Back
[00:00:07] <jepler> jmkasunich: then emc would pop up the "do you want to restart emc" prompt -- not ideal, I admit
[00:01:01] <jmkasunich> ok, one step at a time
[00:01:34] <jmkasunich> do we agree that it would be usefull for the main emc script to pass along cmd line options from DISPLAY= foo option to the GUI?
[00:01:50] <jmkasunich> if so, I'll make that change to the script
[00:02:06] <alex_joni> * alex_joni wonders what exactly a 'Longword' is
[00:02:20] <jmkasunich> (passing a .ngc file from the emc command line to the GUI is another thing entirely)
[00:02:42] <cradek> it's fine with me if you want to make that change.
[00:02:43] <jepler> jmkasunich: Yes, I think it would be good to allow commandline arguments in DISPLAY=
[00:03:07] <alex_joni> anyone wants to find out what a Longword is?
[00:03:22] <jmkasunich> what is the context?
[00:03:23] <jepler> alex_joni: 32-bit integer?
[00:03:33] <alex_joni> jmkasunich: hal driver ;)
[00:03:42] <alex_joni> jepler: not really :))
[00:03:43] <jmkasunich> back in the day, byte was 8, word was 16, longword was 32
[00:03:50] <alex_joni> typedef union
[00:03:50] <alex_joni> {
[00:03:50] <alex_joni> signed long intLong;
[00:03:50] <alex_joni> signed short intWord[2];
[00:03:50] <alex_joni> } LongWord;
[00:04:02] <alex_joni> imagine that..
[00:04:08] <cradek> uh
[00:04:38] <jmkasunich> time honored method for accessing sub-sets of a larger type
[00:04:38] <jepler> "signed long Long"?
[00:04:56] <jepler> er, "signed long int Long"?
[00:04:56] <alex_joni> * alex_joni is really laughing hard
[00:05:10] <jepler> how about a little #include <inttypes.h>
[00:05:16] <alex_joni> signed short int Word ;)
[00:05:24] <alex_joni> oh Word[2] :))
[00:05:26] <jepler> or does this code have K&R style function definitionsm, too?
[00:05:34] <cradek> in hungarian notation it'd be slLlLong?
[00:05:49] <jmkasunich> its a typdef for pete sake
[00:05:57] <jmkasunich> LongWord foo
[00:06:09] <jepler> I'd join the typedef union, too, if I thought I'd get better wages
[00:06:13] <jmkasunich> foo.Word[0] = <16bits>
[00:06:27] <alex_joni> I think sz is more appropiate for the hungarian language (better then 's')
[00:06:30] <jmkasunich> <32bits> = foo.Long
[00:07:25] <jmkasunich> if you know that nobody is going to change endian-ness on you, what is wrong with it?
[00:08:25] <cradek> who says shorts and longs are different sizes?
[00:08:28] <jepler> but why not take it to the logical conclusion .. signed short int bits:1[32]; (unfortunately, that syntax isn't supported by C .. sigh)
[00:09:18] <jmkasunich> jepler: because the underlying machine doesn't support bit access
[00:09:27] <jmkasunich> but probably does support byte/word access
[00:09:44] <jepler> cradek: or that longs are twice as wide as shorts
[00:09:47] <jmkasunich> I bet the guy who wrote that has an embedded background
[00:09:50] <alex_joni> ok, didn't want to stop what you where doing
[00:09:58] <jepler> $ grep printf sz.c; gcc sz.c && ./a.out
[00:09:58] <jepler> printf("%d %d %d\n", sizeof(short), sizeof(int), sizeof(long));
[00:09:58] <jepler> 2 4 8
[00:10:23] <jmkasunich> what box is that?
[00:10:28] <jmkasunich> hal is gonna have problems
[00:10:38] <jepler> jmkasunich: nothing too outlandish
[00:10:39] <jepler> Linux bald.unpythonic.net 2.6.14-1.1656_FC4 #1 Thu Jan 5 22:13:55 EST 2006 x86_64 x86_64 x86_64 GNU/Linux
[00:10:48] <alex_joni> 64bit
[00:10:54] <jmkasunich> gawd damned folks keep changing the definitions
[00:11:04] <jepler> definitions? What do you mean, definitions?
[00:11:28] <jepler> short has never been defined as (exactly) 16 bits, nor has long ever been defined as (exactly) 32 bits.
