Re: Firmwares left
On 2007-09-25, Frederik Schueler <fs@debian.org> wrote:
> As for the rest: our priority is free software, who cares about users?
>
> tg3: Drop. The firmware should be pruned and the driver deactivated
> until someone shows up with a patch upstream accepts.
Upstream is still hostile to any decoupling of firmware from the tg3
driver. Any functional free tg3 driver will have to come from Debian.
(I just checked and the equivalent if_bge driver in FreeBSD is devoid of
firmware blobs.)
----- Forwarded message from David Miller <davem@davemloft.net> -----
From: David Miller <davem@davemloft.net>
Subject: Re: tg3 and request_firmware()
Date: Sat, 29 Sep 2007 00:00:18 -0700 (PDT)
From: Robert Edmonds <edmonds@debian.org>
Date: Sat, 29 Sep 2007 01:40:35 -0400
> Due to the "Free software" interpretation that has won out among voting
> Debian developers, the tg3 driver is to be removed[0] from Debian's
> 2.6.23 kernel. I do not agree with this interpretation and am very
> interested in seeing Debian not alienate users by breaking functioning
> drivers. Unfortunately the eventual removal of all "non-free" firmware
> from the Debian kernel has been in the plans for a long time and it
> seems far too late to reverse this decision.
Debian therefore will become even more irrelevant than it already is.
> * There is a firmware fix necessary for 5701a0 chipsets to operate, but
> this is a very old, uncommon chipset.
And what will you do when we come across a device which has a common
variant which requires a firmware fix to operate correctly?
I really don't agree with this approach to handling things, it's a
slippery slope down which we end up with a non-functional kernel by
default, sorry.
Theses firmware images are part of the driver and are absolutely
required for correct functionality in many cases. You can paper
around cases like tg3 but you will not be able to do this across the
board and it will in fact lead to hurting users. The time already
wasted discussing this issue already has hurt users, in fact.
I could be doing much more productive things than constantly saying no
to this request.
----- End forwarded message -----
--
Robert Edmonds
edmonds@debian.org
Reply to: