[Biojava-l] BioJava 1.3pre2

Thomas Down thomas at derkholm.net
Fri May 2 10:57:25 EDT 2003


Once upon a time, on a computer far far away, Schreiber, Mark wrote:
> Hi -
>  
> I know of two additions that I made to biojava-live that would be 
> good in bj1.3. I made some additions to SeqIOTools which makes the 
> new fileToBiojava and biojavaToFile easier and more flexible to use. 

Fine.

[Aside: if BioJava after 1.3 is going to be JDK1.4-only, can we dump
the use of jakarta-regexp on the HEAD and use the standard Java APIs
instead?  You local, friendly, dependency-fascist...]

> I also made a major overhaul of the org.biojava.bio.proteomics 
> package which would be useful to merge in (alternatively that 
> package should be dropped from 1.3 as its non biojava-live form 
> is buggy and unreliable).

It's a bit hard to tell what's happened just by diffing the
code, since it got reformatted too.  But I couldn't see anything
objectionable.  Shall we merge this up on the 1.3 branch?

> I also know of three recent bug reports that haven't been resolved (too my knowledge).
>  
> 1 Problems with printing CircularLocations (I'll try and cover this 
> in the next few days <sigh>).

Thanks.

> 2 Gap Symbol needs to be Atomic?? (not sure of the details on this 
> but its a fairly recent post).

I'm not sure about this one.  I though that the plan was that any
cross-products symbols which include gaps do not acually occupy
any `volume' in symbol-space.  But I'll re-read the original
message and make sure I'm not missing anything.  It would be good
to have a few words from Matthew about this one.

> 3 Upgrade the Blast result system. Not really a bug, just the addition
> of the ability to get all the information made available by the
> BlastLikeSAXParser events.

No, that's not been done yet.  As I remember it, at Singapore
we agreed to make the datamodel interfaces in bio.search
annotatable, so that aligner-specific bits of data can be
shoved into an annotation bundle.  But were we going to add
any other methods?  I know percentage identity has come up
on a few occasions.



      Thomas.


More information about the Biojava-l mailing list