[Bioperl-l] Bioperl hackathon?

Jeffrey Chang jchang@SMI.Stanford.EDU
Mon, 6 Aug 2001 14:48:40 -0700

At 12:41 AM +0100 8/3/01, Ewan Birney wrote:
>The sorts of things that could be considered are:
>(a) making scripts which use bioperl (jason - do you want to speak
>up?) and/or up'ing the whole documentation
>(b) working with biojava/biopython on more interoperation/infrastructure
>In many ways I am keen on (b) - I think we sort of need one/two days with
>key bioperl/biojava/biopython people together working on the live code
>bases on the same machine. I just don't know whether in the middle of a
>conference we can pull this off as I suspect we will need some serious
>concentration to make interoperation between the three projects and things
>like bioperl-db (I should rename this to biosql...) really work. Hmmmm.

Yes, I definitely second this one.  There's a lot more infrastructure 
that we can leverage off each other.  Some examples:

- regression testing - No, not off all of it, but there's a portion 
of regression tests that's sharable across code bases.  In Biopython, 
we've been collecting examples of database records and blast results 
to test our parsers against.  There's nothing about these records 
that are project-specific, and we could get more efficient coverage 
of things if the various projects shared them.

- checking on external data sources - Everytime a database format 
changes, a server goes down, a server changes APIs, etc. breaks our 
code.  For a while, I've wanted a system that will monitor various 
bioinformatics databases to detect changes and alert us when 
something's different.

- naming standards - There's been some intermittent traffic on the 
mailing lists for ways to name databases and sequence accessions. 
Andrew led a BoF at BOSC (I think) on it.  This is *definitely* 
something we don't want to do indepedently for each project.  That 
would result in a lot of headaches later on trying to map in-between 

But for this to happen will require getting us all in the same room 
hacking together.  While the BoF's at ISMB are useful for getting 
consensus, it doesn't seem to generate much usable code.

So, yes, I think it's vitally important to get a hackathon with 
representatives from multiple groups.

That said, I'd like to echo other people's concerns that cost will be 
an issue.  It's an issue for me and likely for other people in 
Biopython and other projects.  A lot of us are doing this work on our 
free time.  Making time to attend a hackathon is tough enough (but 
can probably be justified), but finding resources to fund it is 
nearly impossible...