grub-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: About firmware facilities


From: Seth Goldberg
Subject: Re: About firmware facilities
Date: Sun, 13 Sep 2009 13:54:33 -0700 (PDT)
User-agent: Alpine 2.00 (GSO 1167 2008-08-23)



Quoting Robert Millan, who wrote the following on Sat, 12 Sep 2009:

On Fri, Sep 11, 2009 at 02:07:10PM -0700, Seth Goldberg wrote:

  I strongly disagree with you in this specific case.  Our experience in
Solaris has demonstrated that PXE firmware is surprisingly robust (when
the right combination of API calls (i.e. those tested by Windows) are
used).  We have been successfully using PXE-based firmware for netbooting
for many years now, and we would like to continue to do so.  Maintaining
a driver collection for NICs is futile, IMHO.  Using the firmware that's
there, and that's reliable should be the goal.  Not all firmware is our
enemy :).

Reliing on proprietary firmware is a compromise.  We don't install the blobs
ourselves, so we're not responsible for them, but it is still problematic
because user has less freedom (firmware bugs is just the most notable
consequence of this).

So our compromise is to use firmware when we have no other choice, or when
the alternative is not reasonable (e.g. not mature or complete enough).

My goal as maintainer is to encourage development of a usable and complete
driver framework.  I'm open to discussion about accepting code for using
hardware support from firmware, but keep in mind it's not our primary goal.

In the specific case of network hardware, I'm more reluctant because it's
a regression compared to what we had in GRUB Legacy.

I agree that choice is very important. In this case, our choice is to rely on PXE firmware, since we've had excellent experiences with it. We added an UNDI network driver to legacy GRUB, so from our perspective, not having PXE in GRUB2 is the regression :).

 Thanks,
 --S




reply via email to

[Prev in Thread] Current Thread [Next in Thread]