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

Re: issues with the AGPL



On Tue, 24 Mar 2009 14:45:41 +0000
MJ Ray <mjr@phonecoop.coop> wrote:

This is helpful. Thanks. More in-line.

> Greg Harris <glharris@panix.com> wrote: [...]
> > Here's what I think needs to be addressed if anyone wants to make
> > actual progress on this subject:
> >
> >  - What exactly is it that someone wants to do that they are
> > prevented from doing by the terms of the AGPL?
> 
> Use it on their website without being liable for the cost of download
> for code that they have not written or modified.

"This site runs free-app plus X; here's where you can get it." If such
a notice is believed not to satisfy some interpretation of specific
language in the license, what needs to be changed in the license
language to resolve that concern? (It ought to be remembered that
contracts (including licenses) as a general rule are legally
interpreted with an implied term of reasonableness unless specific
language imposes a more rigorous obligation.)

> 
> This is the flip-side of the freedom to distribute when you wish and
> the AGPL problem is a consequence of the confusion over whether you
> are compelled to distribute the whole thing or only your changes.
> Also, whether that distribution must be to everyone who has any sort
> of access to it, only to users who are allowed to log in (or similar),
> or only "when you wish" (which is a phrase RMS used to use when
> explaining the four freedoms - I wonder if he still does).
> 
> >  - What other license would be required to allow them to implement a
> >    derived work, offer it to public users, and refuse to share their
> >    changes? (I'm aware that's a loaded phrasing, but that is the
> >    hypothetical problem as seen by those who choose this license.)
> 
> It's not only a loaded phrasing, but I don't understand it.  Surely
> it's not suggesting that we're only objecting to the AGPL because it's
> the AGPL and not some other licence?  That's not it at all.

I was unintentionally obscure. I'm making an assumption that the AGPL
is intended to address the following type of hypothetical situation.
Developer A releases an interesting cloud application with future
potential. Choose-your-random-malicious-corporate-bad-actor Z adds some
nifty widgets to the free application and runs it on its servers for
hefty subscription fees. Z withholds the source code for its
modifications; A is not happy. Because A foresees this possible
outcome, he or she wants to prevent it by appropriate licensing terms.

How that gets accomplished is always susceptible to improvement and
suggestions. If developer A has a valid interest in seeking to preserve
free availability of the source for derived works, there ought to be
some way to accomplish that. And a license that does this, and no more,
ought not to be characterized as non-free.

Now, that leaves open the question of whether the AGPL does "more" in
some unacceptable way. This is where, at least in my opinion,
productive discussion would occur.

> 
> >  - Is this discussion really about nothing more than what
> >    "distribution" means in the context of a public-facing
> > server-based application? 
> 
> Mostly but not exactly: I think it's also about who is a user in the
> context of a private server-based application on a public-visible
> server.

As a working first approximation, it seemed to me that this is exactly
where developer A's concerns get triggered: offering the
derivative application to public users but withholding the source. At
the core, if the application offers no functionality to the operator of
the server without those public users, that's where developer A wants
to see derivative-work source code made available. 

Arguments that start from the premise that the derived work is running
only on "my" machine and that therefore no other concerns are
implicated miss the mark, in my opinion. At least they are unlikely to
persuade developer A to change his or her position. 
> 
> Hope that explains,

It helps.


Reply to: