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

Re: ocaml, QPL and the DFSG: QPL 6c argumentation.

On Fri, Jul 23, 2004 at 02:38:36PM +0100, MJ Ray wrote:
> On 2004-07-23 13:25:04 +0100 Sven Luther <sven.luther@wanadoo.fr> 
> wrote:
> >| 6. You may develop application programs, reusable components and 
> >other
> >| software items that link with the original or modified versions of 
> >the
> >| Software. These items, when distributed, are subject to the 
> >following
> >| requirements:
> >
> >Now comes the second complication. This entry covers software which 
> >is not a
> >modified version of the original software, but which links with it. 
> >[...]
> As I posted earlier today, I am a little uneasy about this 
> interpretation. Nothing I saw seemed to exclude application programs 

Well, the intent is clear in both the wording of the QPL 6 header as well as
the Trolltech annotations. I don't know if the ocaml upstream realize that, i
think that maybe not, but i will get clarification from them about it.

> (or ...) from c3 and c4. Similarly, the reverse for c3: especially, I 

Err, you mean QPL 3 and QPL 4 here ? prefixing the clauses with the licence is
good practice, since it precludes confusion between QPL 3 and DFSG 3 for

> think about modifications which turn part of ocaml into a linked (but 
> seperate) reusable component. What did I miss?

The clear distinction trolltech has put into the QPL licence wording and the
annotations ? Also the QPL 6 header clearly makes a separation between the
"application programs, reusable components and other software items that link
with" and the "original or modified versions of the Software".

I am no lawyer, but i am under the impression that the intent of the licence
writter comes into play when there is dubious interpretation, but the number
of times this distinction is made in the licence doesn't let much doubt about
the interpretation.

> >The upstream author can request a copy of the items, if they are 
> >distributed,
> >but not openly distributed (in which case he just needs to get the 
> >public
> >version). One could argue again that this would mean a breach of the 
> >DFSG #1,
> >since the right of the author to those software would be considered a 
> >fee or
> >royalty, but the same argumentation as above makes me reject that.
> If this is limited to linked works in the way you describe *and* a 
> transfer cost can be required, this clause seems free. It is an 

Exactly my point.

> obnoxious practical problem for non-initial authors, who are forced to 
> release to particular people under particular licences in order to be 
> distributable. Essentially, it looks impossible to totally copyleft 
> anything which the QPL touches.

I don't follow you, the author of the work linking with the original software
is free to chose any free licence, BSD or QPL, and it doesn't say anywhere
that you have to make particular consesion for the upstream author. The only
thing QPL 6c ensures is that you cannot refuse to distribute the linked
software to the upstream author, if you distribute it to other folk under the
QPL 6.

> As I wrote above and elsewhere, I'm not sure about your 
> interpretations yet.

Let's bring all the discussion about this in this subthread though.


Sven Luther

Reply to: