about this linux-libre 'bug' [Fwd: Re: question about a bug mention in an interview back in 2013]

[ng0 <ng0@pragmatique.xyz>]

Hi,

I think this could be useful for extending the documentation
by explaining the limitations of linux-libre in case we
don't already do it in an easy language.
Appended Message Exchange with Alexandre Oliva:


On May  3, 2017, ng0 wrote:

> do you have a bugtracker where progress on this could be visible,

'fraid not.

> Do I have your permission to forward this message in full to our
> developer mailinglist,

Sure

> or do you happen to have an explanation already
> somewhere more public?

I'd have pointed to Bruce Byfield's interview, but evidently that was
not clear enough ;-)

----- Forwarded message from Alexandre Oliva -----

> Date: Wed, 03 May 2017 00:20:56 -0300
> From: Alexandre Oliva
> To: ng0
> Subject: Re: question about a bug mention in an interview back in 2013
> 
> On May  1, 2017, ng0 wrote:
> 
> >> Indeed, I became aware that some users have got the idea that
> >> blocking the loading of blobs is a feature. It's not; it's just a
> >> bug that's quite difficult to fix. The decision on whether or not to
> >> use a piece of software, be it Free or not, should belong to the
> >> users, and it's not our intent to make that difficult.
> 
> > If for example linux-libre is installed on a device with an
> > intel wifi card, and your bug is solved, would this imply that
> > the intel card can be loaded (which currently can't be achieved)?
> 
> I suppose you mean 'the blob that controls the intel card can be
> loaded'.  If so, the answer is yes.  Loading it and running it is a
> user's decision, even when the software is non-Free.  We don't stop
> users from doing so, but unfortunately we make it very difficult
> (modifying the module sources and rebuilding the kernel, or at least the
> module, is currently required).  If the bug was fixed, it would likely
> work out of the box.
> 
> Unfortunately, it looks like fixing this bug is not possible.  Even with
> the internal firmware loader and if we were to enumerate available
> firmware files before ever asking for them, hashed or not, a userland or
> networked filesystem implementation could make a file listing available
> that amounted to *installable* firmware rather than to *installed*
> firmware, and then, once the kernel asks for a file, proceed to install
> it, or ask the user to agree to have it installed.  This sort of
> kernel-directed request and installation is precisely what we wish to
> avoid, and there doesn't seem to be any way to avoid it, and it's not
> like the development of such an automated firmware installer is
> far-fetched: AFAIK it has been done through hotplug scripts.
> 
> -- 
> Alexandre Oliva, freedom fighter    http://FSFLA.org/~lxoliva/
> You must be the change you wish to see in the world. -- Gandhi
> Be Free! -- http://FSFLA.org/   FSF Latin America board member
> Free Software Evangelist|Red Hat Brasil GNU Toolchain Engineer

----- End forwarded message -----

-- 
https://pragmatique.xyz
PGP: https://people.pragmatique.xyz/ng0/