[Bioperl-l] Identifiable and Describable

Matthew Pocock matthew_pocock@yahoo.co.uk
Thu, 25 Jul 2002 00:59:50 +0100 (BST)


 --- Ewan Birney <birney@ebi.ac.uk> wrote: > On Wed,
24 Jul 2002, Matthew Pocock wrote:
> 
> > 
> > Is this arguing for a tiny encoding interface with
> a single method that 
> > eats an identifier and makes a string?
> > 
> 
> THis si the visitor pattern sort of idea I proposed.
> It is a little bit of
> object overkill, but does lay things out explicitly.
> 
> 

Yes - quite similar. Interface-based design-pattern
lore would sudgest that good designs have one
interface for each part of the system that can be
implemented in more than one way. Keep the interfaces
lean and focused and the polymorphisms take care of
your extensibility and adaptability for you. Somewhere
between filoe-pastry and a lump of melted beads ;-)

> 
> > Matthew

See you at BOSC


__________________________________________________
Do You Yahoo!?
Everything you'll ever need on one web page
from News and Sport to Email and Music Charts
http://uk.my.yahoo.com