Python 1.5.2 illegal ?

[Gregor Hoffleit]

Pardon me. Just a request for clarification.

Does CNRI consider the use and distribution of Python 1.5.2 illegal ?

It looks like CNRI is currently owner of the Python sources, and therefore
they can change the license under which they distribute these sources at
their free will. They choose to come up with a new license for their next,
last Python release, 1.6. That's fine.

Still, I got Python 1.5.2 under the terms of the old Python license. Suppose
it happens that I like the old license much better. According to the terms
of the (BSD like) Python 1.5.2 license, nobody could stop me from forking a
new Python development tree, forked from the 1.5.2 sources. That's what
happened to SSH 1.x, and now we have OpenSSH (and everybody is using that
;-). We could even call it FreePython and release it under the GPL. The old
Python 1.5.2 license would allow that!

Now earlier statements in this discussion made it look like CNRI would deny
that we had this right to fork and relicense. Is that impression correct ?
Then it sounds like CNRI is telling us that the 1.5.2 (and earlier) release
was illegal to use and distribute, since its license was invalid.


Similarly, I think PythonLabs had the option to fork Python 2.0 either from
the to-be-released Python 1.6 under the CNRI license, or from the old Python
1.5.2 tree under the old Python license. I understand that much work has
been done from 1.5.2 until 1.6b1, but if the CNRI license really would do
the wrong thing (which is not yet sure), that would be an option, right ?


I haven't researched that in depth, but I have the impression that there are
quite a few Linux apps that might become illegal to use with Python 1.6 if
the CNRI license wasn't GPL compatible. Things that come to mind are PyGtk
and PyGnome, Sketch 0.7.x, Gimp-Python, Gnumeric's Python plugin etc. pp.

Also, it would make linking Python into the to-be-released StarOffice 6.x
impossible, AFAICS.

If that happened, forking from Python 1.5.2 might become a necessity.

    Gregor
    



On Mon, Jul 31, 2000 at 05:35:06PM -0400, Andrew Kuchling wrote:
> CNRI's corporate counsel, for good or ill, does not believe the GPL
> will ultimately stand up in court.  I think this is incorrect -- Next
> backed down and obeyed the GPL's terms in their use of GCC, and both
> IBM and Sun, two entities not noted for disregarding legal niceties,
> consider the GPL binding enough to use for releases of some of their
> own code -- but my opinion is of little weight.