[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

MIT and Expat licenses; licenses ‘similar to’ a BSD license (Re: [DEP-5] [patch] License table: more links and licenses.)



Le Sat, Aug 14, 2010 at 01:26:45PM +0200, Carsten Hey a écrit :
> 
> Shouldn't it be mentioned in the licenses description that the expat
> license sometimes wrongly is referred to as MIT license?

Hi Carsten,

I wonder if the tradition of using the “Expat” name to refer unambiguously to
one of the variants of the “MIT” license is widespread or Debian-specific.  I
see no mention of the Expat license in the lists from Fedora or SPDX™, but
Expat is mentionned in Wikipedia's MIT license page, and on the FSF website.

http://fedoraproject.org/wiki/Licensing
http://spdx.org/licenses/
http://en.wikipedia.org/wiki/MIT_License
http://www.gnu.org/licenses/license-list.html#Expat

I think that the DEP should not fall in the trap of trying to make some
extensive license classification. We actually removed most of what made the
short name parseable during the off-line preparative work, and I will prehaps
go further and propose the removal of the description of the version semantics,
that is for instance: GPL-2 and GPL-3 would be two separate short names, not
two version of the GPL short name. Ontologies and license metadata are probably
better in the scope of another document.

If it provides ‘MIT’ as a keyword, and the full text of the MIT license in
annex, then it will be clear what ‘MIT’ means in the context of the DEP.
Interstinly, SPDX has not yet made up their minds about the MIT license:

http://spdx.org/book/export/html/2548
(Title with no additional content at the time this email is written).

It suggests that it would be better to discuss the issue with them, given the
potential impact of their work, before making our final decision of what
we call MIT, or if we will avoid MIT as a short name.

By reference or directly, we will definitely need to provide to the readers
the full text of the licenses for which we provide a short name in the DEP.


Now, for the BSD:

> Since Berkley removed the advertising clause from their published code
> (but obviously could do this not for code they do not own the copyright)
> referring to the original BSD license just as "BSD" seems to be
> imprecise, I would prefer BSD-4.  Especially the explanation should
> mention that BSD is the original 4-clause variant.  On Debian
> /usr/share/common-licenses/BSD even is a 3-clause BSD license.
> 
> Counting from one to four is more easy than remembering which licenses
> FreeBSD and TNF use nowadays, we should at least consider adding these
> numbers to the licenses description and maybe also make BSD-2
> respectively BSD-3 aliases for FreeBSD and NetBSD.

For the variants of the BSD license that contain specific names in their text,
for instance in the non-endorsement clause, or in the derivation claim of the
NetBSD license, I think that it not possible to use the same keyword for
derived licenses. Said differently, if a work is not copyright of the Regents
of the UCB, it is not licensed under the BSD license. This is also the way
taken by Debian, as the BSD license will eventually be removed from
/usr/share/common-licenses/, and our Policy has been modified to stop allowing
to refer to this directory instead of quoting the full text.

This said, it would be a waste to lose the information that a work has a
license that is identical to the BSD license, except that some names have been
changed. I think that we can implement this by a ‘similar to’ keyword, and
providing in annex a template that must be matched for the keyword to be used.
For intance:

File: pinaillette/tartempion.c pinaillette/tartempion.h
Copyright: 2012 Ulysse Capillo <capillo@tracteur.com>. All rights reserved.
License: Capillo, similar to BSD
	Redistribution and use in source and binary forms, with or without
	modification, are permitted provided that the following conditions are met:
	.
	    * Redistributions of source code must retain the above copyright
	      notice, this list of conditions and the following disclaimer.
	    * Redistributions in binary form must reproduce the above copyright
	      notice, this list of conditions and the following disclaimer in the
	      documentation and/or other materials provided with the distribution.
	    * Neither the name of the Pinaillette nor the name of Ulysse Capillo
	      may be used to endorse or promote product derived from this software
	      without specific prior written permission.
	.
	THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND ANY
[…]

Since I do not think that BSD-3 and BSD-4 could be used for anything else than
BSD unix, and since BSD-2 would be difficult to disambiguate between the
FreeBSD license (generic) and the NetBSD license (specific because of the
derivation claim). I think that it is simpler to refer to the license by the
name of their project, and to use a ‘similar to’ syntax when their text have
been used as templates to create a very similar license for another work. Note
that under these assumptions, not work is anymore licensed under the 4-clause
BSD license: if it could have been labelled BSD-4, then its clause 3 has been
cancelled by the University of California, Berkeley.

For the sake of the reference, here is a link to each license:

BSD with advertisement clause: http://www.freebsd.org/copyright/license.html
BSD after removal of clause 3: http://www.netbsd.org/about/redistribution.html#berkeley
License of NetBSD: http://www.netbsd.org/about/redistribution.html#default
License of FreeBSD: http://www.freebsd.org/copyright/freebsd-license.html
License of OpenBSD: http://www.openbsd.org/cgi-bin/cvsweb/src/share/misc/license.template?rev=1.2;content-type=text%2Fplain
ISC license: http://www.isc.org/software/license (differ by an ‘and/or’).


If consensus converges on using a ‘similar to’ keyword, I will submit a patch.


Have a nice sunday,

-- 
Charles


Reply to: