Komodo in violation of Mozilla Public License? (fwd)

[Dennis E. Hamilton]

Since David has responded, I want to point out a case that I completely
neglected.  Mainly because the MPL simply doesn't apply, in my view.  This
is an important one with regard to browsers (it even applies to the license
on IE5.x and many other commercial products, such as components of MS
Office).

The MPL grants complete license to *use* the Covered Work or a modification
of the Covered Work, without restriction.  This is made very clear in
section 2.1 of the license.  Any code which is in some sense employed to
operate the Covered Work in some customized way is not subject to the MPL
(unless itself a Covered Work or modification thereof).

This programmatic use (and what other kind is there, for computer
software -- when was the last time you communed directly with an application
program?  Actually, when was the first time?) is simply and cleanly excluded
from the reach of the MPL.

Now, this is not what is claimed for the GPL by the Free Software
Foundation.  For the GPL, the FSF argues that any use is viral and the using
program must be available under the GPL.  They do this by making an
extremely broad interpretation of the "derivative work" clause.  This is
what had the FSF issues the LGPL as a separate license.  My sense of this is
that RMS wants to see the transitive extension of the viral property have
maximal reach.  If I were to ask him whether a GPL'd hardware system could
only be used to run GLP'd software, I'd expect him to nod his head
emphatically. Maybe even jump up and down with glee.  I don't have any
quarrel with that (though I think courts may have trouble with that stretch
on derivative work), and there is great integrity and consistency in
Stallman's vision for Free Software.  Those are the GPL rules as currently
interpreted and you play that game or don't make derivatives of GPL'd code.
Simple.

The MPL makes no claim of this kind. First it is specifically stated that
derivative works can have their independent portions subject to a different
license.  Secondly, tolerance for unrestricted "use" is stated at the very
beginning.  Finally, the case of "incorporation" in a larger work also
provides freedom for the non-Covered-Work portions being under an
independent license.

Although the MPL is by no means my favorite Open Source license, it seems to
have successfully contained the GPL virus.  As intended.

-- Dennis

-----Original Message-----
From: Dennis E. Hamilton [mailto:orcmid at email.com]
Sent: Tuesday, April 10, 2001 17:01
To: mertz at gnosis.cx
Cc: Python-list
Subject: RE: Komodo in violation of Mozilla Public License? (fwd)


I had just taken a look and here's my totally-intellectual,
legally-uninformed (TILU) appraisal of the situation:

First, there is a Covered Work (or Works) that were covered by the MPL that
have been used in building Komodo.  That is, some portion, perhaps
substantial, includes code or modifications of code from MPL-covered works.
If that isn't the case, the conversation is over.

So long as Covered Works are relied upon, then the following provisions kick
in without exception:

[ more affirmation of the rules that apply to Covered Works and their
modifications. ... ]

OK, this part is simple.

Here's where it gets tricky:

3.7 Allows for incorporation of the *Covered* *Work* (modified or not) in a
larger work.  MPL is not a viral license and the larger work can have a
different license applied to those portions that are not portions of the
*Covered* *Work*.  However, the *Covered* *Work* (possibly modified) needs
to be identified and the licensing of the covered work preserved.
	You could view this as part of the aggregation provision in the Open Source
Licensing model (and in the GPL).  It is not clear that one needs to
continue to offer the *Covered* *Work* (modifications) as part of
distribution of a larger work, but it would certainly be necessary if the
packaging were such that the Covered Work were easily separable from the
larger work for separate use.  This may just be a place where the MPL is
inprecise.

6.3 Addresses the derivative work case.  Please note that the MPL is *not*
*viral*, though it is not exactly a BSD-like (or Python-like) license
either.  6.3 allows a derivative work to be distributed under a quite
different license and there are instructions about how to use the MPL
license notice with such a work, along with a few simple conditions that
must be satisfied.
	Since Komodo is prospectively a derivative work of whatever MPL'd *Covered*
*Works* they incorporated code from, this is the likely case that applies.
The way to tell is to examine the license for Komodo and see how it
acknowledges reliance on portions of MPL *Covered* *Works* (as modified) and
whether it independently offers the MPL *Covered* *Works* under the MPL.

	Here they be sea serpents.

-- Dennis

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Lulu of the
Lotus-Eaters
Sent: Tuesday, April 10, 2001 14:26
To: python-list at cwi.nl
Subject: Re: Komodo in violation of Mozilla Public License? (fwd)


David Ascher (Komodo Tech Lead) in email asked for clarification of why
I thought the Komodo licensing terms violate the MPL.  I'd like to
clarify for the list, since this IMO is a matter of general concern to
the Python community (David is BCC:'d).

The Mozilla Public License seems to live at:
  <http://www.mozilla.org/MPL/MPL-1.1.html>

The chief sections that apparently conflict with Komodo's commercial
licensing terms:

  3.1. Application of License.
[ ... ]

Komodo is available--as far as I can determine--only under a
substantially different license than the MPL.  Specifically, the
ActiveState license imposes many terms that additionally restrict the
recipients rights.

  3.2. Availability of Source Code.
 [ ... ]

ActiveState does not make source code for their modifications available.
That seems pretty staightforward.

[ ... ]

Yours, Lulu...