On Sunday 08 December 2002 12:38, James Sparenberg wrote:
> On Sun, 2002-12-08 at 01:37, Alex Chudnovsky wrote:
> > I agree completely. What lacks sorely in XFree86, IMHO, is the fallback
> > mechanism - if this driver doesn't work, use another driver, and if
> > nothing works, use some worst case default ( 640x480 generic VGA driver
> > or VESA FB driver for videocard, autoprobe for mice). As in (cough)
> > Windows. Anyway, it is pretty hard these days to find EGA card or (cough)
> > CGA card.
>
> This is also a position that could help when say replacing a blown video
> card.  Many is the time that once I got "something" working, and could
> then adjust from there, to get a nice product. For ex-windytes (windows
> users) the ability to limp into a visible mode and then properly adjust
> with MCC would be a really big help when changing hardware etc.

Yes, and not only that. My wet dream for successful Linux desktop is XFree86 
not needing configuration file at all - i.e it should use one if it finds it, 
but if it finds nothing, then it should be as follows :
- Video card is autoprobed, if the suitable driver found, it is used, or 
generic VESA, or generic VGA.
- Mice are autoprobed, all mice found are used.

And about mice : 
Why should one re-run XFdrake anew or manually change XF86Config-4 if all he 
does is just replace some PS/2 mouse with some new USB mouse? IMHO, mice 
should be autoprobed. OTOH, if XF86Config-4 contains mouse definition, it 
should be used, but if it is incorrect,  it should be considered omitted, 
warning should be issued, and mice should be autoprobed.
-- 

Regards,
Alex Chudnovsky
e-mail : [EMAIL PROTECTED]
ICQ : 35559910


Reply via email to