[bip] [Fwd: Re: domain name for community site]

[Andrew Dalke]

On Aug 22, 2007, at 5:10 PM, Titus Brown wrote:
> One thing that came across quite clearly at the BoF was that even if
> many of us disliked what we'd see of BioPython, we didn't know what we
> *did* want to do.  There were a lot of ideas flung about, though, and
> hopefully they'll come up again in more concrete form.

When Biopython started there was the same "we should do something,
but what?" feeling.  At a couple of the BOSCs we talked about
possibilities.  For example, bioinformatics tools based on Zope.
(This *was* a while ago.)

There was no cohesive team developing Biopython with shared vision.
As a result, Biopython ended up as a set of only somewhat related
modules, with only minimal API commonality.  Which is being worked
on and improved, but only slowly.

Which is why I say Biopython is not monolithic.  It's a lot
of not-really-fitting-together pieces.  Making it hard to learn.

> For the moment I'd prefer to think about code snippets and  
> tutorials and
> getting some people involved in actually writing a bit of code.  At  
> some
> point something will coalesce, I'm guessing, and then we can discuss
> melding it with BioPython, or releasing it as a series of small
> projects, or ...

For the long term, unless there's a non-trivial project using
the code, I don't think there will be a good, useful coalescence.
I do think tutorials and code snippets are useful, and probably
the best approach for this sort of group.  Something like the
ActiveState cookbook, perhaps?

Stating the obvious - make sure the license is pinned down early
for code and documentation.


				Andrew
				dalke at dalkescientific.com