[copyleft-next] Update 2012-08-15

Theodore Ts'o tytso at mit.edu
Thu Aug 16 14:23:19 UTC 2012


On Wed, Aug 15, 2012 at 11:38:22PM -0400, Richard Fontana wrote:
> > 
> > Can you clarify what this means, as OSI does not confirm "successors" per se?
> 
> It rests on an admittedly fanciful assumption that both the FSF and
> the OSI would at least begrudgingly accept the role of deciding
> whether copyleft-next 2.0 was worthy of being a successor to
> copyleft-next 1.0.

This raises the question of whether the FSF approves of copyleft-next
1.0 at all.  And/or if the OSI declines to accept that responsibility
altogether, indeed it means "there will never be an 2.0 license".

If the goal is to worry about the trust problem, ultimately, people
will use the 1.x license of they trust it on its own merits.  Whether
or not they want the "or later" option at all will be part of this
question --- and that brings up the tradeoff of possible additional
license incompatibility problems if you leave the option up to the
person who initially puts their code under the license, versus people
choosing not to use the license at all because they don't trust the
organization(s) that are entrusted with the ability to upgrade the
license.

For example, I've always used a GPLv2 only license because I wasn't
sure my goals and interests were aligned with the FSF's, and as it has
turned out when the GPLv3 squirted out of its process, I have been
happy that I had always licensed my code under a GPLv2-only option.

Ultimately, I think you can't weasel out of this.  I'd suggest either
not allowing license upgades at all, or if you do, you will need to
instantiate your own self-perpetuating body, and charge it with very
clear goals about what sort of upgrades are acceptable and which are
not.  The clearer the instructions, the more likely people will be
willing to trust an "or later" clause.

Regards,

						- Ted


More information about the copyleft-next mailing list