[GNU-linux-libre] Re: Freedom issues with non-free firmware in external files

[Yavor Doganov]

Diego Saravia wrote:
> > Could you please explain what this invitation entails?
> 
> you are telling to the world: we have free code to load firmware, we
> need  free firmware

No.  You are telling to the world: we have code to load your firmware,
so "be my guest".  Nobody would even figure out that this particular
part needs work, because... things will magically work and users won't
complain.  The right thing to do is what we've been saying in such
situations: "These things do not work with free software (yet)".

What's worse is that you're telling *your world*, that is, the users
of *your distro* who your promised to provide a 100% computing
environment, that you failed to keep that promise.  (This was the most
astonishing thing when I switched to gNewSense -- some devices that
I've always thought had free drivers turned out useless, and Debian's
promise to me as a user was really gibberish.)

> why develop gnash in that case? Its look as a contradiction

No contradiction.  Gnash is developed by the folks who need Flash
played with free software.  Apparently some users demand Flash
performance, so Gnash is useful for them.  There's nothing inherently
bad in Gnash, because it doesn't load any secret magic numbers from
anywhere to operate properly; it just tries to parse the data you (as
a user) pass.  It is a goal of the GNU project to provide a
replacement for just about any program and/or activity that is useful
to people, so here it goes.

> > And a kernel module that relies on non-free firmware is not the same
> > like gnash, evolution-exchange or empathy/pidgin/etc.
> 
> why not?

Those packages are only useful for processing external data that the
system is not concerned about, or communicating via some network
protocols to achieve the same purpose.

What makefiles I run `make' on, or what input files I feed `m4' with,
is not something that should concern anyone.  I see no difference
between these use cases and Gnash.  Or, to put it in "modern" clothes,
I see no difference that you used Google's MUA to reply to me.

> in fact the driver could load free firmware if it exists.

This doesn't matter until such firmware actually exists.  If it
happens, there are many ways to deal with this scenario.