Alice99
Stephan Rudlof
sr at evolgo.de
Thu Feb 10 13:00:23 UTC 2000
I have
(8->1854 16->2064 32->1566 )
on a
Pentium mobile 366, Linux 2.2.10
and a
(--) Mach64: PCI: Mach64 Rage LT Pro rev 220, Aperture @ 0xfd000000,
Registers @
(--) Mach64: Memory type: SGRAM (5)
(--) Mach64: 1024x768 panel (ID 1) detected; clock 65.14 MHz
(--) Mach64: Clock type: Internal
(--) Mach64: Maximum allowed dot-clock: 230.000 MHz
(**) Mach64: Mode "1024x768": mode clock = 66.000
(--) Mach64: Virtual resolution: 1024x768
(**) Mach64: Video RAM: 8192k
(--) Mach64: Using hardware cursor
graphics with a LCD screen.
Isn't it a little bit worse?
Stephan
Bruce ONeel wrote:
>
> Hi,
> It's old and slow with a 16 bit i/o bus... PB540c with 100hmz 603e
>
> Full Screen:
> 640x480x8 (789 804 1570 956 1303 )
> 640x400x16 (1241 1241 1297 1248 1553 )
>
> In a window:
> 640x480x8 (1980 2009 2105 2425 2528 )
> 640x400x16 (2681 2693 2742 1152 2836 )
>
> Yes, the 16 bit color is 640x400 and I know we're not measuring the same
> thing since my display is so much smaller.
>
> cheers
>
> bruce
>
> "Raab, Andreas" <Andreas.Raab at disney.com> wrote:
> > Hey,
> >
> > Just FYI: Squeak relies *heavily* on a single function that needs to
> > implemented by the graphics driver. If your driver doesn't support this
> > function directly, some weirdo Windows simulation code will be run and that
> > makes things *really* slow. If you wish to find out if you've a problem here
> > you might want to run the following code:
> >
> > #(1 4 8 16 32) collect:[:depth|
> > Display newDepth: depth.
> > Smalltalk garbageCollect.
> > Time millisecondsToRun:[10 timesRepeat:[Display forceToScreen]]].
> >
> > Here are the results from my machine (PII 400Mhz, ATI Rage LT AGPx2),
> > running Squeak 2.7 in fullscreen mode on Win98:
> > 1024x768x8: ( 68 251 410 716 724 )
> > 1024x768x16: ( 65 601 684 458 424 )
> > 1024x768x24: ( 117 265 304 800 337 )
> > 1024x768x32: ( 114 1099 1192 898 239 )
> >
> > Note the funky differences! For optimal performance you should run Squeak in
> > the same display depth as your card is in but sometimes there are subtle
> > differences so it's best to actually try it out.
> >
> > [That reminds me: Somebody should write a benchmark tool that measures
> > typical operations (just like the above) so that one can get an idea of how
> > fast a squeaky system one is running on. Any volunteers?!]
> >
> > Andreas
> >
> > > -----Original Message-----
> > > From: Rik Fischer SmOOdy [mailto:riks at cs.pdx.edu]
> > > Sent: Wednesday, February 09, 2000 12:20 AM
> > > To: Milan Andric
> > > Cc: recipient list not shown
> > > Subject: Re: Alice99
> > >
> > >
> > > In message <389BA7DA.B7017887 at cs.pdx.edu>you wrote:
> > > > Rik,
> > > >
> > > > I have a dual PII w/Voodoo3 video card at home and I found Squeak to
> > > be
> > > > terribly slow (probably because of video subsystem). I still haven't
> > > > received your feedback on whether should we use Alice 99 or not.
> > > > Please reply asap, so I can adjust my work. In the meantime I'm
> > > working
> > > > in Squeak... (which sometimes really hurts...)
> > > I looked in http://www.egroups.com/list/squeak/
> > > and found nothing about voodoo.
> > >
> > > Perhaps someone on the squeak mailing list has seen
> > > some sort of impedance mismatch between Squeak and voodoo?
> > > If so, please advise.
--
Stephan Rudlof (sr at evolgo.de)
"Genius doesn't work on an assembly line basis.
You can't simply say, 'Today I will be brilliant.'"
-- Kirk, "The Ultimate Computer", stardate 4731.3
More information about the Squeak-dev
mailing list
|