[00:11:29] <jmkasunich> byte/word/long have been 8/16/32 since the days of the 8085
[00:11:54] <jmkasunich> the lack of solidly defined FIXED LENGTH types is a real problem
[00:12:05] <jepler> ummmm, hello? <inttypes.h>
[00:12:17] <jmkasunich> when was that added?
[00:12:23] <jmkasunich> C99 maybe?
[00:12:31] <jepler> well, it's in The Single UNIX � Specification, Version 2
[00:12:32] <jepler> Copyright � 1997 The Open Group
[00:12:36] <jepler> http://www.opengroup.org/onlinepubs/007908799/xsh/inttypes.h.html
[00:12:47] <jmkasunich> what about non-unix?
[00:13:20] <jmkasunich> all the C out there that is running in microwaves, motor drives, etc, etc
[00:13:46] <alex_joni> cradek: you there?
[00:13:48] <jepler> they should, uh, take a lesson from unix circa the mid 90s?
[00:13:55] <alex_joni> or maybe jeff, I need some help
[00:13:57] <cradek> no
[00:14:00] <cradek> I'm hiding
[00:14:08] <alex_joni> In file included from /home/juve/emc2/src/hal/drivers/hal_vti.c:165:
[00:14:09] <alex_joni> /usr/include/pci/pci.h:25: error: redefinition of typedef 'u8'
[00:14:09] <alex_joni> include/asm/types.h:41: error: previous declaration of 'u8' was here
[00:14:18] <jmkasunich> lol
[00:14:19] <cradek> hahaha
[00:14:23] <jmkasunich> there you go
[00:14:25] <alex_joni> not funny ;)
[00:14:31] <cradek> sure it is
[00:14:43] <alex_joni> but include/asm/types.h doesn't sound like something that should get included for modules
[00:14:54] <jmkasunich> btw, I was doing typedefs for things like u8 back in the early 90s
[00:14:56] <jepler> /usr/include/pci/pci.h doesn't either
[00:15:19] <jepler> (sound like something that should be included by a module)
[00:15:22] <alex_joni> jepler: there should be a different pci.h ?
[00:15:27] <jepler> alex_joni: btfoom
[00:15:33] <alex_joni> I think it was always this one..
[00:15:41] <alex_joni> btfoom?
[00:16:09] <jepler> $ find /usr/src/linux-headers-2.6.12-magma/include/{linux,asm} -name pci.h
[00:16:10] <jepler> /usr/src/linux-headers-2.6.12-magma/include/linux/pci.h
[00:16:24] <jepler> "beats the **** out of me", means "I have no idea"
[00:16:46] <jmkasunich> haha, I hadn't heard that one before
[00:16:46] <alex_joni> heh.. yeah
[00:17:07] <alex_joni> * alex_joni looks at that one
[00:17:17] <cradek> fwiw, it's pronounced "B T foom"
[00:17:23] <jepler> cradek: it is?
[00:17:28] <cradek> I'm pretty sure
[00:17:40] <jmkasunich> so jepler, is u8 (as defined in asm/types.h) or uint8_t (as defined in inttypes.h) preferred?
[00:17:49] <cradek> huungry
[00:17:57] <jmkasunich> the fact that both exist simply confirms that things are fscked up
[00:19:04] <jepler> bbl
[00:19:20] <jepler> jmkasunich: I'd say uint8_t since it's a standard
[00:19:56] <jmkasunich> who
[00:20:01] <jmkasunich> who's standard?
[00:20:16] <jmkasunich> seems like asm/types is the linux standard
[00:20:30] <jmkasunich> whose I mean
[00:21:15] <jmkasunich> moot point I guess (just the embedded guy in me grumbling - the standard, if any, is a very recent developlent)
[00:21:17] <alex_joni> whose's ?
[00:21:43] <jmkasunich> so the guy who was assuming word = 16 and long = 32 isn't stupid, just old
[00:24:05] <jmkasunich> anyway, the hal types should probalby be changed to use inttypes.h or asm/types.h
[00:24:14] <alex_joni> yup
[00:25:48] <alex_joni> night all
[00:25:52] <alex_joni> * alex_joni heads to bed
[00:27:56] <SWPadnos> hmm - are there some relatively mindless but tedious things that I can do on the plane tomorrow (like changing the HAL types)?
