[License-discuss] list currently on opensource.org/licenses [was Re: notes on a systematic approach to "popular" licenses]
luis at lu.is
Tue Jan 10 22:34:22 UTC 2017
Changing subject line, since I think this discussion is somewhat distinct
from the main subject of the other email (how to do much better going
forward). And it is going to get into a silly level of detail ;)
(That said, the quibbling over even these fairly small changes between
people who agree 99% of the time is indicative of the politics involved,
which may be a useful reminder/lesson for all of us ;)
On Tue, Jan 10, 2017 at 1:06 PM Richard Fontana <fontana at sharpeleven.org>
> > - I don't recommend merely updating the existing "popular and..." list
> > through a subjective or one-time process. The politics of that will be
> > messy, and without a documented, mostly-objective, data-driven method,
> > it'll again become an outdated mess.
> Luis, I agree.
> I just want to point out something I've said privately (and I think
> publicly as well, if not in a few years), which is that the current
> version of the "popular or strong communities list" is in my opinion a
> mess. It takes the original (flawed IMO) ~2006 list and does the
> * Changes MPL 1.1 to MPL 2.0 (which of course didn't exist in 2006 and
> which is significantly different from MPL 1.1)
I'm not sure "significantly" is quite the right word, at least when
compared to GPL v2->v3. The primary feature of the license (the file-level
copyleft) is intended to have the same scope, with drafting changes only
for clarity. Some secondary features changed (notably patents and
compatibility), but by and large the license is intended to be, and has
widely been used as, a drop-in replacement for the older, deprecated
Contrast to GPL v2 -> v3, where a variety of important qualities of the
license changed (particularly the obligations of the licensee), such that
the two licenses are widely viewed not to be substitutes for each other.
That last statement is an admittedly subjective view, and so perhaps not an
easy one for OSI to make, but it's also hard to argue with, given that an
entire critical segment of the software industry (mobile/embedded) had
clearly come to the same conclusion by 2013 (when the new list was created).
To put that last point a slightly different way, while the primary author
of the license views GPL v2 as deprecated, there is a huge swathe of the
community that still views it as the primary form of the GPL. Discussing
only GPL v3 is, in that sense, substantially different from discussing only
MPL v2 (where there has been no substantial deliberate push to keep people
* In contrast to MPL, the existence of significantly different
> OSI-approved versions of the GPL and LGPL is ignored
That is inaccurate. The version numbers previously provided in the 2006
list were removed, and the links in opensource.org/licenses (which to the
best of my knowledge is the only location of the "updated" list) go to a
disambiguation page, providing information on both licenses.
That said, perhaps my intent would have been better-served by saying "(v2
and v3)" rather than deleting the license information altogether - I'd have
no objection to doing that in the list on opensource.org/licenses.
> * Ignores the fact that CDDL's current license steward has for several
> years had a minor (1.1) update which has not been submitted for OSI
Given that it has not been submitted, it's hard for opensource.org/licenses
to link to it :) But it is fair to point it out as a problem with either
the 2006 or 2013 versions of the list.
> I had thought it might be preferable to return to the original
> "popular list" and just make clear that it is the product of a
> now-distant point in time, but I now believe this solution would
> probably be seen by many as worse than the current approach.
As I pointed out yesterday and when I drafted these in 2012, I think OSI
must offer *something* beyond the comprehensive lists to first time
learners - which is to say, many readers of opensource.org/licenses. So,
yes, the 2013 list is not ideal! But it is better, in 2017, than retaining
the 2006 list or offering only a heavily-disclaimed version of the 2013
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the License-discuss