I've had thoughts about the "rule vs. standard" question but I haven't been able to reach any conclusion, so I'll just throw them out there.  

My first thought was that the appeal of a rules-based license is that it is the form probably preferred by people who write code because it most closely fits with how they process a decision tree. So I expect it will be more appealing from a license-adoption standpoint. But others have described the flaws; to me lack of flexibility in changing circumstances and the high likelihood of silly results if no common sense is applied are the two most important.

The problem with the standard- based license is that it requires much more trust that the enforcer will be reasonable and not stretch ambiguous boundaries too far. But when you have a license that is perceived as ideologically-based, then there will naturally be a fear of over-enforcement. So what you're inhibiting is the adoption of the software, even by those of good intention. Perhaps the FAQs will be a partial solution, they can at least explain in more detail what the intent was and perhaps provide a governor on over-enforcement. 

So I think it depends on what your goals are, and if you want copyleft software to be more widely-disseminated you have to find a way to provide reassurance that the enforcement will be forgiving of human error, ambiguity, and the differing of reasonable minds. 


On Fri, Apr 19, 2013 at 5:27 PM, Richard Fontana <fontana@sharpeleven.org> wrote:
This is a Harvey Birdman Rule violation cure.

On or about Wednesday, 17 April 2013, in San Francisco, in the late
afternoon, I had a short conversation with Mike Linksvayer concerning
copyleft-next that I believe triggered HBR.

As I recall: Mike asked me if I was still planning on drafting an
Affero-ish variant of copyleft-next. I said I was, and that I had
recently wondered whether my previous assumption (start with a
non-Affero copyleft-next and work on the Affero variant afterwards)
may have had things backwards.

The substantive obstacle I have been facing with copyleft-next (as I
think I explained on this list a couple of weeks ago) relates to the
'rule vs. standard' dichotomy; I believe a simpler alternative to the
GPL ought to (ideally) require greater ease of application and
interpretation. I do not know how to achieve this, and perhaps it is
not achievable. That is to say, the best we can hope for wrt making
copyleft-next "easier" (beyond the kind of structural and textual
clarity and simplicity already present in existing versions) is to
improve in small ways on some of the basic definitions in
copyleft-next and rely on the development of useful authoritative
interpretation. However, I'm still not ready to concede that.

Anyway, as I recall, I said to Mike that perhaps drafting an
Affero-ish variant of copyleft-next might produce insights into how to
draft what I've previously been thinking of as the vanilla version. I
have also previously and separately wondered whether an Affero-ish
copyleft-next ought to be MPL-like in its approach to the copyleft
boundary issue. This is all related to suggestions, made at times by
Michael Meeks and by Bradley Kuhn, that a "Lesser AGPL" may be useful.

- RF

copyleft-next mailing list