[00:28:14] <jmkasunich> sure
[00:28:26] <SWPadnos> ok - lay 'em on me ;)
[00:28:34] <jmkasunich> I think the haltypes isn't that tedious, they are typedefs in hal.h (change in only one place, done)
[00:28:38] <jmkasunich> (I think)
[00:28:42] <SWPadnos> yep
[00:29:30] <SWPadnos> I'll be reading more in the shell scripting book, plus I'm bringing my tcl/tk book
[00:32:28] <SWPadnos> maybe I should add an "include" command to halcmd
[00:32:39] <SWPadnos> (not mindless by any means)
[00:33:11] <jmkasunich> nested hal files?
[00:33:17] <SWPadnos> kinda
[00:33:51] <SWPadnos> include may not make much sense to a non C programmer
[00:34:12] <jmkasunich> non-programmer anyway
[00:34:21] <SWPadnos> yeah
[00:34:21] <jmkasunich> even bash lets you include things
[00:34:31] <SWPadnos> . or source
[00:34:45] <jmkasunich> right, the syntax varies, but the concept is the same
[00:34:52] <SWPadnos> exec or load (or run) would make sense
[00:34:56] <SWPadnos> sort of
[00:35:42] <SWPadnos> I guess the other halcmd thing is to add readline support
[00:35:49] <SWPadnos> are you working on that?
[00:35:53] <jmkasunich> not today
[00:36:02] <jmkasunich> I was thinking of it as a 2.2 thing
[00:36:14] <jmkasunich> is it readline or getline?
[00:36:48] <SWPadnos> don't recall
[00:36:57] <SWPadnos> ok, I can wait
[00:37:10] <jmkasunich> looks like readline
[00:37:11] <SWPadnos> I'll be on vacation this week, and I should have some time to tinker
[00:37:21] <jmkasunich> I made a mistake in the TODO, I wrote getline
[00:37:59] <jmkasunich> the goal is to have history and tab completion
[00:38:04] <SWPadnos> yep
[00:38:32] <SWPadnos> if that's a 2.2 thing, it's OK with me. There's not much left for release though
[00:38:35] <SWPadnos> that's a good thing
[00:39:08] <jmkasunich> zoiks... just looked at the readline man page... that's definitely a 2.2 thing - there is a lot to it
[00:39:16] <jmkasunich> (at least it seems that way)
[00:40:21] <SWPadnos> yeah - it definitely looed a bit extensive when I glanced at it before
[00:40:35] <SWPadnos> and there was no obvious way to set the list of things that can be tab-completed
[00:40:44] <jmkasunich> yeah, I was just looking for that
[00:41:48] <SWPadnos> I don't remember if I ever found the method
[00:42:36] <SWPadnos> I know it's there, but doing things like limiting to types (for the show command) and then to hal ites of a particular type - that's not trivial (yet)
[00:43:13] <jmkasunich> "the actual completion performed is application specific"
[00:43:23] <jmkasunich> but no clue as to how the application controls it
[00:43:41] <SWPadnos> right
[00:43:43] <SWPadnos> me either
[00:43:50] <jmkasunich> google probably
[00:46:01] <SWPadnos> well - I'm about to see if cradek's script wotrks on kubuntu ...
[00:46:39] <SWPadnos> heh - nope. no gksudo command ;)
[00:46:59] <SWPadnos> good thing I'm stupid, and logged in as root
[00:50:02] <jmkasunich> ok, that's the first step in the taming of mini
[00:50:43] <SWPadnos> size args?
[00:50:58] <jmkasunich> no, passing args from the ini file to the GUI
[00:51:05] <jmkasunich> now we need to decide what the args should be
[00:51:17] <SWPadnos> that should be automatic, actually. wish should accept the --geometry XxY@ox;oy
[00:51:39] <jmkasunich> yeah, cradek/jepler convinced me that the convention should be respected
[00:51:41] <SWPadnos> but then mini will resize, so I guess it does need to know
[00:52:03] <jmkasunich> problem is Sherline wanted a custom size, and ray coded it do ignore the --geometry
[00:52:22] <SWPadnos> do you know why the script is started with /bin/sh, then restarts itself in wish?
[00:52:41] <SWPadnos> is it so the shell can find wish for it?
