Re: Final text of AGPL v3
On Monday 19 November 2007 02:56:23 pm John Halton wrote:
> On Mon, Nov 19, 2007 at 11:26:21PM +0100, Francesco Poli wrote:
> > The term "user" is not clearly defined. If I get an "access
> > denied" error page through a browser, am I a user of the web
> > application? When I visit a portal, am I a user of the browser?
> > Of the portal application, as well? Of the server-side scripting
> > engine, perhaps? Of the web server? Of the kernel the web
> > server runs on top of? Of the router OS? And so forth...
> > Where do we draw the line?
> I'm inclined to say, "At common sense", taking into account the
> intended functionality of the software. So someone getting an
> "access denied" message is not a "user" of the software (even
> though the software might need to be executed to generate that
> Similarly, as I said in my previous email a few minutes ago, while
> the definition of "interaction" could be clearer, I think it has to
> be taken as meaning "interaction" in a direct, top-of-the-stack
> sort of way. As for the kernel, scripting engine etc., as a remote
> user you don't interact with those directly, but only through the
> intermediary applications sitting on top of them.
I find myself agreeing with Francesco on this point... and as a web
developer who does a lot of work incorporating OSS software into
commercial websites, it's something that is of particular concern to
me. The issue I'm wondering about is a mod_perl application licensed
under the AGPL... once it's started it melds itself deeply into
apache (hence mod_perl). What is the extent of the "contamination"
And, of course, web applications are often a large set of scripts...
dozens upon dozens of individual scripts. If I write a single new
script that adds some level of functionality, but in no way changes
anything else to the AGPL'ed program or uses any existing
functionality, am I required to share my change?
What if something like Zope were AGPLed... would all applications
written ontop of Zope also be AGPLed?
The ultimate issue with the AGPL is that it separates authorship from
distribution. Under the GPL what I am *required* to distribute as
the original author is VERY clear... I only have to distribute what I
distribute. Nothing more and nothing less. But, with the AGPL my
distribution requirements are linked to some other event, and event
which is very difficult to track down since it occurs in a virtual
world without easily defined and discrete events.
I don't know about DFSGness at this point either... but I can say
that I think the license is a pretty awful idea designed to target a
very specific segment of software developers and sure seems like a
use restriction on first glance.
So, let go
...Oh well, what you waiting for?
...it's all right
...'Cause there's beauty in the breakdown