[License-review] For Approval: License Zero Reciprocal Public License

[Kyle Mitchell]

John,

Thanks again.  I think I understand your line of reasoning much better now.

You're absolutely right that copyright owners cannot control every use, and that copyright misuse can kill a license.  Fair use is the classic category of meaningful exceptions.  And you're right to point out 17 USC 117 as another, specific to software.  But I think you overestimate what's protected under that section.  I think L0-R exercises rights well within what copyright affords.

Broadly speaking, 117 permits the "owner of a copy of a computer program" to make copies and adaptations of software as "essential steps" in using it on some computer.  "Essential" is a high bar, but as a threshold matter, "owner of a copy" is not the same as "licensee".

The licensed-not-sold concept comes up in two relevant places under US law, the "first sale doctrine" and the 117 limitation.  The one to read, as I see it, is Vernor v. Autodesk, 621 F.3d 1102 (9th Cir. 2010):

  We read [three prior opinions] to prescribe three considerations that we may use to determine whether a software user is a licensee, rather than an owner of a copy.  First, we consider whether the copyright owner specifies that a user is granted a license.  Second, we consider whether the copyright owner significantly restricts the user's ability to transfer the software.  Finally, we consider whether the copyright owner imposes notable use restrictions.[Note 11]  ...

  Note 11:  Although use restrictions were not dispositive in [the three prior opinions], we considered them in each case.  [Citations omitted.]

Applying the test to L0-R, slightly out of order for clarity:

1.  L0-R is a license.

3.  L0-R imposes a very notable use restriction, a time limit on use with other than published Open Source.

2.  Redistribution, source and binary, is subjected to conditions that ensure the use restriction applies to each recipient.

At least as I read it, the law does not rigidly couple copy-for-distribution to copy-for-use.  It's possible to permit distribution---conditions 1 and 2 of L0-R---without permitting everyone to whom copies are distributed to use without restriction---against the terms of condition 3.  In other words, it's not the law that anyone who comes upon a copy, in circumstances either foul or fair, receives an automatic right to use and adapt, despite any restrictions the copyright holder might like impose.  When there is an outright, unrestricted sale, the copyright owner loses some rights to take back, sue, or impose new restrictions.  But not every transfer of a copy makes an outright, unrestricted sale.

As for the FSF's definition of Free Software, you're absolutely right about freedom 0, "freedom to run the program as you wish, for any purpose".  But consider what that definition has to say about copyleft:

  Certain kinds of rules about the manner of distributing free software are acceptable, when they don't conflict with the central freedoms.  For example, copyleft (very simply stated) is the rule that when redistributing the program, you cannot add restrictions to deny other people the central freedoms.  This rule does not conflict with the central freedoms; rather it protects them.

L0-R sets a rule about freedom 0, imposing source and licensing requirements.  Those requirements protect freedoms 1 and 3, via "access to source code" and license requirements.  GPL and AGPL set rules about freedom 3, "freedom to distribute copies of your modified versions to others".  GPL and AGPL likewise buttress freedoms 1 and 3, also via access to source and license requirements.  I don't think L0-R restricts the "central freedoms" by setting freedom-protecting source and licensing rules about freedom 0, any more than the GPL restricts them by setting freedom-protecting source and licensing rules about freedom 3.

There is an implementation difference.  FSF's definition doesn't mention rules about other freedoms, other than the one the GPLs chose to hook into.  But it leaves open, and begs, the question of whether they could also be Freedom-enhancing, on balance.  GPL-3.0 and AGPL-3.0 are older than Vernor.  Were GPL-3.0 rewritten today, perhaps the authors would still choose to take distribution restrictions out of the toolbox, and leave use restrictions behind.  But both come from the toolbox called copyright.

Once the question of L0-R under OSD is resolved, I'd be very interested to see what FSF has to say.  But I should be careful how much I take on at once.

Many thanks,

K

-- 
Kyle Mitchell, attorney // Oakland // (510) 712 - 0933