Subject: Re: Followup on Exhibit B licences
From: Matthew Flaschen <matthew.flaschen@gatech.edu>
Date: Wed, 07 Mar 2007 07:05:04 -0500

Christiaan Erasmus wrote:
> Rick
> 
> Please let me know if we should take these discussions private as we are
> starting to drift to a point were it may not add value to the OSI
> discussion list readers any more?
> 
> In our case we have to change the jurisdiction in the MPL clause 11 from
> the "Northern District of California" to South Africa, where we are
> based, which then results in a new license being created as stated in
> clause 6.3, which will then have to be re-submitted for approval.

You don't have to.  IANAL, but it is questionable whether it even makes
a difference.  If neither you nor the alleged infringer is in that
district I suspect the jurisdiction clause would be ignored.

> This sounds like a small irrelevant change but a litigation in America
> would be extremely costly and software patents are also not recognised
> in South Africa.
> 
> ...
>
> We are therefore currently stuck with the OSL that I am having a look at
> but as stated I think that the better option is to wait for the
> GPL3/LGPL3 finalisation.

You should be aware that the most recent GPL3 draft
(http://gplv3.fsf.org/draft) has an explicit patent license, and
explicitly disallows choice of law clauses.

> As you will see in the OSL "This License is Copyright © 2005 Lawrence
> Rosen." so how could it have been OSI approved in 2004?

Old version.

> On the contrary, this is about what is to the benefit of the OS
> community and the OSS landscape as a whole.
> 
> I also believe that there are many more code bases that will be open
> sourced if a solution can be found.

This isn't worth much, if it requires compromising the OSD.

> As stated I do not think that the TPL, if used for a web based
> application, contravenes the OSI definitions or the spirit of open
> source.

A license is either OSD-compliant for all applications, or not.  I
believe yours is not, primarily because of the #10 issue.

>>>> Do you concur, or am I missing something?
>>> Attribution is an approved OSI concept so I do not see the problem.
>>
>> 1.  The OSI does not "approve concepts".  It certifies licences.
> 
> Agreed, but do you agree that attribution clauses are present in some of
> the certified licences?

Attribution is a very vague term, and Exhibit B type attribution has not
been approved.

>> I can only guess -- struggling to be charitable -- that the notion of
>> code and license reuse is a new thought to most of you, as (I would
>> guess) open source is, in general.
> 
> This is where different ideologies come in to play and why not all OSI
> approved licence applications can be distributed together, most basic
> example being GPL and BSD?

Well, new BSD is GPL compatible, but I see your point.  However, these
are separate licenses.  I'm still allowed to use any old BSD (or GPL)
code in any application, with any interface.  I can't say that of your
license.

Matthew Flaschen