[Biopython-dev] Possible contributor

[Jeffrey Chang]

Hi Rasmus,

> We (the CCPN project - www.ccpn.ac.uk) think we might make a natural
> contributor to BioPython
[... description of project]

> We do have a few questions:
>
> Where do I find a copy of the license and conditions of distribution?

Biopython is distributed with the Biopython license.  I have put it 
online at:
http://www.biopython.org/static/LICENSE

It's basically the Python license.  I don't believe it can be 
distributed with the LGPL, so we have a license incompatibility.  The 
Biopython license would not mind being distributed with LGPL, but the 
LGPL wouldn't like it much!


> Is this an integrated project (thus with extensive coordination
> requirements) or a collection of independent deposited software? What
> obligations would we be taking on?

It's in-between the two.  People are essentially independently 
in-charge of their own code, but there is some oversight on what goes 
into the project.  People do make minor changes (e.g. bug fixes) to 
other portions of the code base, but major changes and changes to the 
API are discouraged without discussion.  More coordination comes in 
when it's time to make releases, when I need to make sure that 
everybody's code is in working order and ready to be released.

If you were to submit your code as part of biopython, we would probably 
give you your own package in the Bio namespace, probably Bio.CCPN.  
You'd be free to do whatever you wanted under there.


> Do we have to deposit to your CVS repository and who would get write
> access?

If you want the code to be distributed with Biopython, then it would 
have to be in the CVS repository.  We have thus far been relatively 
liberal about handing out write access, and haven't run into any 
problems yet.  We would likely be able to give out accounts to anyone 
in your project that needs it, within reason, I suppose.


> We follow a slightly different set of style guidelines, in that we use
> internalUpperCase instead of separated_by_underscore, and we generate 
> our
> own HTML documentation format. Would this be a problem?

If you're familiar with our code base, you'll notice that we don't 
always follow our own guidelines consistently!  :)  It is unlikely that 
it will ever get unified, unless we happen to come upon a large 
increase in resources.

As for documentation, that's not a problem.  Brad has been wanting to 
move to a more distributed documentation format, to make it easier for 
package maintainers to write their own documentation.


> Can we just put our name and URL on the ScriptingCentral page, and 
> might
> this be better than contributing to the CVS?

Yes, it would certainly be appropriate for you to do that!

I don't know if it would be better than contributing to CVS -- depends 
on what you want to get out of it.  While the projects cover the same 
general area, I'm not sure how much overlap there is between the two 
projects, that is, how many people now are using both Biopython and 
CCPN.  If the overlap is low, then many people would end up downloading 
a lot of code they don't intend to use.  If the overlap is high, then 
distributing together would simplify the installation process.

Also, please consider the distribution cycle.  Historically, Biopython 
has had releases about once every 6 months.  That's about the amount of 
time to accumulate enough new code, fix bugs, and for someone (me 
currently) to make the release.  If you want faster releases, you'll 
probably have to help out with the builds!

That said, why would someone want to contribute to Biopython?  
Biopython does have a stable, full-featured infrastructure, with nice 
net access, and web, CVS, mailing lists.

Also, please read over the Contribution Guide, which talks about other 
considerations and requirements for contributing to Biopython:
http://www.biopython.org/docs/developer/contrib.html

Jeff