[License-review] Approval: Server Side Public License, Version 1 (SSPL v1)
Richard Fontana
fontana at sharpeleven.org
Tue Oct 16 23:06:55 UTC 2018
On Tue, Oct 16, 2018 at 09:03:02AM -0400, Eliot Horowitz wrote:
> This license is being submitted for approval by its steward, MongoDB, Inc.
[...]
> Open Source Definition item 9. License Must Not Restrict Other
> Software.
I haven't yet thought too much about SSPL but I see this (whether SSPL
conforms to the letter [or spirit] of OSD9) as the main policy issue
raised by this license, though there may be additional practical
problems with it, particularly concerns around ambiguous terminology
in section 13 (and I suppose the two sets of concerns are closely
related).
> This item of the OSD states that the license must not place
> restrictions on other software that is distributed along with the
> licensed software. "For example, the license must not insist that all
> other programs distributed on the same medium must be open-source
> software." We think our license is consistent with this element as it
> is written. First, the Server Side Public License does not place any
> restrictions on the use of any software, only conditions.
I'm not sure that distinction matters.
> Second, OSI
> itself says, “Yes, the GPL v2 and v3 are conformant with this
> requirement. Software linked with GPLed libraries only inherits the
> GPL if it forms a single work, not any software with which they are
> merely distributed.” But the converse is not necessarily true: if
> software is not “linked” it does not follow that a source code sharing
> condition violates the definition.
That is definitely true.
I think we can agree that SSPL generally would significantly broaden
the scope of the copyleft obligation relative to what we see
specifically in AGPLv3 (at least based on my impression of how the
FSF, AGPLv3's license steward, interprets the license, as well as the
interpretation I have encountered in occasional discussions over the
years with community (non-business-associated) projects using AGPLv3,
as for example GNU Mediagoblin). It occurs to me that SSPL
approximates being the license that for a while many in the commercial
world feared AGPLv3 was.
I'd also note that SSPL is somewhat akin to License Zero, and raises
some of the same issues.
OSD9 is clearly meant to be harmonized with FSF and prevailing
community interpretation of the copyleft requirements of the GPL
license family. So one of the basic questions to consider is, how much
more expansive in scope than AGPLv3 section 13 can a network-services
sort of copyleft obligation in an AGPL-ish license be before the
license runs afoul of OSD9.
One further note: I've heard some raise concerns about the elimination
of the GPLv3 compatibility provision in SSPL section 13. Given the
unusually broad scope of the section 13 obligation, the absence of any
provision for GPLv3 compatibility, and the modern tendency to use
diversely-licensed FLOSS stacks, license compatibility issues may turn
out to be a more significant practical concern than they ordinarily
would be in an AGPLv3 scenario.
Richard
More information about the License-review
mailing list