Re: the laptop proposal -- ideas, thoughts, and criticisms
On 15-Sep-99 Werner Heuser wrote:
> Sean 'Shaleh' Perry schrieb:
>> Why framebuffer?
> - usually for every new graphic card there has to be a Linux
> X server developed. Framebuffer is kind of 'generic' graphic card
> support, at least if the card manufacturers follow
> the VESA VBE standard.
> - framebuffer devices could use the hardware better, they avoid the
> current video memory mapping 'concept' (I'm not a hardware wizard,
> please correct me, if I'm not clear here)
At the same time, I would like to keep the laptop kernel the same as Debian's.
Just with the needed pieces.
>> pentium ops are actually minimal compared to the -O2 output from egcc, plus
>> code is larger
> Do we have kind of a benchmark, let's say for X build with options.
X is not very cpu bound. There have been tests done, do not know where. The
gain is a few percent (I think it is less than or near 5).
>> the xserver in potato works with all supported Neomagic chips, a separate
>> neomagic server is no longer needed
> I checked http://www.debian.org/Packages/unstable/x11/ today.
> 'neomagic' isn't in the headlines. Where shall I look?
It is part of the xserver-svga -- no need for a separate server.
>> the hardware detection is much larger than laptops and is being looked into
> You wrote in another mail: " Several groups are working on a generic
> hardware detection interface -- once finished any linux can use
> these to detect hardware."
> I agree. We shouldn't make our own laptop hardware detection program.
> But the projects I checked, didn't support certain laptop hardware
> detection yet. For instance, I would like to have IrDA controller,
> ACPI and VESA/VBE correctly detected. Of coarse the actual list
> is longer. - Therefore my intension is, to have some laptop people
> working together with a good hardware detection program.
> Personally I prefer the Mandrake - Lothar - Project (but
> this has to be researched). And an according Debian package.