[BioPython] Phylogenetics trees

[Andrew Dalke]

Greg, of Greg and Lisa, said:
>When I wrote my tree classes, I was unaware that Mavric existed. It is
>pretty much exactly where I wanted my project to be in the near future, and
>I don't think my own code offers any additional benefits. If Rick is
willing
>to allow Mavric to be a part of BioPython, I think it should probably just
>replace my code.

There is a licensing difference between Mavric and biopython.  Mavric
is distributed under the GNU GPL while biopython uses the Python license,
"with the serial numbers scratched off."  It's basically the same as
the modified BSD license.

The GPL says:
> But when you distribute the same sections as part of a whole which is
> a work based on the Program, the distribution of the whole must be on
> the terms of this License, whose permissions for other licensees extend
> to the entire whole, and thus to each and every part regardless of who
> wrote it.

I want my contributions to stay BSD-like, or rather, want some way for
people to use my contributions under a BSD-like license.  That clause
of the GPL means that if Mavric code is added to biopython then any
of my work included in a biopython distribution must also be licensed
under the GPL.

Whether or not my contributions are then available under both the
current license and the GPL or only the GPL, I'm not enough of an lawyer
to say.  (That is, can someone get the biopython distribution, pull
out the Mavric parts and use only the BSD-like license on the rest?
Probably, if the copyright is owned by me.  Probably not if anyone made
any changes under the expectation that the changes were under the GPL.)

That means I will need some other way to distribute my software, which
is a distinct negative point - two code bases, or some way to extract
one part from the other, including notifying people about the differences
in the parts they are modifying.

Therefore I argue against including Mavric code in biopython unless the
inclusion is made under a more lenient license, like the LGPL.
(Lenient in this case means accepting of non-GPL software.)

I suppose this might end up in a yet another licensing flame war (*sigh*)
so I'll just point to http://www.linux.com/news/articles.phtml?aid=7125
for a description and leave it at that.

                    Andrew
                    dalke@acm.org