Re: Social Contract: Practical Implications
> > They are not exempt.
> > They are also not programs. So, the additional constraints the DFSG
> > puts on programs do not apply to licenses.
On Sun, May 09, 2004 at 12:13:24AM -0400, Glenn Maynard wrote:
> "The license must allow modifications and derived works, and must allow
> them to be distributed under the same terms as the license of the
> original software."
> This does not say "everything but license texts", or even "programs" (a
> word which is used in some other clauses of the DFSG). It's a a violation
> of the DFSG for a license text--as a text, not a set of terms-to be
Nor does it say "all derived works".
That said, at least some of the folks on debian-legal say that the
when interpreting the DFSG, the spirit matters more than the letter.
> You mention the word "program" in the DFSG.
> I think the use of the word "program" in the DFSG is unclear. We clearly
I think the DFSG is referring, specifically to a "computer program" (as
opposed, for example, to a "television program" or a PHS grant program).
What sort of ambiguity do you see?
> The recent GR makes it clear that we're to apply the DFSG to everything
> in Debian. Are we really going to start nit picking the use of the word
> "program" in the DFSG, just like people used to nit pick the use of the
> word "software" in the Social Contract?
If people are going to nit pick about the meaning of the DFSG, why not
nit pick about the meanings of the words in the DFSG?
> Do we really think DFSG#6 "must not restrict anyone from making use of the
> program in a specific field of endeavor" should only apply to programs, so
> a documentation license that forbids use in military training would be
> considered free?
This gets back to the spirit of the thing.
> I think the only reasonable thing to do is to read the word "program" in
> the DFSG as "software". The words aren't synonymous, but to do otherwise
> would be ignoring the result of the recent GR and the new SC text, bringing
> seven DFSG clauses into question.
I think program is a subset of software. It's a significant subset,
but I don't think it's an improper subset.
It's true that most definitions of software indeed include the phrase
"computer program", and that program is a fairly broad concept (including
both interpreted and compiled code, and special purpose languages),
which is also subject to interpretation.
However, software also includes definitions which do not use the word
"program" at all. Anything in the form of bits can be software.
> The only case where I think there may be actual value in the distinction
> is the source requirement: whether we want to allow fonts in main without
> source, and so on. I don't know what the right thing to do is, in that
> case--but I don't think rule lawyering the word "program" in DFSG#2 in
> order to sidestep the issue is a good approach, any more than it was a
> good idea previously to rule lawyer the word "software" in the SC.
At least some fonts have existence as a program (in some cases, the
underlying language is metafont, in other cases it's postscript, there are
many other such languages). Whether all do is a matter of interpretation.