Re: Suggestions of David Nusinow, was: RPSL and DFSG-compliance - choice of venue

[Brian Thomas Sniffen <bts@alum.mit.edu>]

Matthew Garrett <mjg59@srcf.ucam.org> writes:

> My goal is to maintain Debian's standards of freedom at the point that
> they are and where I believe they should be. You believe that those
> standards should be in a different place. Given the fundamental
> difference in viewpoint, I'm not convinced either of us is ever going to
> convince the other of anything of significance.

Sure we can.  I might convince you that they're in the wrong place --
and certainly debian-legal is the right place for that discussion.  Or
you might convince me that they are in the right place.  Neither of
those is an axiomatic belief.

And trying to convince people in either direction is a perfectly fine
thing to do.  It helps ensure that the current interpretations of the
DFSG and SC track the project's desires.  I don't expect to see DFSG 4
taken out this decade, but that's no reason not to talk about it.

On the other hand, your attempts to silence debate on the issue
because you're happy with the status quo is worrisome.  I don't see
any benefit from so doing, and I see a risk that the project's
opinions will fall out of synch with its procedure.  Lots of shouting
is a good thing in this sort of activity.

>> I think that compromise is in the wrong place -- DFSG 4 was supposed
>> to be a compromise with the rest of the world, but as far as I can see
>> the rest of the world punted on that.  There isn't anybody using patch
>> clauses to release free software -- djb's software is still in
>> non-free and will stay there, and the La/TeX people have been great
>> about the renaming/patch issue.
>
> That's more than thinking that it's a hair too loose. Patch clauses
> cause practical difficulty rather than philosophical difficulty. In that
> respect they're similar to writing code in an obscure language. I'd
> argue that writing code in Brainfuck is more of an impediment to freedom
> than any patch clause.

Well, no more than writing something big directly in machine code --
and I'm not really clear what the projects' opinion is on write-only
code.  I think a kernel written in direct machine code, or a similar
program in Brainfuck, has no source code.  There is no reasonable form
for modification, so it's not Free Software.  But it's never been
necessary to form a collective opinion on that, so I haven't seen many
arguments either way.

But holding up a probably-non-free thing and proclaiming that it's
more non-free than a patch clause doesn't do much to convince me that
patch clauses are free.  Yes, source-less programs are worse.  But
that doesn't make practically unmodifiable programs free.

-Brian

-- 
Brian Sniffen                                       bts@alum.mit.edu