Re: GR proposal: the AGPL does not meet the DFSG (take 2)
On Thu, Nov 19, 2009 at 11:59:54AM +0100, Wouter Verhelst wrote:
> On Sun, Nov 15, 2009 at 02:37:51PM +0100, Bill Allombert wrote:
> > Hello,
> > I would like to move the discussion to debian-vote where it belongs.
> > I'd like to apologize to have started this cross-post in the first place.
> > (please CC me).
> Actually, I'm thinking it's probably more on-topic on -legal in this
> stage. But whatever.
Well, I already sent a previous answer to debian-legal (by mistake, but still).
> > On Sun, Nov 15, 2009 at 04:04:49AM +0100, Wouter Verhelst wrote:
> > The idea was that if you distribute it in source form, someone else might
> > start to run the software on the public internet and then the 'your modified
> > version must offer' clause take effect.
> > So if you are a service provider, is the AGPL trivially bypassable by
> > 1) not accepting it and 2) having someone else doing the modification for you
> > but never actually providing any service themself?
> I'm not sure, but I'm also not sure how that is relevant to the
> discussion of whether the AGPL is or is not free?
> If this requirement cannot be bypassed, then the clause does apply and
> any argument on the freeness of the license that is based on such
> bypassing cannot be valid.
My point was that compliance with section 13. lay squarely on the shoulder
of the developer modifying the software since it is obvious that the AGPL
drafters did not intent to put any liability to the service provider, and
so my objections 2.2 and 2.3 stand. In particular, the developer has to offer
the source code for as long as a service provider is using it, even if the
developer is unaware of it.
> > So it is not "with every network conversation", but it is "all users".
> I guess the wording could/should be modified. If the license were to say
> that the offer must not be removed, and must be put in an appropriate
> place that is available to all users interacting with it and should not
> be hidden from view, that would probably be better. Then again, that's
> quite a convoluted to say something like that.
> Having said that, I'm not convinced it makes the license non-free. If
> you use a license on a piece of software that it was not meant for in
> the first place, then of course it will cause problems. AIUI, the AGPL
> was written for web apps, not for general-purpose software, and in that
> context this phrasing causes no ill effects.
There is nothing fundamental in webapps which make webapps code improper for
use in non-webapps applications. The right to take code in a project and
reuse it for a totally different purpose is a fundamental free software right.
> If someone were to use the AGPL on a web server, that would be a
> problem. But I don't think that's the case, is it?
Someone might want to reuse code in an AGPL software inside a webserver,
but the AGPL prevent that. How a license that disallow use in web servers can
be more free than a license that disallow use in genetic research ? How can it
satisfy DFSG 6 ?
In any case, thanks a lot for your contribution to the debate.
Imagine a large red swirl here.