On Sun, Jun 20, 2021 at 3:40 PM Bradley M. Kuhn <bkuhn(a)ebb.org> wrote:
This may be neither here nor there, but it lead me to look again at
what the
Termination provision of copyleft-next. I see that it hasn't changed much
since the original 2012 drafts.
Here's the 0.3.1 version:
10. Termination
Your license grants under section 1 are automatically terminated if You
a) fail to comply with the conditions of this License, unless You cure
such noncompliance within thirty days after becoming aware of it, or
b) initiate a patent infringement litigation claim (excluding
declaratory judgment actions, counterclaims, and cross-claims)
alleging that any part of My Work directly or indirectly infringes
any patent.
Termination of Your license grants extends to all copies of Covered
Works You subsequently obtain. Termination does not terminate the
rights of those who have received copies or rights from You subject to
this License.
To the extent permission to make copies of a Covered Work is necessary
merely for running it, such permission is not terminable.
Admittedly it isn't too interesting. The most notable thing is the
absence of the complexity of GPLv3 section 8.
Most notably, I see it still contains a patent infringement
litigation
poison pill:
> b. Initiate a patent infringement litigation claim (excluding
> counterclaims and cross-claims) alleging that all or part of My
> Work directly or indirectly infringes a patent.
Upon reading this, I wonder if this is really necessary? I remember when I
first read it back in 2012, I thought âoh, great use of a patent poison
pillâ, but now I'm dubious.
A few years ago, on a CHR mailing list (which may or may not be the
same as the mailing list you referred to) someone once asked if anyone
knew of any case where a FOSS license patent
peace/retaliation/termination provision of this sort had actually been
invoked. As I recall, no one knew of any. This is not surprising. One
reason is that the patent license grants in FOSS licenses are
typically fairly narrow anyway.
These provisions are in most of the non-minimalist FOSS/FOSS-like
licenses that have been publicized since ~1998. It's slightly
interesting that the post-2014 simple noncopyleft licenses UPL and
BSD+Patent (not to be confused with the license sometimes known as
BSD+Patents) deliberately do not include such a provision, in an
attempt to avoid incompatibility with GPLv2 based on the orthodox lore
that the Apache License 2.0 patent termination provision is
incompatible with GPLv2. You might say that these two licenses broke
the taboo that any modern FOSS license with an express patent license
grant has to have a corresponding patent peace provision.
As far as I know, these patent peace provisions were an importation of
a concept in standards agreements, which typically feature patent
licenses that in certain respects are generally broader than what you
see in FOSS licenses.
The reason I'm now dubious about it is that patents are even more
weaponized
and more consolidated than they were previously. Frankly, I'm am just *not*
sympathetic to a litigation between two patent holders, which is the most
likely use of this clause. Specifically, it's a case like this:
Company Foo sues Company Bar for patent infringement. Company Bar exercises
this clause, terminates rights, and then immediately sues a counterclaim.
Foo has no rights under copyleft-next, but Bar retains their rights.
But what about declaratory judgment lawsuits? What about situations where
Bar was actually the patent aggressor out-of-court and Foo preemptively
acted? (I was recently reading that something like this happened in the
real world.)
So, in summary, I think a smart patent aggressor can manipulate the
situation and find themselves being the one in the cross-claim/counterclaim
situation.
So, upon consideration, I think we should cut the patent poison pill
entirely. This has the upside of making the license shorter by 3 lines /
180 chars / 25 words. Can we discuss?
Sure. I think it's worth thinking about why these provisions are in
FOSS licenses to begin with. I think it's probably a bit of a cargo
cult thing. For certain corporate types, I think there's a view that
this is fair given that a patent license grant is being given. It's
possible those corporate types have an unrealistic view of the scope
of the patent licenses they'd be granting. I'm not sure if anyone
really cares otherwise. Note that there are counterparts to these
clauses in typical CLAs.
Back in the era of the drafting of GPLv3, I had the sense that some
people in the technical community viewed these provisions with some
enthusiasm because they thought they were powerful tools to be used
against patent aggressors. I believe the FSF contributed to this view
by using the nonstandard term "patent retaliation clause" to describe
them, a term which I think has since died out. To this day I remember
one stet comment on the unusual patent retaliation clause in
Discussion Draft 1 of GPLv3 (which didn't survive) from someone who
praised the FSF for taking action against "trolls". Of course they did
not mean "troll" in the standard sense of a non-practicing entity. But
trolls in that sense are (from what I have seen) the main practical
patent-related problem for FOSS, and they are mostly immune to license
provisions of this sort. There have been a few situations where a
quasi-troll has turned out to be a GPL licensee, but the likely
"retaliation" in that situation is a copyright infringement claim.
There might be an argument that eliminating a patent termination
clause will discourage patent-holding companies from adopting the
license. But at this point it's pretty clear that the adoption of
future copyleft licenses won't be driven by corporate users, except in
some secondary effect sense.
Meanwhile, since I have your attention, bkuhn, there was at least one
Hindering Backchannels Rule (HBR) cure I was supposed to do based on
at least one discussion we had in the past. One was following
CopyleftConf in early 2020 but I think there was an earlier one from
April 2019. Anyway I can't really remember what we talked about so I
think these HBR violations are not curable, unless you remember.
Richard