I'm posting this as someone who went actually through the FSF's RYF process
to the end.
> There is no such thing as a middle-ground with the FSF.
The FSF is _the_ libre software lighthouse. As such, it has set its own set
of well-established principles that are almost written in stone. Not everyone
has to agree to those principles, but that's what principles are for; not to
get crushed by nonentities at the first sign of trouble. Those principles
don't come out of nowhere, they grew historically and they will continue to
grow.
The FSF is not beyond reason even though they have principles. When we were
communicating with them, they were able to find a middle-ground with things
that came up and they offered practical solutions.
Imagine the FSF would find a "middle-ground" with proprietary blobs,
potential spyware and whatnot. I don't think that would be a middle-ground,
that would be plain ignorant the very least, like the larger part of the
industry is. The RYF certification would be absolutely worthless!
> The FSF will not endorse any post-2008 Intel-based (or AMD-based, until
further notice) computer, until every last kilobyte of the Intel Management
Engine (or AMD PSP) has been reverse engineered or zeroed out (currently
Purism can zero out somewhere between 90 and 95% of the IME).
The ASUS KGPE-D16 that can run with libre boot firmware has been released in
2010. It's still available brand-new on the free market and it is still
powerful. So powerful we run all our servers with it.
More importantly, there is a new (as in 2017) mainboard coming up from Raptor
Engineering, the TALOSII (see other threads in this forum). That's where the
progress comes from these days: from small companies with moral principles
and ingenuity.
They don't have to "zero out 95% IME". If you go with the non-SAS version,
there are no proprietary blobs, 100%. AFAIK this board is eligible for RYF.
> In terms of software: it’s an excruciatingly slow process (months long,
and moving targets), even if we already go above and beyond any requirements
for GNU Linux distro endorsement.
I've had a different experience and it was a positive one. We actually
enjoyed the contact with the FSF and their professionalism _even though_ they
were understaffed at that time. Furthermore RYF (=Respects Your Freedom
hardware product certification) is not meant for GNU/Linux, but for the
underlying hardware.
> To be clear for anyone reading my previous reply: I’m not blaming the FSF
for sticking to their principles (it’s their job), I’m just stating the
current state of things, factually speaking. i.e. 99% free isn’t 100% and
therefore there is “no middleground” when it comes to certification—but
then nobody else is even remotely close to 100% with modern mainstream
hardware, either.
It would be great to have more modern hardware. We don't need modern hardware
that disenfranchises you though. Until we have something better™, we'll
stick to "old" hardware that respects our privacy and does what we expect it
to do.
One question remains, why do some (if not most) people _still_ not understand
in 2017 what this all is about. Why do people still strike a blow for
companies (looking at you, Intel & co.) that treat you like crap. It really
beats me.
