 | jimlay | hmm... |
18:14:01 | scan[ibook] | the shared memory method is less general, because it only works on a local pgserver, and it's not useful unless the driver uses a normal pixel format.. but it lets you do fast graphics |
18:14:06 | scan[ibook] | that's how the SDL support works |
18:14:18 | jimlay | so that part is implemented? |
18:14:22 | scan[ibook] | yes |
18:14:24 | jimlay | the shared memory thing. |
18:14:31 | jimlay | And the ability to get the pixel format, |
18:14:34 | scan[ibook] | picogui.org has some screenshots of it running |
18:14:42 | jimlay | what demo uses that? |
18:15:07 | scan[ibook] | there's a little "shmtest" demo, plus if you grab the latest CVS of SDL 1.2 you can compile it with picogui support |
18:15:18 | scan[ibook] | there's a picogui readme in the SDL CVS tree |
18:15:46 | jimlay | cool. |
18:16:33 | scan[ibook] | hmm... are there any licensing issues with making parts of X into a shared library? |
18:16:35 | jimlay | how is your server catching the segv's? |
18:16:46 | jimlay | nope. |
18:16:59 | scan[ibook] | SIGSEGV handler, in pgserver/gcore/signals.c |
18:17:08 | scan[ibook] | it tries to close down the video and input drivers cleanly |
18:17:16 | jimlay | X - free to manipulate, steal, rape, pillage, charge people for use of if you can trick them. |
18:17:40 | jimlay | So long as you include this "discalaimer: this product might not be y2k compliant" |
18:17:45 | scan[ibook] | hehe |
18:17:57 | jimlay | I know how handlers work, but the server is displaying the popup. |
18:18:15 | jimlay | Is this just the pgui client library catching the signal and causing hte popup? |
18:18:21 | jimlay | on the server? |
18:18:26 | scan[ibook] | oh |
18:18:38 | scan[ibook] | client segfaults are handled in the client lib, the client lib just creates a dialog box |
18:18:42 | jimlay | No XFree is on the MIT liscense - it's free'r then that GLP. |
18:19:00 | jimlay | ok. I thought for a moment the server was actualy catching the segv. |
18:20:21 | jimlay | EXTREAM COOLNES! windows still work when there process is SIGSTOP |
18:20:27 | scan[ibook] | :) |
18:20:35 | jimlay | only the part that they actually have to draw doesn't update. |
18:20:44 | scan[ibook] | which app? |
18:20:59 | jimlay | it's.. . how do you say beautiful just doesn't cut it. |
18:21:11 | jimlay | the block out. I just mean the window all still works. |
18:21:21 | scan[ibook] | hehe.. if you think that's cool, try running something newer like connect four |
18:21:38 | scan[ibook] | blackout is kinda old, haven't updated it for some of the new canvas widget features |
18:21:49 | scan[ibook] | in connect four, the game board can resize even when the app is stopped |
18:21:57 | scan[ibook] | most apps work like that |
18:22:00 | jimlay | lol.. that was a big typeo. bock out = blackout. |
18:22:10 | scan[ibook] | figured that was what you meant :) |
18:23:35 | jimlay | that's really cool. connectfour has everything handle in the server so only the input fails, the window handles complete redrawing. |
18:24:18 | scan[ibook] | It just maintains a list of primitives in the canvas widget, and redefines the canvas' coordinate system to have pgserver do scaling for it |
18:24:46 | scan[ibook] | has the nice side effect of keeping the network bandwidth low too |
