Andreas Beck wrote:
>
> That is actually where the wrench should be applied. Using a HW-Mousecursor
> will have much less benefit than modifying X to ggiFlushRegion as
> appropriate. This will dramatically lower the required PCI Bus bandwidth.
>
> > So, I guess it would have a noticable effect, regarding the mouse.
>
> Well ... it might be noticeable, but if you want to work with the thing, the
> other method brings much more.
>
> Marcus: Did you join XFree yet ? AFAIK XF4 has Xinerama support, which is
> basically what you need.
+
+However it would be better to solve the primary problem, which is XGGI
+not supporting Xinerama setups.
+
>
> --
> = Andreas Beck | Email : <[EMAIL PROTECTED]> =
Hi guys,
My point is:
implementing a graphic mouse that would use interrupt programming would
be a good minimax (minimum programming, maximum result) option in an
apparence point of view (interrupt driven mouse drawing does not need
GGIFlush to happen)
And, if GGI_flush need to be rewritten, or Xinerama setup to be
supported I am ready to help, but give me the marbles to play.
Whatever the solution is: let's go.
However the opinion (I share with myself) is that I do not intend to
wait for Xinerama :
As a pragmatic guy I would say:
On the opposite of XGGI, XF 4.0 is STILL in core developpement, they are
claiming a lack of developpers, they are already late, the reactivity of
the group might therefore be lower than this community. It is a product
that will come clearly later than announced
As a user I would say:
The point is also that XGGI is more advanced and (for I am using it
every day in multi-display at the office). The "Beta tester" point of
view is: it currently work well (excepting a few details, like the
moouse). XGGI strategies for now are better because they are currently
working well.
As a coder I would say:
The spririt of XGGI as a modular, well layered architecture is also
better (on the coder point of view) than the all in one approach of XF
4.0. And, I could easily developp a FB device driver that glue perfectly
with XGGI, without knowing anything of XGGI. XGGI strategies for now
are better because they are currently working well.
Having grown up with PCs for 10 years I would say:
XF is reaching a critical mass, it is crumbling under it's own mass,
the same way microsoft windoz did:
- it is mixing GUI, and OS specific features (such as drivers ).
- it is quite a restricted community of programmers.
- it has imposed itself as the only alternative, so people are demanding
more features, they cannot provide at the rate of the demand. (And it
participate to the lack of programmer). XGGI can.
Conclusion:
1) XGGI has a future because it can attract developpers (it is a
readable well separated code): you need not to be Xfree specialists to
help.
2) XGGI has a future because it is more reactive to users, (see point
1).
3) Xinerama belongs to XF 4.0 => it will be late anyway. And perhaps, we
may forget Xinerama until it has proven to work ?
4) As a product that works, XGGI imposes de facto it's own point of
view.
Now, the point may be to have XGGI better known along the community, and
announce officialy it works? (I do all I can to have a photo this week)
Annexe:
Here are two extracts from the XF official home page:
<<
Q.A15- Can I use more than one video card in the same machine with
XFree86?
Although it is technically possible to use multiple PCI-based SVGA cards
in the same machine, none of the servers currently support this.
For XFree86-4.0 we are working on true multi head support.
>><<
1. What is XFree86 3.9.16?
XFree86 3.9.16 is a ``work in progress'' snapshot of the XFree86 4.0
development code. It is one of in a series of snapshot releases leading
up the the 4.0
release. The plan is to release a new snapshot every four to six weeks
...
>>
They also precise, the current release is not for users ... the first
release of XF 4.0 was expected in july the 20th ...
Julien Tayon
--
Easter-eggs Linux System Integrator
17 bis rue du Dr Zamenhof - 92500 Rueil Malmaison - France
Phone: +33 (0) 1 41 29 10 01 - Fax: +33 (0) 1 41 29 11 71
mailto:[EMAIL PROTECTED] - http://www.easter-eggs.com
