[Biopython-dev] biopython on github

Bartek Wilczynski bartek at rezolwenta.eu.org
Thu Mar 26 01:02:59 UTC 2009


On Wed, Mar 25, 2009 at 10:55 PM, Brad Chapman <chapmanb at 50mail.com> wrote:
> Hey all;
> Good discussion on this; I touch on a few points from different
> threads below.
>
Indeed, I'm very happy that we got the ball rolling and more people
now take part in the discussion.

> I fall in between these two viewpoints. Git has more complications and,
> unless we manage those, we risk introducing additional barriers to
> contribution. Imagine looking at biopython on git hub and seeing 10
> different branches for different users, many of which may be old and
> out of date. This could lead to the impression that we are not
> organized toward a single goal. If you are still interested, how
> do you know which ones could use your help and what they are for?
>
> The solution to this is documentation on the wiki. We rely too much on
> the mailing list and expect people to keep up. Peter read my mind on
> this:
>
> Peter:
>> We should probably all write something on the wiki page for our
>> personal forks, describing what you're using it for, what at the main
>> branches likely to be of interest etc.
>
> I started a page over the weekend doing this:
>
> http://biopython.org/wiki/Active_projects
>
> It's a skeleton so add or subtract away. My idea for this is that it
> is for longer projects that could use outside help. It's not reasonable
> to spend time writing up things you'll be finishing in a week or so; for
> that bugzilla does fine keeping interested parties up to date.
>
> Another idea on this page is a specific wish list of libraries for
> future work. This is a starting point for anyone who comes into
> Biopython fresh and would like to take something on. Also, it encourages
> people who have developed external libraries to deal with problems we
> are interested in to consider folding them into Biopython.

Great ideas. I fully agree that we need clear documentation if we want
more people to contribute.

>
> Me:
>> > There is a lot of good material in this thread for new potential
>> > developers. Tiago, it would make sense to condense what you've
>> > written and include it with the Contributing guide:
>
> Tiago:
>> Just a followup on this: I think it makes no sense to put much of the
>> new content before there is an official step of moving to github.
>
> We are serious about moving to Git and need to have the documentation in
> place so others can learn it. You wrote up a lot of good stuff, and it
> will be lost on the mailing list.

Continuing on that topic. I think there are three (more or less
separate) issues here:
1) Describing git usage technically, to make sure all developers have
a smooth transition to git from CVS
2) Describing typical ways to use git in biopython. This is very
important to calrify how we are going to use cool features of
git/github in biopython. I'm not advocating here to write it very
precisely and I'm fully aware that it's going to change over time as
we learn to use things better, but writing things up will help us
understand how we want to use git/github.
3) General contributing guide with coding style and testing framework etc.

I think that point 3 is quite well separated from the other two
points, which are more git related. I think it is also nicely handled
by the current wiki page:
http://biopython.org/wiki/Contributing. It might be mildly adapted to
include some info on git branches, but these will be minor things.

Points 1 and 2 are not so easily separable, but I don't think it's a
major problem. Current version of the
http://biopython.org/wiki/GitMigration
 touches upon them, but it is meant as a temporary info, so it does
not describe how things should be done after we really make the
switch. I think we need to spearate these issues (temporary
arrangements vs. final desired procedures), so I made a new wiki page:
 http://biopython.org/wiki/GitUsage
which is meant as an early draft of such guidelines. This page is
meant to serve as a technical tutorial describing typical tasks in
biopython development.

Please feel free to modify/expand this page and/or send comments to
the mailing list.
I've tried to keep it close to our current development model, but
there is a lot of room for discussion and I'm very open to new ideas.

cheers
  Bartek



More information about the Biopython-dev mailing list