18:25:16 | jimlay | yeah it's simple enough, but X11 just is designed backwords from that perspective. |
18:26:49 | jimlay | what skins have color targets? |
18:26:56 | jimlay | just the default? |
18:27:11 | scan[ibook] | color targets? |
18:27:25 | jimlay | like a color fb. |
18:27:55 | scan[ibook] | oh.. the zaurus skins |
18:27:57 | jimlay | the zauros does. |
18:28:07 | jimlay | but there isn't a good desktop skin is there. |
18:28:19 | scan[ibook] | you don't have to use those skins... |
18:28:22 | jimlay | the default skin (ie no skin) is kind of featureless. |
18:28:24 | scan[ibook] | they're only really for emulating handhelds |
18:28:36 | scan[ibook] | methinks you need to know the difference between skins and themes :) |
18:28:39 | scan[ibook] | I should put this in the FAQ... |
18:28:55 | jimlay | methinks I just figured that out from "you don't have to use those skins..." |
18:29:00 | scan[ibook] | the 'skins' are just configuration files that make pgserver emulate various devices |
18:29:05 | jimlay | and ls them (tab) (tab) |
18:29:21 | scan[ibook] | themes are the actual databases that define the look and feel |
18:29:58 | file | I got my ADSL router |
18:30:02 | scan[ibook] | themes are pretty flexible.. you can even load multiple themes at once |
18:30:08 | scan[ibook] | file: does it work yet? |
18:30:25 | file | scan[ibook]: some jackass didn't set the jumpers! |
18:30:36 | file | on our line, for ADSL |
18:30:54 | scan[ibook] | heh |
18:31:01 | jimlay | pgserver -nc beveled.ths ; doesn't work. |
18:31:07 | jimlay | what's the correct extension. |
18:31:13 | file | so the guy said maybe by 9pm |
18:31:21 | jimlay | it says to ask you if I have anyqyestions. I read the README. :) |
18:31:37 | scan[ibook] | jimlay: you have to use themec to compile .ths themes into .th themes, then load them with -t or the config file |
18:31:49 | jimlay | danke |
18:31:53 | scan[ibook] | jimlay: I should put a note in the README to read README.configfile and the output of "pgserver -h" :) |
18:32:39 | scan[ibook] | you can load as many themes as you want, for example "pgserver -t aqua.th -t mod_dropshadow.th -t my_background.jpg.th" |
18:32:43 | jimlay | thinks picogui kicks ass. |
18:32:47 | scan[ibook] | there's a pgbg tool to package background images into themes |
18:32:49 | scan[ibook] | :D |
18:32:55 | scan[ibook] | glad you like it |
18:33:08 | jimlay | it's not my fault that pgserver --help does't produce anything. |
18:33:27 | scan[ibook] | oh yeah.. been meaning to fix that bug |
18:34:39 | jimlay | you know what else kicks ass. Berlin takes 20 minutes to build (server only) and has way less features. |
18:34:50 | jimlay | This takes <1 min to build and has way more faetures. |
18:35:03 | scan[ibook] | berlin's design is a lot more general than picogui's.. just doesn't have nearly as many features done now |
18:35:33 | jimlay | ok well I'm goin g to go eat and then business meating. So talk to you later. |
18:35:36 | scan[ibook] | should get some lunch |
18:35:44 | scan[ibook] | ok, ttyl |
19:16:10 | scanline | is back (gone 10:18:54) |
19:36:51 | gonkulator{work} | is back too |
19:37:44 | gonkulator{work} | poketh scanline |
19:38:15 | scanline | is pokethed |
19:38:20 | gonkulator{work} | whats up? |
19:38:45 | scanline | PicoBot: what's up? |
19:38:45 | PicoBot | Replacement program for uptime.. URL: http://www.burdell.org/up.php3 |
