[License-review] Approval: Server Side Public License, Version 2 (SSPL v2)

[Richard Fontana]

788977On Tue, Mar 12, 2019 at 2:43 PM Josh Berkus <josh at berkus.org> wrote:
>
> On 3/12/19 11:18 AM, Richard Fontana wrote:

> > Josh, I'm honestly really puzzled at your reaction. Looking at the
> > license-review archives from the point in time SSPLv1 was submitted, I
> > believe there was quite an energetic and serious discussion about the
> > OSD conformance of SSPL from an unusually wide variety of commenters
> > including several who don't normally post here (which continued to
> > some degree after SSPLv2 was submitted). Look at Lukas Atkinson's
> > summaries of the discussions. I just don't see a dramatic failure in
> > *this* case.
>
> I'm talking about the last month of discussions, before the SSPL was
> withdrawn.  I don't think the submitters consider that serious
> discussion, and I *know* that folks who have been following L-R don't.

Okay, I've quickly looked at the last month or so, and what I see is
that substantive discussion of SSPL was largely dying down by that
time (despite Eliot's efforts to propose a further modification of
Section 13) and the remaining discussion was basically a tussle
between Bruce Perens and Kyle Mitchell, most of which I thought was on
topic but maybe not all of it. I guess that's what you're talking
about then but I'm not sure.

> I'm at OSLS now, and literally within 10 minutes of posting that I got
> two people coming up to me and saying "thank you for posting that".
> Further, within the last month, I've had folks pinging me on social
> media, asking me WTF is wrong with the OSI.  I've even reviewed (and
> talked a significant OSS contributor out of posting) a polemic on why
> the OSI should be shut down on the basis of this thread.
>
> The OSI only has authority to the extent that we are widely regarded as
> an impartial arbiter of what is and is not open source.  It's important.
> And on the SSPL, we are *not* widely perceived as fair or impartial.

You or those you're referring to seem to be equating OSI with "people
who participate on license-review", a public mailing list open to
public participation subject to the code of conduct. I know that OSI
itself somewhat encourages this conflation of OSI with the communities
around its public mailing lists, but at a certain point we have to
keep them distinct. If I don't limit the scope of the criticism to the
past month, this sounds to me like a complaint that most of the active
participants in the license-review threads on SSPL were hostile to
SSPL. It is obviously true that most license-review commenters on SSPL
were critical of the license, though it had its defenders. Eliot
acknowledged this in saying "based on its reception by the members of
this list and the greater open source community, the community
consensus required to support OSI approval does not currently appear
to exist regarding the copyleft provision of SSPL".  Participants on
license-review are expected to adhere to the code of conduct, but they
are not expected to be neutral or non-opinionated, and I don't think
it is fair to expect everyone to abandon all previously-held views on
a license submitter's business model if it has some relevance to how a
license will likely be used in practice.

In this case, OSI never acted as an arbiter, because MongoDB, Inc.
removed the opportunity to do so.

> >> Where was the legal
> >> discussion around whether the things that Mongo wants are actually
> >> licensable?
> >
> > But that's not the question we face here, although it was certainly
> > touched on by a number of commenters. Rather, it's whether the license
> > "conforms to the Open Source Definition and provides software
> > freedom."  General discussions about how far it would be appropriate
> > or legally practicable to take copyleft (without leaving free
> > software/open source behind) might be appropriate for license-discuss
>
> I don't see these as separable; "how far can/should we extend copyleft"
> is directly related to any discussions around OSD compliance, especially
> 6, 9 and 10.

I think as to "should" I agree with you, but if "can" means "can a
copyleft license do this in some way that works in some relevant legal
system", I think that is less central (I may have made this point in
connection with Van Lindberg's paper critiquing SSPLv1). In any case
there was discussion of both "can" and "should", I thought.

-- 
Richard Fontana
Open Source Initiative