[Bioperl-l] Unigene proposal and basic implementation

[Lincoln Stein]

On Tuesday 16 April 2002 12:02, Ewan Birney wrote:
> On Tue, 16 Apr 2002, Lincoln Stein wrote:
> > Personally I don't mind creating two new top level directories for each
> > new class, but whatever the convention is, it should be consistent across
> > the entire project.
> >
> > Here's another style to discuss:
> >
> > 	Bio::Seq
> >
> > 		::IO
> > 		::IO/Interface.pm
> > 		::IO/embl.pm
> > 		::Interface.pm
> >
> > 	Bio::Seq::RichSeq.pm
> >
> > That is, instead of having a Bio::SeqI and a Bio::SeqIO, we have a
> > Bio::Seq::Interface and a Bio::Seq::IO::Interface
> >
> > Yes, this will break everything, but we can create faux packages to
> > smooth over the differences.
>
> This is alot of changes, and I am very leery of that. Consistency is good,
> but I hate changing things so radically (the I --> Interface).

It's not as bad as it looks.  The I -> Interface change doesn't affect any 
user code at all -- just internal bioperl modules.  A find2pl script will fix 
it in one line, and we can create empty stub packages around to act as 
forwarding addresses:

	package Bio::SeqI;
	use Bio::Seq::Interface;
	1;

We haven't reached legacy status yet I hope!

Lincoln