Re: [Nolug] Microsoft Story of Interest

From: Andrew S. Johnson <andy_at_asjohnson.com>
Date: Wed, 31 Dec 2003 11:24:40 -0600
Message-Id: <200312311124.41058.andy@asjohnson.com>

On Wednesday 31 December 2003 10:04 am, Chris Reames wrote:
> > Ron Johnson wrote:
> >
> > It might not be as bad as we fear. Intel sells a *lot* of chips
> > into linux servers, which are their high margin products...
> >
> > However, if I were a really malevolent virus writer, I'd be
> > salivating right now (unless the EFI "drivers" are embedded in
> > EPROMs).
>
>
> >On Behalf Of Andrew S. Johnson
> >1) Unless this new code comes from the US, it likely won't be a MS
> >derivative. Nobody in Asia wants to pay licensing fees when free
> >is an option, so it might even be Linux based.
> Maybe they will follow the Linux model and take submissions from multiple
> sources and integrate a working EFI 'package'.
> Then again, it may turn out to be FYOB (Flash Your Own BIOS).
>
> >2) If this new system makes it easier to support something like a small
> >Compact Flash card to hold all that is needed to boot, including an OS,
> >then I can see an upside. No more special hardware or programs to
> >flash an EEPROM. Any computer with a USB port could bail you out.
> Andrew, can you elaborate the point about the USB port?

I have a 7-in-1 USB flash reader that I could connect to just about any
computer built in the last few years and write to the CF card. Today,
you'd have to have an EEPROM programmer ($200+) and the hard-
to-find EEPROM that matches the one you have. Maybe not so hard
if you have a DIP, but the newer ones that are square and sit in a socket
I've haven't seen at Radio Shack.

Look at this board from VIA and you'll see what I'm talking about:

http://www.viavpsd.com/product/epia_MII_spec.jsp?motherboardId=202

>
> >4) It seems a prerequsite for this that the hardware be standardized
> >at the lowest levels, meaning that chipsets all act the same and use
> >the same instructions. If not, how would this work? If you have to have
> >some code beyond EFI, then aren't we back at the BIOS we have now?
>
> My thoughts are that there will be a central 'variable processor' that calls
> subprograms that are EFI updates.
> The hardware may flash EEPROM during some seeking process. This would make
> manufacturers independent with no global standard other than working with
> the EFI seek process.
> The upside is hardware features will not be limited to bottom of the barrel
> foot dragging standards of legacy compatibility.
> The upside/downside is either the hardware manufacturer or software
> programmers will have to establish compatibility between the hardware and
> software.
> The downside is that certain hardware may be totally incompatible with
> specific OSs. Microsoft dream world here. Intel makes chip X for Microsoft
> and chip Z for Linux servers. Everybody builds PC's with chip X. No more
> Linux distro's until chip X is cracked or patched. Same with other hardware
> (just my parinoia).

The big downside I see to a hardware standard at the chipset level is
that it will prevent the adoption of new features until they become standard.
Patents will have to be cross-licensed, blah blah blah. I don't see anybody
making chips for distros, but we already see different chips for different
CPU's. I'm not worried about MS here, it's Intel trying to squeeze out AMD
and make us pay $900 per CPU that scares me. If you think Intel is over-
priced and underpowered now, just wait until they've gotten rid of their
competition.

Andy Johnson

>
> Just my two cents...does anybody have change for a penny?
> Chris
>
> ___________________
> Nolug mailing list
> nolug@nolug.org
>

___________________
Nolug mailing list
nolug@nolug.org
Received on 12/31/03

This archive was generated by hypermail 2.2.0 : 12/19/08 EST