[00:52:48] <jmkasunich> tcl strangeness, vanishes in the mists of time
[00:52:50] <jmkasunich> I think so
[00:53:19] <jmkasunich> not rays idea or anything specific to emc tho
[00:53:32] <SWPadnos> no - I see it in every tcl program
[00:55:21] <jmkasunich> IMO, the right thing to do is to add a command line option to mini that triggers the sherline behavior, and let it do the normal behavior for everybody else
[00:55:56] <SWPadnos> that does make sense
[00:56:04] <jmkasunich> "--resize 80" to force resize to 80% of the screen perhaps?
[00:59:20] <SWPadnos> well, it's way too big on a 3840x1200 desktop
[00:59:37] <jmkasunich> its way too big on 1600x1200
[01:00:03] <jmkasunich> I see two lines that set the size to 90% of the screen
[01:00:10] <SWPadnos> ok. it's waaaaaaaay too big on 3840x1200 ;)
[01:00:21] <jmkasunich> but I'm not sure just disableing those lines will make it do the right thing
[01:00:39] <jmkasunich> and dinner time is coming, I don't want to dig into it right now
[01:00:52] <jmkasunich> * jmkasunich really doesn't want to bust mini, ray would be most displeased
[01:01:42] <SWPadnos> nope. I've got to get to bed as well. gota 6 AM plane to catch tomorrow
[01:01:59] <jmkasunich> have a safe flight
[01:02:15] <SWPadnos> thanks - we got bumped to first class, so it should be nice :)
[01:02:37] <jmkasunich> we? family/vacation trip as opposed to work?
[01:02:42] <SWPadnos> yep
[01:02:46] <jmkasunich> nice
[01:02:49] <jmkasunich> enjoy
[01:02:50] <SWPadnos> err -yes, vacation
[01:03:00] <SWPadnos> thanks. too bad it's almost as cold in Texas as it is here
[01:03:13] <jmkasunich> we're going to florida in a couple of weeks
[01:03:14] <SWPadnos> that's why we took the bump. we were supposed to leave today
[01:03:18] <SWPadnos> that should be nice
[01:03:27] <jmkasunich> gonna see the first baseball game of spring training ;-)
[01:03:32] <SWPadnos> unless the Texas weather moves in
[01:03:36] <SWPadnos> heh - cool
[01:04:06] <jmkasunich> my wife has a T-shirt that says "Life begins when the season starts"
[01:04:12] <jepler> SWPadnos: IMO that trick dates from a long time ago, before the #!-line was widely supported. It also makes the script work whether wish is in /usr/bin or /usr/local/bin.
[01:04:23] <SWPadnos> hmmm - I just installed firefox, tcl and tk docs and dev packages - anything else you think I should have?
[01:04:41] <SWPadnos> right - OK. I was thinking that it would take the path problems out of it
[01:04:41] <jmkasunich> build-deps emc2?
[01:04:48] <SWPadnos> ah, right
[01:04:59] <jmkasunich> wish I had remembered that one the other night
[01:05:24] <jmkasunich> when I was doing an install at 3am... make, error, apt-get, make, error, apt-get, etc, etc
[01:05:46] <SWPadnos> heh - that's a nasty annoying cycle
[01:05:50] <SWPadnos> much better with DSL though
[01:05:54] <jmkasunich> yeah
[01:06:26] <jmkasunich> well, time for dinner, then we have plans
[01:06:33] <jmkasunich> see you folks tomorrow
[01:06:38] <jmkasunich> jmkasunich is now known as jmk-away
[01:06:51] <SWPadnos> see you
[01:06:57] <SWPadnos> next week ;)
[18:22:28] <staggerlytom> bye
[19:28:59] <alex_joni> hello
[19:29:10] <cradek> hi alex
[19:32:50] <cradek> alex_joni: can you look at the logs on solaris and see if anyone is still using the apt repository there?
[19:36:19] <alex_joni> I can try that..
[19:39:56] <alex_joni> remember the alias?
[19:40:01] <alex_joni> was it /emc2/ ?
[19:40:04] <cradek> yes
[19:42:18] <alex_joni> does 64.105.106.249 ring a bell?
[19:43:15] <alex_joni> there are a few (about 4-5 accesses / day)
[19:43:15] <cradek> h-64-105-106-249.sfldmidn.dynamic.covad.net.
[19:43:20] <cradek> nope
[19:47:55] <alex_joni> does that answer your question?
