Re: DFSG#10 and the Open Source Initiative
On Thu, May 27, 2004 at 11:16:56AM -0300, Humberto Massa wrote:
> Yes, I can see. I don't agree, tough.
I assume you meant "though". :)
> >>>>>> This essay is hereby licensed to the reader, granting full
> >>>>>> rights of modification and redistribution, provided every
> >>>>>> derived work of it does: (1) maintain my copyright notice
> >>>>>> intact and positioned near to the top of the text AND (2)
> >>>>>> mentions the original title AND (3) is distributed under this
> >>>>>> same license.
> It strikes me as a free, strong-copyleft, GPL-incompatible license.
> But it fails miserably your reading of the DFSG in two accounts:
> 1. it restricts the *positioning* of the copyright license; AND
This is unacceptable.
If I reuse a couple pages, and my copyright notices are at the end (which
I tend to do, in source), I have to put this one person's copyright notice
at the top. The only thing that I believe would make this acceptable is
if it was actually required by copyright law ("put it near the top or it
doesn't count"), but that goes against my (poor) understanding of copyright
> 2. you can't erase _every_ _single_ instance of the original title;
> you can erase N-1 instance of the original title, but not all of
I'm not sure; I havn't seen this, so I havn't thought about it too much.
My feeling is that it's essentially a "verbatim statement", albeit a
very short one.
I think verbatim statements are at least one case where restrictions to
modification really have been allowed beyond copyright notices and DFSG#10:
the Apache license.
So, I agree that my interpretation is imperfect, but I don't believe yours
is an acceptable alternative.
> This is (at least partially) the logic that makes me conclude that
> DFSG#10 should not be a grandfathering clause: it would lead to a
> stricter interpretation of the DFSG, rendering simple free stuff
Reading DFSG#10 as a grandfathering clause does not make the rest of the
DFSG stricter; it just adds a few exceptions.
I think Branden has better arguments against DFSG#10 than I do:
"It is counterproductive to lead people to believe that they can slap the
BSD, GPL, or Artistic licenses on a piece of software, apply perverse
interpretations to the terms of that license, insist that Debian endorse
their viewpoint, and then excoriate us as hypocrites if we do not
subordinate our judgement to theirs."
We should be interpreting DFSG#1-9 in ways that we believe are correct,
and not force-fitting our interpretations to be sure the GPL always passes.
If we adopt your interpretation, instead of arguing on the merit of a
restriction, we'll find people asking questions like "this is similar to
the GPL's requirement XXX; doesn't that make it DFSG-free?"
> And, all of this without excluding: if DFSG#10 is a grandfathering
> clause, does it reach GPL#8, too? In other words, can a
> GPL#8-restricted work be in main? Or even in non-free? I think this
> alone grants the need for a clarification of the DFSG #10.
This problem appears with your interpretation of DFSG#10, too. I agree
it's a problem--unless DFSG#10 is read as a complete no-op, it's hard to
exclude GPL#8, but it's clearly non-free.
(Of course, DFSG is "necessary but not sufficient"--even if a work passes
the DFSG, Debian isn't obliged to include it. Consistency is good, though.)
>  insert 1000-post-long flamewar about the meaning of the word
> software without coming to any conclusion here. :-)
This flamewar is over, replaced with the "program" and "source" flamewars. :)