Lee Brown wrote:
> On Tue, 07 Nov 2000, Steffen Seeger wrote:
> >
> >portable graphics drivers,
>
> This is the issue to which I was oh so subtly referring. This works to the
> advantage of KGI and not Linux.
I cannot understand the point you are trying to make.
When KGI started, the illusion was that it would have a realistic chance
to get accepted one day, not because of political moves and 'being
friends with the right people', but because of the good and new
ideas in there that do help to solve quite some of the issues related
to graphics hardware support under Linux.
And they are in there, starting from modular driver designs up to a
handling of graphics hardware similar to the SGI RRM. According to James
even some of the kernel developers seem to have realized that some
ideas in there are quite useable. Most of them are also implemented,
in contrast to other approaches working to some extend;
it hasn't undergone reality testing (with a real-world X server using it).
One idea was to attract graphics hardware _vendors_ to write drivers
for KGI, by the 'arrangement' that other platforms could be supported
from the same source base by nothing more but a recompile. From my point
of view, this is the __ONLY__ chance you have to get a graphics hardware
vendor to write a driver for any 'minority' platform. The main reason
for this is that in order to produce an optimized version of this driver,
you will need to spend development time in the order of the product
lifetime. So you will be quite picky about the systems you support.
And this brings me to the main point why I will _never_ give up: portability.
Portability is the main clue behind the whole Open Source movement: If
you write an Open Source program, you want to keep the investment you have
made into it. All Open Source licenses have one point in common: They take
one essential right from the developer, the right to decide what the
future of his work should be. Applied to device drivers,
they give you, the user, the right to re-use this driver in your own
environment, which basically is the act of porting it.
So, from my -- admittedly very idealistic -- point of view, if you don't have
an easy way to recycle the device drivers in a new environment, you can as well
dump the whole effort, it will only be temporarily. And this is the reason
why KGI from it's very beginning has been laid out to make it easy for other
groups, projects, whatever, to take the ideas and code in there and adopt them
to their needs. As this actually happened, KGI has succeeded -- in that way.
So one might argue that so far KGI seems to have worked to the advantage of
Linux, not to itself.
Friendly,
Steffen
_______________________________________________________________________________
Steffen Seeger mailto:[EMAIL PROTECTED]