[19:48:05] <cradek> yes, it means one person is using it still
[19:48:27] <cradek> there are no "covad" emails in my emc folder
[19:48:38] <cradek> hmm
[19:52:46] <alex_joni> it's not from the same person, but 3-4 accesses
[19:52:56] <cradek> ah, ok
[19:53:11] <alex_joni> so I think we can remove stuff, and put a webpage that redirects
[19:53:23] <cradek> that's a good idea, I didn't think of that
[19:55:28] <alex_joni> cradek: how about I suggest we add Eric as a developrer?
[19:55:33] <alex_joni> developer
[19:55:43] <cradek> who's eric?
[19:56:09] <alex_joni> Eric H. Johnson
[19:56:14] <alex_joni> the one with the VTI driver
[19:56:29] <cradek> ah
[19:56:33] <alex_joni> he already has sent code, and I put it in CVS
[19:56:35] <cradek> yes if he's submitted a driver
[19:56:42] <cradek> no question
[19:57:08] <alex_joni> he has some additions now, and I can repeat the procedure, but at one point it will be too much overhead
[19:57:32] <cradek> we should add him if he's doing active development.
[19:58:26] <alex_joni> not sure if he will ever work more than this driver, but it might have started his interest
[20:12:03] <cradek> I think that's fine
[20:12:17] <cradek> even if he wants to just maintain one thing, that's great
[20:13:10] <alex_joni> I sent an reply asking if he wants that..
[21:28:22] <jmk-away> jmk-away is now known as jmkasunich
[21:42:59] <cradek> chris@buster2:/usr/src/emc2$ bin/emc_module_helper insert /usr/src/emc2/rtlib/hal_vti.ko
[21:43:02] <cradek> insmod: error inserting '/usr/src/emc2/rtlib/hal_vti.ko': -1 Operation not permitted
[21:43:09] <alex_joni> that's expectable
[21:43:09] <cradek> chris@buster2:/usr/src/emc2$ dmesg|tail -2
[21:43:09] <cradek> [ 9928.188358] VTI: Exiting with auto detect failed
[21:43:09] <cradek> [ 9928.188364] VTI: ERROR: vti_init_card() failed
[21:43:13] <alex_joni> ;)
[21:43:19] <alex_joni> seems to be working as it should
[21:43:20] <cradek> can it return a better errno?
[21:43:29] <alex_joni> not sure..
[21:44:04] <cradek> ENXIO is No such device or address
[21:44:17] <cradek> ENODEV is No such device
[21:44:39] <alex_joni> that is easy I think..
[21:44:45] <alex_joni> just return ENODEV
[21:44:54] <cradek> are you sure?
[21:45:11] <alex_joni> * alex_joni has no idea ;)
[21:45:14] <cradek> haha
[21:45:19] <alex_joni> but I can try..
[21:45:23] <cradek> maybe -ENODEV
[21:45:35] <cradek> must be documentation somewhere...?
[21:46:22] <jepler> -ENODOC
[21:46:31] <alex_joni> jepler: lol
[21:47:25] <jepler> FATAL: Error inserting sunhme (/lib/modules/2.6.12-magma/kernel/drivers/net/sunhme.ko): No such device
[21:47:31] <jepler> looks like random devices in the kernel use ENODEV
[21:47:48] <cradek> I wonder how they return it
[21:48:14] <cradek> return -ENODEV;
[21:48:21] <cradek> I get the prize
[21:48:57] <jepler> static int __init happy_meal_probe(void)
[21:49:03] <jepler> if (!cards)
[21:49:03] <jepler> return -ENODEV;
[21:49:38] <alex_joni> yup
[21:49:48] <alex_joni> * alex_joni is already compiling ;)
[21:50:08] <cradek> jepler: your kernel is out of date again already...
[21:50:14] <jepler> cradek: oh foo
[21:50:26] <cradek> jepler: for some reason the modules are incompatible
[21:50:56] <jepler> cradek: oh foo
[21:51:12] <jepler> cradek: you change anything besides the CONFIG_4G?
[21:52:07] <cradek> jepler: no
[21:56:48] <jmkasunich> cradek: got this from the update manager wrt rtai-modules-2.6.12-magma 1.3.3-1+cjr1: Changes not found, the server may not be updated yet.
[21:57:06] <jmkasunich> is that because its not in the mainstream repositories, or ?
[21:57:15] <cradek> let me check
[21:57:43] <cradek> jmkasunich: do an apt-get update, that's not the new version
[21:58:32] <jmkasunich> does "reload" on the update manager to the apt-get update?
