On 10/02/2012 01:49 PM, Clint Byrum wrote:
> Excerpts from Alan Robertson's message of 2012-10-02 11:16:47 -0700:
> Count me as a +1 for GPLv3, and a suspicious +1 for AGPL. I like the
> way this license works in principle, but I don't know that it is always
> used in the appropriate way or quite as enforceable as the GPL. Also it
> is considered quite toxic by many organizations, who see it as a deal
> killer to have to expose their exact running code to users. I've heard
> it described as "shareware for open source" by some.
>
> With Assimilation, its hard to say if AGPL is appropriate. The license
> was specifically created to address web based applications like MediaWiki
> or WordPress, where service providers will be handling users' data a lot
> and could in theory abuse their position (see: Facebook) as the trustees
> of said data.
It is also intended to encourage users of the system to provide their
modifications even when they are not distributing their product - ala
Google, Yahoo, and other SAAS-like vendors.
> Assimilation will, at some point, have something that external users
> might connect to such as a web app. For that, AGPL makes sense. But for
> the nanoprobes and even the centralized collector, I'm not sure AGPL does
> much to protect users, and it might taint the implementation enough to
> drive users to other solutions.
I know of no reason for the nanoprobes, or the CMA (centralized
infrastructure) to be AGPL. The thing I like about the Neo4j approach
is that they make the base capability available via GPLv3. I don't have
anything specific and concrete planned to include in an AGPL version,
but I could imagine some things similar to what you suggested. It's
less obvious than it is for a web server, or a database server...
*In terms of "toxic" and "drive users to other solutions"...*
I suspect one of the intended side-effects of the Neo4j approach is that
it tends to drive organizations that can afford to pay for the software
to purchase the commercial version. The commercial versions contain no
more features than the corresponding open source versions, but they
don't have that nasty, *toxic license* (from the perspective of a great
many lawyers).
*I suspect that Neo4j's hope is that it does indeed drive them to
another solution* - purchasing a commercially licensed version. The
overwhelming majority of organizations are already comfortable with
paying for software they use, and getting it under one or another
commercial licenses.
You could argue that this makes it a FUD tax... I'm good with that ;-)
Kevin's comments about the administrative annoyance/complexity/etc of
AGPLed software is a good point. The point would be to have it to be a
FUD tax, not make the admin's life harder. It's worth further thought.
The truth is I don't yet understand the AGPL as well as I do the GPL -
much less know how it works out in practice - so I appreciate his
experience.
--
Alan Robertson <[email protected]> - @OSSAlanR
"Openness is the foundation and preservative of friendship... Let me claim
from you at all times your undisguised opinions." - William Wilberforce
_______________________________________________
Assimilation mailing list - Discovery-Driven Monitoring
[email protected]
http://lists.community.tummy.com/cgi-bin/mailman/listinfo/assimilation
http://assimmon.org/
