[Bioperl-l] How to become a bioperl hacker... (and come to
birney at ebi.ac.uk
Thu Feb 20 09:37:34 EST 2003
Bioperl is a completely open group - anyone can join the lists
and we welcome any contributions - but from the outside people might
"who decides who is in Bioperl"
"who gets to go to Singapore"
"can I come along?"
So... here is short, rough idea how this works out:
- There is a small bunch of people who actually make releases and
bundles. Jason was the 1.0 release guy (he made the tar balls), I am the
1.2 release guy (for my sins). They fix alot of the bugs and get alot of
hassle... ... they are helped by ...
- People who write the bulk of the code and documentation. These are
Lincoln, Heikki, Hilmar (for bioperl-db), SteveC, Elia, Peter and BrianO.
These people have cvs accounts on open bio and often commit.
- People who have cvs accounts and fix areas of it or tweak things.
These include BradN (assemblies), Aaron and others with a cvs commit
- People who don't have cvs accounts but run their own code bases and
make good responses to posts on the list about decisions, eg, Ian Korf
(though we slurped in Ian's excellent BPlite a while ago, Ian does not
maintain it in Bioperl), Aaron, Michele etc.
- People who respond to other queries on list, make useful comments and
generally follow things
- People on the list.
There is only one real gate-keeping point in all of this - when you get a
cvs read/write account. this is when you can contribute directly to
bioperl without going through anyone else. When do we decide who gets a
- If you contribute code back to specific people, or post
code/documentation fixes to the list then - as long as you are not clearly
at odds with the overall flow of things (...and we are a very tolerant
bunch of people...) - we'll give you an account
- It is far faster than most people expect. Most people who get a cvs
account feel we give them too early. On the other hand, the developers
really find it a drag commiting other people's code - you have to futz
- What actually happens is that one of the main developers posts to the
root-l list on open-bio suggesting someone gets an account; they give a
quick biopic (this is Chad... working on assemblies) and effectively takes
responsibility for that person for the first couple of weeks (ie, if
someone misbehaved, we'd expect the person who proposed them to keep an
eye on them for a while). Often one other developer will say "yup, this
guy is ok".
- For the hackathons (we've only done two so far!) there is a local
sponsor (Apple - many thanks Apple!) and a local contact (Elia). Elia and
Apple contact the whole "open bioinformatics foundation" for people to
come along, with a rough assignment of "X number of people per project".
So - this year - we have 5 bioperl, 4 biojava, 3 biopython, 2 GMOD, 3
Ensembl and some others. The list is distributed amongst the invitees and
people are nominated to come along. The criteria is good work.
- We then discuss what we are going to do (documenting BioSQL, BioPipe
and database access is what we decided to it...) and then we do it ;)
So --- to become a Bioperl hacker you should:
- use Bioperl (!)
- contribute back - often this is documentation fixes and then
bug-fixes as you push a module hard
- contribute to the discussion about new features and ideas
- write your own modules - a Bioperl module should use Bio::Root::RootI
for its exceptions, Bio::Root::IO for complex IO - if it uses sequences,
it should work with Bio::SeqI etc etc...
Anyone - and I do mean anyone - can join and indeed become the main
people. In fact, this is how both myself and Jason and Lincoln all joined
just by contributing (though Lincoln had a big reputation before he
joined...) and we just rate people by their coding and documentation
contributions, nothing else ;)
So... do join ;)
More information about the Bioperl-l