[Python-Dev] Still no new license -- but draft text available

[M.-A. Lemburg]

Guido van Rossum wrote:
> 
> > Is the license on 2.0 going to look the same ? I mean we now
> > already have two seperate licenses and if BeOpen adds another
> > two or three paragraphs will end up with a license two pages
> > long.
> 
> Good question.  We can't really keep the license the same because the
> old license is very specific to CNRI.  I would personally be in favor
> of using the BSD license for 2.0.

If that's possible, I don't think we have to argue about the
1.6 license text at all ;-) ... but then: I seriously doubt that
CNRI is going to let you put 2.0 under a different license text :-( ...

> > Some comments on the new version:
> 
> > > 2. Subject to the terms and conditions of this License Agreement, CNRI
> > > hereby grants Licensee a nonexclusive, royalty-free, world-wide
> > > license to reproduce, analyze, test, perform and/or display publicly,
> > > prepare derivative works, distribute, and otherwise use Python 1.6b1
> > > alone or in any derivative version, provided, however, that CNRI's
> > > License Agreement is retained in Python 1.6b1, alone or in any
> > > derivative version prepared by Licensee.
> >
> > I don't think the latter (retaining the CNRI license alone) is 
> > possible: you always have to include the CWI license.
> 
> Wow.  I hadn't even noticed this!  It seems you can prepare a
> derivative version of the license.  Well, maybe.

I think they mean "derivative version of Python 1.6b1", but in
court, the above wording could cause serious trouble for CNRI
... it seems 2.0 can reuse the CWI license after all ;-)
 
> > > Alternately, in lieu of CNRI's License Agreement, Licensee may
> > > substitute the following text (omitting the quotes): "Python 1.6, beta
> > > 1, is made available subject to the terms and conditions in CNRI's
> > > License Agreement.  This Agreement may be located on the Internet
> > > using the following unique, persistent identifier (known as a handle):
> > > 1895.22/1011.  This Agreement may also be obtained from a proxy server
> > > on the Internet using the URL:http://hdl.handle.net/1895.22/1011".
> >
> > Do we really need this in the license text ? It's nice to have
> > the text available on the Internet, but why add long descriptions
> > about where to get it from to the license text itself ?
> 
> I'm not happy with this either, but CNRI can put anything they like in
> their license, and they seem very fond of this particular bit of
> advertising for their handle system.  I've never managed them to
> convince them that it was unnecessary.

Oh well... the above paragraph sure looks scary to a casual
license reader.

Also I'm not sure about the usefulness of this paragraph since
the mapping of a URL to a content cannot be considered a
legal binding. They would at least have to add some crypto
signature of the license text to make verification of the
origin possible.
 
> > > 3. In the event Licensee prepares a derivative work that is based on
> > > or incorporates Python 1.6b1or any part thereof, and wants to make the
> > > derivative work available to the public as provided herein, then
> > > Licensee hereby agrees to indicate in any such work the nature of the
> > > modifications made to Python 1.6b1.
> >
> > In what way would those indications have to be made ? A patch
> > or just text describing the new features ?
> 
> Just text.  Bob Kahn told me that the list of "what's new" that I
> always add to a release would be fine.

Ok, should be made explicit in the license though...
 
> > What does "make available to the public" mean ? If I embed
> > Python in an application and make this application available
> > on the Internet for download would this fit the meaning ?
> 
> Yes, that's why he doesn't use the word "publish" -- such an action
> would not be considered publication in the sense of the copyright law
> (at least not in the US, and probably not according to the Bern
> convention) but it is clearly making it available to the public.

Ouch. That would mean I'd have to describe all additions,
i.e. the embedding application, in most details in order not to
breach the terms of the CNRI license.
 
> > What about derived work that only uses the Python language
> > reference as basis for its task, e.g. new interpreters
> > or compilers which can read and execute Python programs ?
> 
> The language definition is not covered by the license at all.  Only
> this particular code base.

Ok.
 
> > > 4. CNRI is making Python 1.6b1 available to Licensee on an "AS IS"
> > > basis.  CNRI MAKES NO REPRESENTATIONS OR WARRANTIES, EXPRESS OR
> > > IMPLIED.  BY WAY OF EXAMPLE, BUT NOT LIMITATION, CNRI MAKES NO AND
> > > DISCLAIMS ANY REPRESENTATION OR WARRANTY OF MERCHANTABILITY OR FITNESS
> > > FOR ANY PARTICULAR PURPOSE OR THAT THE USE OF PYTHON 1.6b1 WILL NOT
> > > INFRINGE ANY THIRD PARTY RIGHTS.
> > >
> > > 5. CNRI SHALL NOT BE LIABLE TO LICENSEE OR ANY OTHER USERS OF THE
> > > SOFTWARE FOR ANY INCIDENTAL, SPECIAL, OR CONSEQUENTIAL DAMAGES OR LOSS
> > > AS A RESULT OF USING, MODIFYING OR DISTRIBUTING PYTHON 1.6b1, OR ANY
> > > DERIVATIVE THEREOF, EVEN IF ADVISED OF THE POSSIBILITY THEREOF.  SOME
> > > STATES DO NOT ALLOW THE LIMITATION OR EXCLUSION OF LIABILITY SO THE
> > > ABOVE DISCLAIMER MAY NOT APPLY TO LICENSEE.
> >
> > I would make this "...SOME STATES AND COUNTRIES...". E.g. in
> > Germany the above text would only be valid after an initial
> > 6 month period after installation, AFAIK (this period is
> > called "Gewährleistung"). Licenses from other vendors usually
> > add some extra license text to limit the liability in this period
> > to the carrier on which the software was received by the licensee,
> > e.g. the diskettes or CDs.
> 
> I'll mention this to Kahn.
> 
> > > 6. This License Agreement will automatically terminate upon a material
> > > breach of its terms and conditions.
> >
> > Immediately ? Other licenses usually include a 30-60 day period
> > which allows the licensee to take actions. With the above text,
> > the license will put the Python copy in question into an illegal
> > state *prior* to having even been identified as conflicting with the
> > license.
> 
> Believe it or not, this is necessary to ensure GPL compatibility!  An
> earlier draft had 30-60 days.  But the GPL doesn't, so this was deemed
> incompatible.  There's an easy workaround though: you fix your
> compliance and download a new copy, which gives you all the same
> rights again.

Hmm, but what about the 100.000 copies of the embedding application
that have already been downloaded -- I would have to force them
to redownload the application (or even just a demo of it) in
order to reestablish the lawfulness of the copy action.

Not that I want to violate the license in any way, but there
seem to be quite a few pitfalls in the present text, some of
which are not clear at all (e.g. the paragraph 3).

> > > 7. This License Agreement shall be governed by and interpreted in all
> > > respects by the law of the State of Virginia, excluding conflict of
> > > law provisions.  Nothing in this License Agreement shall be deemed to
> > > create any relationship of agency, partnership, or joint venture
> > > between CNRI and Licensee.  This License Agreement does not grant
> > > permission to use CNRI trademarks or trade name in a trademark sense
> > > to endorse or promote products or services of Licensee, or any third
> > > party.
> >
> > Would the name "Python" be considered a trademark in the above
> > sense ?
> 
> No, Python is not a CNRI trademark.

I think you or BeOpen on behalf of you should consider
registering the mark before someone else does it. There are
quite a few "PYTHON" marks registered, yet all refer to non-
computer business.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/