19:38:49 | scanline | ha |
19:38:52 | scanline | PicoBot: forget up |
19:38:52 | gonkulator{work} | ok |
19:38:52 | PicoBot | scanline: I forgot up |
19:39:05 | scanline | PicoBot: up is the direction opposite to the force of gravity |
19:39:06 | PicoBot | OK, scanline. |
19:39:09 | scanline | PicoBot: what's up? |
19:39:10 | PicoBot | i heard up was the direction opposite to the force of gravity |
19:39:37 | gonkulator{work} | gotcha |
19:41:36 | gonkulator{work} | is at work... |
19:41:41 | gonkulator{work} | waiting for files to download |
19:42:17 | gonkulator{work} | an hour left... |
19:42:24 | gonkulator{work} | decides to go get a drink |
19:43:22 | scanline | hands gonkey a designated driver |
19:47:26 | file | oh crap - not good |
19:47:28 | file | I'm on 5 hrs/mth for dialup!!!!! |
19:47:33 | file | and my DSL isn't even setup yet |
19:47:48 | gonkulator{work} | file: uh oh |
19:47:55 | gonkulator{work} | file: quick DIAL and UN-DIAL |
19:48:10 | file | lol |
19:49:35 | gonkulator{work} | oh well |
19:49:38 | file | gonkulator{work} : they haven't setup DSL on my line yet! |
19:49:45 | gonkulator{work} | that sucks |
19:49:52 | file | yes |
19:49:57 | gonkulator{work} | they didn't do that for me until I had yelled at them for like 3 months |
19:50:06 | gonkulator{work} | had to wait almost forever for the DSL to come on |
19:50:11 | file | the guy told me the jupmer switch is marked as "Outstanding" |
19:50:15 | file | er jumper |
19:50:32 | gonkulator{work} | loads up windows XP |
19:55:20 | file | bbl |
19:56:27 | gonkulator{work} | later file |
20:34:58 | carpman{work} | *yawn* |
20:35:01 | carpman{work} | PicoBot: seen scanline |
20:35:02 | PicoBot | scanline was last seen on #openzaurus 1 minutes and 18 seconds ago, saying: kergoth: hey... I didn't notice before that snes9x isn't GPL, it's "Free for noncommercial use", so I don't think I could keep a picogui/arm port of it in sourceforge CVS :( [Thu Sep 12 15:39:57 2002] |
20:35:18 | carpman{work} | PicoBot: seen gonkulator{work} |
20:35:19 | PicoBot | gonkulator{work} was last seen on #picogui 38 minutes and 51 seconds ago, saying: later file [Thu Sep 12 15:02:41 2002] |
20:35:27 | scanline | hi carpman{work} |
20:35:30 | carpman{work} | hi scanline |
20:52:39 | gonkulator{work} | is here |
20:52:54 | carpman{work} | hi gonkulator{work} |
20:52:59 | gonkulator{work} | hey carpman{work} |
20:59:33 | gonkulator{work} | time to work |
20:59:36 | gonkulator{work} | is away: working |
21:37:35 | gonkulator{work} | hmm... |
21:37:37 | gonkulator{work} | time to go home |
21:37:41 | gonkulator{work} | almost |
21:49:08 | carpman | is back |
21:49:12 | scanline | hi carpman |
21:49:14 | carpman | PicoBot: seen gonkulator{work} |
21:49:14 | PicoBot | gonkulator{work} was last seen on #picogui 11 minutes and 32 seconds ago, saying: almost [Thu Sep 12 16:43:56 2002] |
21:49:16 | carpman | hi scanline |
22:02:34 | laodamas | poke -v scanline |
22:03:27 | scanline | GNU poke version 0.0025636: Unknown command 'scanline |
22:10:04 | carpman | poke -f file |
22:10:31 | carpman | YAY! I found it!!! |
22:10:44 | carpman | http://www.penny-arcade.com/dl/pa-a2k.mp3 |
22:10:54 | carpman | The Answertron skit!!! |
22:12:26 | scanline | laodamas: my phone is being evil... IRC is more reliable |
23:00:14 | gonkulator | hey laodamas |
23:00:16 | gonkulator | is back (gone 42:37:50) |
