Bioperl: Bioperl CORBA stuff

Ewan Birney
Wed, 26 Apr 2000 17:00:04 +0000 (GMT)

This is mainly for Jason, but just so other people are kept in the loop -

I have

	a) committed my changes to bioperl-corba-server

	b) started the bioperl-corba-client repository

I work intermittently on the corba stuff (it is great for me as it doesn't
really require that much coordination) but I guess as we both are working
on it now, we should be more careful. As I expect you will have more
spare time than me (not certain), you should drive the project - but we
should both post to this list to keep everyone in the loop.

Ok. So - what works?

 Server side we have 

	PrimarySeqIterator (aka - databases for database searching of
	A half-cooked Seq object
	A half-cooked SeqFeature/SeqFeatureIterator objects
	A half-cooked BioEnv (root place to get objects object).

I think the key thing to have is the SeqDB object. Once we have a working
SeqDB object ontop of the Bio::DB::SeqI interface then we are truely
open for business.

All server objects inheriet off Bio::CorbaServer::Base. This does 
ref counting for memory management and also holds the POA (portable object
adaptor) needed when one object makes some new objects.

bioperl-corba-client just has PrimarySeq and PrimarySeqIterator objects
and a SeqIO system that treats a saved .ior file as a SeqIO file (cute).

Again, there is a Bio/CorbaClient/Base object for reference counting.

What needs to be done

a) SeqDB server side

b) more client side objects

c) testing across networks etc. and memory handling.

For the other projects - if you are listening in - bioperl-corba-server
will provide basic access to bioperl sequence objects, including all the
nasty parsing stuff and these index systems and therefore give you an
immeadaite sequence parsing functionality...

It should be pretty sweet if we can get this all to work together.

Over to you jason...

=========== Bioperl Project Mailing List Message Footer =======
Project URL:
For info about how to (un)subscribe, where messages are archived, etc: