[BioSQL-l] towards schema freeze

[Hilmar Lapp]

On Thursday, March 20, 2003, at 11:02  AM, Yves Bastide wrote:

>>> * biosequence.molecule is missing (was in the bioperl-release-1-1-0 
>>> branch, but not on the main one).
>> This became column alphabet. The molecule attribute of Seq::RichSeqI 
>> should be in the annotation bundle.
>
> OK.  But it doesn't seem to be filled by load_seqdatabase.pl
>

Just realized I'm ahead of myself. molecule() is still implemented as a 
hash key. Arrgh, gotta fix this.

Do you care about molecule()?

The reason I'm asking is that storing keywords and all of that cr*p in 
the annotation bundle is not on the stable branch (yet).

There are changes of the API in RichSeqI that came along with it 
(keywords() -> get_keywords/add_keyword/remove_keywords being the most 
notable), although RichSeq.pm provides a backward compatible 
implementation of keywords(). I'm wary though of back-porting this to 
the branch. Ewan, do you have an opinion on that?

	-hilmar
-- 
-------------------------------------------------------------
Hilmar Lapp                            email: lapp at gnf.org
GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
-------------------------------------------------------------