Re: A GNU Distribution

[Brandon Invergo]

> Of course, none of these are deeply technical points.  But I don't see
> the differences between the existing distros as all that deeply
> technical, either, so it's not a barrier in my mind. 

Indeed, I think most distros start as a single descriptive goal: "I want
the most user-friendly distro," "I want the fastest distro," "I want a
distro that focuses on simplicity and elegance" and so on, rather than
differentiating themselves in any technical way. Proclaimed technical
differences in the end (such as speed) are subjective or negligible, in
my opinion. Ok, so what is the goal behind building a canonical GNU
system? Well, that's pretty easy: GNU already has several pages of
philosophy that have been guiding the project since the 80s.

So: 
1) The GNU system was announced and its purpose was declared.
2) All of the pieces are now arguably in place for creating this system.
2b) Meanwhile, other systems have been put together in an indirect way
(via non-free GNU/Linux distributions) to meet the GNU project's goals. 
3) The present, short-term goal would be to finally fulfil the plans
laid out in #1

I'm repeating myself a bit from my previous emails but I think that the
project should be completed with its original stated goals as the
primary motivation to do so. 

Being user-friendly? To me, this first of all mostly implies a
user-friendly desktop environment, which Gnome already has as its
guiding principal and has achieved nicely. The GNU system architects, in
my opinion, shouldn't have to worry about that as a primary concern; as
long as they include Gnome as the desktop environment, they get user
friendliness for free. System tools, like package management, just need
to be integrated into that environment (using, eg. packagekit) to
benefit from its user-friendliness.

Being the number one free OS? I think that will come for free as well,
though it will depend on some momentum. Let's face it, when people think
"free OS", they think GNU. Then, when they want GNU, they turn to one of
the free GNU/Linux distributions. If a canonical system is released,
people will turn to that instead (though I don't anticipate the other
non-free distros going away, of course; their presense as alternatives
is good). I think everyone would love the giant "Download GNU" button on
www.gnu.org to just download GNU rather than or in addition to
presenting us with a bunch of choices. Anyway, the point is that GNU has
some 30 years of reputation behind it already. The release and
subsequent ongoing refinement of an official GNU system would generate a
lot of (hopefully sustained) buzz. Will it become the number one system
with the Linux-libre kernel? Quite probably. 

Will it become the number one system with any Linux-derived kernel?
Realistically, it will be difficult in the near future due to hardware
support. My laptop, like many, uses integrated Intel wifi, which
notoriously depends on non-free firmware. Now I know better, and I know
what freedom-supporting options I have. But if buying specific hardware
is a prerequisite, you lose a lot of the people who would install the
system to try it out on a whim (an important group); unless they're
already familiar with the argument for freedom, they might just view the
lack of hardware support on a very practical level and never even give
the system a try. Nevertheless, this is not a problem for GNU system
architects; it is a major task for the Linux-libre developers and for
free software activists. Hopefully, such hardware restrictions will be
stripped away in the ongoing efforts to work with manufacturers, and
eventually all of the users of Ubuntu et al would have no technical
reason for not switching to GNU. Until then, this is no reason not to
release the system for those who want a truly free OS and who understand
what is required to run it. 

Sorry, another wall of text from me. Also, I would like to say that yes,
I understand that I'm a new kid on the block and that my words don't
carry much weight. But, I'm excited to be here and to contribute what I
can... :)

-brandon

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