[Bioperl-l] pipeline

David Block dblock@gnf.org
Tue, 12 Mar 2002 07:59:39 -0800


Anybody remember my talk at BOSC (Ewan - you were talking about 
hard-to-find bugs :)

Genquire almost went to full-on automatic schema or code generation.  
There is frustration and duplication in writing the table.sql file _and_ 
the code to get/set it.  Genquire does automagically create the get/set 
methods based on a hash of attributes, and that is a tremendous 
advantage in development time.

I basically abuse perl to do that - not sure it could even be done that 
way in Java.

Of course, when I posted my solution to the Perl Object Oriented 
Persistence list, they congratulated me on reinventing several wheels.

OTOH, rolling my own meant I understood at least some of the bugs I 
found :)

On Tuesday, March 12, 2002, at 04:40  AM, Ewan Birney wrote:

> On Tue, 12 Mar 2002, Imre Vastrik wrote:
>
>>
>> P.S. I'm in no way suggesting that Ensembl as a whole should move
>> towards automatic O<->R mapping and suchlike. I'm merely trying to
>> tickle people's brains with less "orthodox" views.
>>
>
> Completely Understood. It is worth thinking through other options fully 
> in
> any case rather than rejecting things out of hand.
>
>
>> From Ensembl's perspective it is Arne's call and it was interesting to
> note that in the Java port the call was to go for hand written over the
> Castor system due to performance issues.
>
>
Perl may not have as many issues - Genquire doesn't.  I wouldn't 
recommend a dependency
on an entire framework, since your hands may be tied all of a sudden.

Just waking up on the West Coast...
--
David Block          (858)812-1513
Bioinformatics        http://www.gnf.org
dblock@gnf.org        Just ridin' the Coaster...