From: Jon M. Taylor <taylorj@ecs.csus.edu>
  To  : ggi-develop@eskimo.com
  Date: Thu, 30 Jul 1998 22:36:29 -0700 (PDT)

Re: kgicon, etc

On Thu, 30 Jul 1998, Edward S. Marshall wrote:

> On Thu, 30 Jul 1998, Emmanuel Marty wrote:
> > 4) steffen's KGI spec is ready and solid, and I of course _still_ want a
> > native KGI. we will probably not get it into the kernel if we rewrite the
> > console bits from roof to top, be them superior or so, but I still want it.
> 
> The problem, of course, is that it can't be considered superior to the
> current console code by the mainstream kernel developers simply because it
> doesn't work on all current Linux ports. Or has some work been done lately
> on getting it running on the SPARC, Alpha, MIPS, PPC, etc. ports?

	See my previous post - the console layer should be
architecture-independent.  KGI probably should be also.  I wanted to
create a cgsix KGI driver sometime last year, so when we got a new system
drive for the Sparc10 my dad has sitting in his office here at home I got
him to agree to let me put Linux on the Sparc and do some development. 
Unfortunately, I just could not get f!@#ing RedHat 5.0 to install
correctly over the network, nor could I get SILO to work right which left
my dad unable to boot into Slowlaris to do his GIS work which he needed to
be able to do for his business.  

	I tried for like two weeks to make it work, but in the end I had
to admit defeat and kill the Linux partitions.  I ended up having to
reinstall Slowlaris to get SILO off the bootblock.  The damn machine has
been off for the last several months, too, but I have been focusing either
on school or kgicon and I will probably be moving out of my folks' place
real soon now that I have finished my last class and can get a real job. 
So, my Sparc days may be over soon.  But, if anyone else out there has a
Sparc/Linux box to play with, I'll bet that it would be very easy to port
the fbcon cgsix driver to kgicon.  Same is probably true for any other 
architecture's drivers.

> Not a complaint at all; just something to consider when thinking about
> strategies for getting the whole EvStack/GGI Console concept accepted.
> Once it's working on a handful of other platforms, and the concept is
> shown to be solid in a portable way, you'll likely be golden with the
> kernel folks.

	Well, nobody but us x86 folks (and PPC with Jason) have stepped
up to the plate yet....

Jon

---
'Cloning and the reprogramming of DNA is the first serious step in 
becoming one with God.'
	- Scientist G. Richard Seed