Text 13695, 97 rader
Skriven 2009-12-20 07:42:00 av WAYNE CHIRNSIDE (1:123/140)
     Kommentar till en text av JAMES BRADLEY
Ärende: Re: All better now
==========================
-=> JAMES BRADLEY wrote to WAYNE CHIRNSIDE <=-

 -=> ROSS CASSELL wrote to JAMES BRADLEY <=-

 RC> 12 Dec 09 15:24, you wrote to Wayne Chirnside:

 JB> Long long ago, in a galaxy... I digress. I recall reading the order of
 JB> control that occurs after a boot, and took from those words that the
 JB> firmware on the vid (I *think* it's a BIOS routine.) runs through its
 JB> business before POST on the main board, and then any other controllers
 JB> (IE: NIC SCSI/RAID...) before the kernel is called.

 RC> A. Layer One = BIOS
 RC> B. Layer Two = Operating System
 RC> C. Layer Three = Programs

 RC> The individual I was conversing with, was insisting that his bootloader
 RC> was loading a video driver, before his BIOS even posted and executed
 RC> the bootloader.

 WC> The firmware in BIOS'ed video card *was* in fact linked to
 WC> lilo.conf in the MBR *before* the motherboard BIOS
 WC> attempted to take over, and failed dramatically,
 WC> to reestablish itself.

 JB> Your machine locks with the NVidia card, and is fine without? This
 JB> leads back to an address conflict, no? BIOS isn't being told to drop
 JB> its onboard vid, or is reserving something you need for the NVidia.

Nope, doesn't lock *unfortunately*, it corrupts
any partitions file systems attempting to be mounted beyond
repair.

This only because the Nvidia driver tainted the kernel and
modified lilo.conf.

 WC> Anyway, as I mentioned all before as well
 WC> simply google the following keywords:
 WC> Linux GeForce MX-400 experiences

 JB> This is why I suspect the canned NVidia driver. I know ATI has a
 JB> aftermarket/isolationism driver that is run in an executable file, and
 JB> I suspect NVidia is guilty of the same tomfoolery. 

Yup, log onto their web site, read the disclaimerand whap
down comes their binary file which both taints the Linux kernel
as well as altering lilo.conf to resolutions the Mach 64 and my other monitor
just can't handle.

JB> I'm not up on it,

Yet subbing binary for executable, same same, you nailed it!
Unlike some others here offering up advice.

 JB> and I can only write from what I know in the here and now. That's the
 JB> way it used to work, and what is up to date aint me. <G>

I aint so untouched by time either.

 WC> IOW's plenty of other people found this out before
 WC> me and but also unfortunately after I followed
 WC> some very unsafe and ill advised advice.

 JB> I'm all ears, if you'd care to further my understanding of "the latest
 JB> and 'greatest'".

We'll go gmail as soon as I send and get a reply from you.
I don't post either my cable nor gmail account anywhere

 WC> Right now that card and it's software driver sits in the bottom
 WC> of a drawer with my never intending to mess with such
 WC> a piece of poorly designed firmware and software *both*.

 JB> Might be a good candidate for a "friends" Windows machine. <G+D>

As Maurice researched they're not so swell there either.

 WC> As a *final*, I hope, comment on this issue I
 WC> may very well get an Nvidia equipped motherboard
 WC> with compatible BIOS only because quite soon
 WC> Nvidia's driver will be made available as part
 WC> of the new Linux kernel and thus not taint the kernel
 WC> rendering it worse than useless with another video card
 WC> as does Nvidia's proprietary software for the GeForce MX-400 .

 JB> It might be a keeper after all of that. Maybe the N-S bridge by the
 JB> same company can be copasetic with it. What's the buzz on their
 JB> chipsets? I tend to taint my opinions with the company's past
 JB> performance. This *does* sound like a doozie!

It may be a keeper indeed once I get through all the work arounds
I'd intended before a catfight ensued here held me up
a week and a half minus the 70 hours it took
me to get a stable Linux U.S. again after following some bad advice.
 
--- MultiMail/Linux v0.49
 * Origin: Doc's Place BBS Fido Since 1991 docsplace.tzo.com (1:123/140)