[21:58:37] <cradek> yes
[21:58:49] <jmkasunich> already did reload, no change
[21:59:15] <jmkasunich> lemme cancel the update manager and use apt
[21:59:26] <cradek> are you using the dsplabs repository?
[21:59:38] <cradek> deb
http://dsplabs.cs.upt.ro/emc2/ breezy emc2
[21:59:38] <cradek> deb-src
http://dsplabs.cs.upt.ro/emc2/ breezy emc2
[21:59:56] <jmkasunich> no
[22:00:00] <cradek> ahhh
[22:00:07] <jmkasunich> you are putting your debs in a different place now?
[22:00:08] <cradek> I didn't updated the other repository
[22:00:33] <cradek> we moved it off alex's home machine. He was going to set up a redirect, but I guess he hasn't yet.
[22:00:57] <jmkasunich> Hit
http://solaris.cs.utt.ro breezy Release
[22:01:06] <jmkasunich> thats alex's box?
[22:01:08] <cradek> yes
[22:01:22] <cradek> please switch to dsplabs
[22:01:27] <jmkasunich> so I should go into sources.list and change that
[22:01:27] <jmkasunich> ok
[22:01:31] <cradek> yes
[22:01:45] <jmkasunich> is it really upt.ro? or utt.ro?
[22:02:07] <jmkasunich> cradek deb
http://dsplabs.cs.upt.ro/emc2/ breezy emc2
[22:02:10] <alex_joni> better upt
[22:02:18] <alex_joni> both work though
[22:02:21] <jmkasunich> ok
[22:02:24] <alex_joni> but utt will eventually be forgotten
[22:02:50] <alex_joni> cradek: I don't think the redirection will work on all the hits
[22:02:55] <alex_joni> only on emc2/ hits
[22:03:16] <jmkasunich> hey, I read back and saw you guys trying to figure out if people were still using that box
[22:03:27] <jmkasunich> bet that was me
[22:03:31] <alex_joni> yes, bet so too
[22:06:47] <alex_joni> cradek: I've set up the redirect
[22:06:59] <cradek> oh, I was just trying to figure out how to rsync between them
[22:07:07] <alex_joni> don't
[22:07:07] <cradek> I guess I will not bother
[22:07:09] <cradek> ok
[22:07:11] <alex_joni> it's not worth it
[22:07:20] <alex_joni> that box might eventually die out ;)
[22:07:28] <cradek> it's too slow to send everything from here twice!
[22:07:34] <alex_joni> probably change it's name completely..
[22:07:48] <alex_joni> thought you said dsplabs works ok..
[22:07:54] <cradek> yes it's fine
[22:07:58] <cradek> it's my connection that's slow
[22:08:09] <cradek> my up speed is only 45kB
[22:09:02] <cradek> jmkasunich: does that give you the right packages?
[22:09:04] <alex_joni> ok
[22:09:13] <jmkasunich> yeah, cjr12, right?
[22:09:23] <cradek> alex_joni: maybe solaris/emc2/* should just give a 404
[22:09:30] <cradek> jmkasunich: yes I think so
[22:09:33] <alex_joni> I can give that easier
[22:09:39] <alex_joni> just remove the alias from apache
[22:09:40] <jmkasunich> and they have descriptions, noting the 1G issue
[22:09:45] <cradek> great
[22:09:52] <alex_joni> cradek: want me to do that?
[22:09:59] <cradek> alex_joni: yes
[22:10:09] <cradek> alex_joni: we'll hear from anyone who gets an error message then.
[22:10:45] <alex_joni> ok
[22:12:19] <alex_joni> done
[22:12:26] <cradek> thanks
[22:12:34] <cradek> I really appreciate your hosting
[22:13:00] <alex_joni> the least I can do for deb stuff
[22:43:42] <cradek> new emc2 & emc2-dev packages in the ubuntu repository
[22:45:32] <alex_joni> jepler: still there?
[22:51:51] <cradek> new emc2-axis package in the ubuntu repository
[22:56:12] <LawrenceG> cradek: Thanks.... installed and testing.... running 3dchips now
[22:56:34] <alex_joni> LawrenceG: what PC?
[22:56:53] <LawrenceG> a P200 with 192mb
[22:57:05] <alex_joni> oh, ok ;)
[22:57:26] <LawrenceG> using mini as I seem to have opengl problems with this box
[22:57:46] <alex_joni> oh, ok