[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: PBS License

On Tue, 17 Jul 2001, Edmund GRIMLEY EVANS wrote:

>John Galt <galt@inconnu.isu.edu>:
>> >> GPL-ish stuff, the only problem is that you theoretically cannot use the
>> >> OpenPBS license on contributed code, since it implies restrictions (there
>> >> goes DFSG 3).  In fact, the only way you could theoretically contribute
>> >> code is to make the contributions PD, since ANY license implies
>> >> restrictions of SOME type on redistribution.
>> >
>> >I don't follow this argument. In what way does the X11 licence
>> >restrict redistribution?
>> With the XFree86 license, it's tough to find a restriction, but they have
>> one:
>>  The above copyright notice and this permission notice shall be
>>    included in all copies or substantial portions of the Software.
>> this necessarily restricts the redistribution to copies carrying a copy of
>> the license.
>I don't see that as a restriction on free redistribution as it doesn't
>prevent anyone from distributing an unmodified copy of the Software.

This is true, but you asked for a restriction.  I gave you one, albeit a
trivial one.

>It does constitute a (very minor and trivial) restriction on free
>redistribution of modified versions of the Software, yes, but clause 5
>didn't say anything about modification.

I think that the trivial restriction may very well violate clause 5.  BTW,
clause 5 IS about modification, it's the clause treating with
modification.  Other than the preamble, nothing else gives you the right
to modify, so I would HOPE that clause 5 says something aboug

>> >>  Not only is this non-free,
>> >
>> >Even if contributions did have to be PD, why would this make it non-free?
>> DFSG 3: "under the same terms as the license".  Even if we put aside my
>> strict definition of restriction, this license is not going to qualify
>> under it's own clause 5.
>    3. Derived Works
>       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.
>A PD modification can be distributed under any terms at all so it can
>certainly be distributed under the same terms as the original licence
>(whatever they were).

No, actually PD is NO terms at ALL.  If there's a restriction, it isn't
public domain.  If there's a restriction, under Berne it's copyrighted,
since the only way under Berne to release your copyright interest is to
gift it, and an encumbered gift is not legally a gift.

>You seem to have a different interpretation of DFSG 3 which means that
>a modifier must have the right to be as restrictive as the original
>author. It seems to me unlikely that this is what was intended by the
>authors of the DFSG. I also think my interpretation is the more
>natural one just from looking at the text of DFSG 3 "in isolation".

Not the right to be as restrictive, the right to not be bothered with
multiple licenses on the same file.  "under the same terms as the original
license" means to me that a license must allow modifications to go out
under the same license, not for restrictive reasons, but for ease of use
reasons.  DFSG 3 gives the end-user the right to not have a license
enforce a tangled mess of licenses on the final merged code.  Thus Debian
is ensuring the author's option to Do What is Right (tm), and that is to
minimize the number of different licenses on the same bit of code.

>> >> but the packager must realize that they are going to have to give away any
>> >> authorship rights on their modifications and release them to the public
>> >> domain.
>> >
>> >I don't see what is to stop an evil packager from releasing a patch
>> >with a licence saying that the patch may be freely redistributed but
>> >not modified, for example.
>> How would it propagate without the implicit restriction of coupling the
>> license with the product?
>That's an SEP (Somebody Else's Problem). You can't distribute it by
>pigeon post out of a black hole, either, but who cares?

But that's not a license's call.  If a license stated that you couldn't
redistribute by pigeon post out of a black hole, that'd be a DFSG 6
violation, as it's been discussed before that the fact that something's
impossible doesn't necessarily disqualify it as a field of endeavor (IIRC
it was use of RTLinux patches in the *BSD kernel...).

>> >>  In fact, you cannot even require that your name stay attached to
>> >> your changes after they leave your hands, as that could be construed to be
>> >> a restriction.
>> >
>> >Your name would presumably appear in the licence. If anyone were to
>> >remove the licence, the code would become unredistributable, because
>> >no one would have permission any more. This would arguably violate
>> >clause 5, so I would guess that your name would have to stay attached
>> >even if you made your modification PD.
>> If it's PD, what license?
>I meant the author's statement that the software is PD. Perhaps this
>isn't technically a licence, but it serves the same function.
>Anway, let's not get too sidetracked. As far as the original question
>in concerned, you are claiming:
>(1) The PBS licence forces modifications to be freer than the original

Let's avoid loaded words and go with the less loaded "under differing
terms".  Besides, it's less of a misstatement of my position.  Whether or
not the difference is one of freeness, it's still a difference.

>(2) That disqualifies it because of DFSG 3.

Given my caveat, yes.

>You are probably right about (1), but I don't accept (2) because I
>think you have misinterpreted DFSG 3.

Possibly, but I'm not convinced that I have yet.


The early worm gets the bird.

Who is John Galt?  galt@inconnu.isu.edu, that's who!

Reply to: