[Biojava-l] Open source, standards etc.

[Thomas Down]

On Sat, Jan 22, 2000 at 08:01:02PM +0100, David Martin wrote:
> 
> Open source is definitely the way to go to ensure that the most utility
> is gained from the project. How open source is a matter of question. My
> gut feeling is that the core object libraries should be LGPL or similar to
> allow them to be incorporated into anything without a problem.
> 
> Applications that form part of the archive should probably be GPL or a
> similar license.
> 
> Others may disagree.

As you can probably tell, I'm certainly a believer in OSS.
The LGPL (do we /have/ to call it Lesser GPL these days?)
sounds like it could be the way to go -- I'll check it over
more carefully in the next few days.  Applying licences
designed for C code to Java can be a bit strange when the
ambiguous term of `linking' crops up.

> As regards standards, it is probably worth looking at the OMG biomolecular
> sequence analysis IDL and aiming to make biojava effectively an open
> source implementation of that. In any case ensuring that the OMG
> interfaces are included in biojava would IMHO be a Good Thing (tm) I
> haven't had time to build anyhting yet as the standards were still in
> development last time I was doing stuff in Java. (then a change of job
> etc. etc.. but I have a new postdoc who might pick up the ball and run
> with it.)

The OMG interfaces are interesting, for sure.  But personally
I'm not entirely sure I like the way they are based on the
sequence == (string + annotations).  Take a look at the concept
of a sequence as implemented in the Sanger Centre biojava core:

  http://www.sanger.ac.uk/Users/td2/biojava_core_20000121/

It ought to be possible to use our own interfaces in the core
but add a bridge to the OMG interfaces, if that looks like the
best way to go.

Any views on this?

Thomas.
-- 
``Science is magic that works''  -- Kurt Vonnegut.