Subject: Re: BSD and MIT license "compliance" with the MS-PL
From: Matthew Flaschen <matthew.flaschen@gatech.edu>
Date: Fri, 17 Apr 2009 13:47:18 -0400

Tzeng, Nigel H. wrote:
> 3.2 requires source availability of modifications to source code.  1.9 
> Defines what modifications are.

Yes, but 3.6 says 3.2 applies for /any/ executable-only distribution
(meaning there has to be source offer) of Covered Code.  The Mozilla FAQ
corroborates this.  It doesn't say it only applies for executables made
from modified code.  See http://www.mozilla.org/MPL/mpl-faq.html ,
specifically "I want to distribute Firefox (or other MPL-covered code)
that I have compiled myself but not changed. What do I have to do?"

>> Anyway, I only mentioned MPL as a side point.
> 
> It's not a side point if you assert that all copyleft licenses require
> source distribution with all binary distributions and one does not.  

If MPL didn't require source distribution, then MPL wouldn't be
copyleft.  That doesn't change the definition of a word.  But you have a
very heavy burden in trying to prove Mozilla interprets their own
license wrong on this fundamental point.

>> In this instance, I disagree with license list.  The FSF defines
>> copyleft as, "Copyleft is a general method for making a program or other
>> work free, and requiring all modified and extended versions of the
>> program to be free as well."
> 
> That is the definition for strong copyleft yes.

No.  The difference between strong copyleft is whether new files are
included in the copyleft.  Trivial example:

Original program made from A.c and B.c, both under weak copyleft license.

Derivative program made from A.c, B.c, and X.c.  X.c does not need to be
under the copyleft license.

>> Clearly, if I release a Ms-PL program there is no guarantee all future
>> derivatives will remain free.  One or more branches of the source code
>> could be closed off forever.
> 
> This is true of all permissive licenses.

True.  Ms-PL is permissive and not copyleft

  Perhaps the MS-PL license was not
> accidently named and appropriately categorized as a permissive weak-copyleft 
> open source license.

Where is it categorized as copyleft (besides the FSF list)?

> It is probably not in our interests to call any copyleft license "viral"
> even if we don't particularly care for it.  There are already too many
> nuances in open source licenses to voluntarily reopen that can of worms.

It's not copyleft, so if you don't like viral invent a new word.  To me,
though, it's mostly reclaimed.

Matt Flaschen