23:00:23 | scanline | He's gone to dinner |
23:00:26 | gonkulator | oh |
23:00:31 | gonkulator | well... that sucks |
23:01:50 | gonkulator | scanline: when are you going to dinner? |
23:01:53 | gonkulator | PicoBot: seen carpman |
23:01:53 | PicoBot | carpman was last seen on #picogui 50 minutes and 59 seconds ago, saying: The Answertron skit!!! [Thu Sep 12 17:17:08 2002] |
23:06:40 | gonkulator | hmm... |
23:06:44 | gonkulator | pokes carpman |
23:06:52 | gonkulator | PicoBot: seen carpman |
23:06:53 | PicoBot | carpman was last seen on #picogui 55 minutes and 58 seconds ago, saying: The Answertron skit!!! [Thu Sep 12 17:17:08 2002] |
23:06:57 | gonkulator | oh yeah |
23:07:35 | gonkulator | PicoBot: seen file |
23:07:35 | PicoBot | file was last seen on #picogui 3 hours, 12 minutes and 15 seconds ago, saying: bbl [Thu Sep 12 15:01:34 2002] |
23:07:42 | gonkulator | hmm... |
23:08:36 | scanline | gonkulator: I dunno.. not really hungry yet |
23:08:42 | gonkulator | ok |
23:08:48 | scanline | plays Zelda DX in picogui with OpenGL acceleration |
23:09:10 | gonkulator | got caught in the rain at school |
23:09:17 | scanline | ah |
23:09:46 | gonkulator | doesn't like it when hail hits him in the nipple |
23:09:51 | scanline | hehe |
23:09:58 | gonkulator | ESPECIALLY when I am not expecting it |
23:12:10 | gonkulator | pokes file |
23:12:25 | scanline | whoa... I forgot that those octopus things kill you if you try to hurt them |
23:15:01 | carpman | hi gonkulator |
23:15:41 | gonkulator | hey carpman |
23:17:50 | gonkulator | pokes carpman |
23:18:17 | carpman | gonkulator: I'm here, talking to glitch on the phone |
23:18:22 | gonkulator | oh, ok |
23:18:26 | gonkulator | tell him I say "hi" |
23:18:32 | gonkulator | twice ;-) |
23:21:38 | gonkulator | oh, and if possible, use the word "cheese" |
23:22:21 | carpman | I did |
23:22:29 | gonkulator | yay |
23:22:40 | gonkulator | a |
23:22:49 | gonkulator | er, oops |
23:23:18 | scanline | waves to glitch |
23:23:49 | carpman | he waves back |
23:26:58 | file | damn NBTel! |
23:27:13 | file | it's 9:33pm and they never setup my line at the wirign center |
23:27:15 | file | er wiring |
23:27:39 | file | a work order for today, usually means to have it done today! |
23:38:47 | gonkulator | is away: stuff |
23:49:11 | Xentac | is back (gone 09:50:48) |
23:51:18 | file | Xentac: BAH BAH BAH |
23:51:39 | Xentac | what's wrong? |
23:53:14 | file | my ADSL |
23:53:17 | Xentac | why? |
23:53:18 | file | the bastards never did the stuff at their end |
23:53:30 | Xentac | what does that mean? |
23:53:32 | file | has the router |
23:53:40 | file | it means there's no ADSL on my line yet |
23:53:49 | Xentac | oh, I see |
23:54:18 | file | it sucks |
23:54:21 | file | so close, and yet so far |
23:55:27 | Xentac | hehehe |
23:55:28 | Xentac | that's ok |
23:55:41 | Xentac | I ordered my ADSL in like September of 98... and didn't get it till like April of 99 |
23:56:05 | file | Xentac: the tech guys are just too lazy to change some jumpers on my line! |
23:56:20 | Xentac | nods. |
23:56:56 | file | I'm gonna keep calling :) |
23:57:08 | Xentac | ok... |
23:58:20 | carpman | "Please state the nature of your DSL emergency" |
23:58:42 | file | carpman: A lazy technician is too lazy to change the jumpers on my line to provide DSL service. Please assist. |