porting from SGIs to Linux?

[Roland Mainz]

Keith Packard wrote:
> > > So my second question is, I've heard that the new X.org release offers
> > > software support for pseudo-colors?
> >
> > Not that I've heard of. Some of the recent changes would make it
> > easier to implement, if someone was interested in doing so, but
> > there's not a whole lot of interest in pseudo-color in the X development
> > community these days.
> 
> Egbert Eich has indeed put together some software simulated pseudo color
> support.  Note that this won't be any faster than doing it yourself in the
> application though, and may in fact be significantly slower except in the
> remote case as it won't be able to perform any application-specific
> optimizations of the recoloring process.

Erm, the emulation may not be slower, it depends on the application
usage and whether the Xserver can have it's own CPU. PseudoColor usually
means the application has to push less data to the Xserver and then it's
up to the Xserver whether it can process this faster than an
application-side color-index-to-TrueColor conversion.
For example applications like a flight control system (which does
screen-wide recoloring) is much faster in PseudoColor
(2560x2000x24bitimage blit vs. one change in the colormap - guess what's
faster :)) as it doesn't have to update the whole screen each time an
item gets recolored (for example recoloring a path to red/orange in
emergencies etc.).

> I'm not sure which drivers support this new software pseudo color option.

In the Xorg server it's currently only the "neomagic" driver, for
Solaris/Xsun almost every driver either has accerlated PseudoColor
support or PseudoColor emulation (even the SunRays) ...

----

Bye,
Roland

-- 
  __ .  . __
 (o.\ \/ /.o) roland.mainz at nrubsig.org
  \__\/\/__/  MPEG specialist, C&&JAVA&&Sun&&Unix programmer
  /O /==\ O\  TEL +49 641 7950090
 (;O/ \/ \O;)