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

Re: Bug reports of DFSG violations are tagged ‘lenny-ignore’?

On Tue, 2008-10-21 at 15:55 -0200, Alexandre Oliva wrote:
> Hi,
> Sorry if this breaks threading, I'm not on this list, I was merely
> referred to the discussion through the archives.  If you respond to
> this e-mail, please address your replies directly to myself as well,
> so that I can respond without further breaking threading.
> I applaud Debian's willingness, even if hesitant and heated, to tackle
> this issue.
> I'm here to point out that there's some exaggeration and false
> dilemmas being presented as objections to the liberation of the kernel
> in Debian main.
> I saw references to e100 and ATI video cards.  Although it is true
> that linux-libre and other proposed means to remove their blobs might
> cause some hardware to malfunction or fail, I haven't come across any
> such hardware myself.  I do have computers with e100 network cards and
> ATI video cards, and they keep on working just fine with linux-libre.
> Now, I don't know whether the absence of the firmware causes degraded
> functionality, or whether it is just not necessary for the specific
> hardware I own.

Correct.  The e100 driver works with a large family of controllers, some
of which need the driver to supply firmware and some which don't
(presumably it's built in).  The r128 and radeon kernel drivers only
deal with 3D functionality and the X server can drive the 2D part
without them.

> But assuming that the removal of these pieces of
> firmware would break all instances of said hardware is a mistake.

Unfortunately that's something we don't really know, and it may take
some time to discover just what the impact is.  So while I'm continuing
to work on separation of firmware, I leave it to the kernel and release
teams to judge whether these changes are worth making before lenny.

> Anyhow, if you do decide to ship a Free kernel in Debian GNU/Linux
> main, please remember that cleaning up the kernel binaries is only
> part of the work; if the corresponding sources still contain the
> non-Free blobs,

They don't - there is a script to strip them from the upstream sources.


Attachment: signature.asc
Description: This is a digitally signed message part

Reply to: