[Bioperl-l] bioperl reorganization
cjfields at illinois.edu
Sun Jul 19 19:35:11 UTC 2009
On Jul 19, 2009, at 10:42 AM, bix at sendu.me.uk wrote:
>> Nice graph! We should probably attempt something similar to
>> Graph::Dependency to actually visualize some of the clusters.
> We've done visualization long ago:
> (I can't find my post to the list about it right now)
Yes, I recall the same. Actually, I found Tim Bunce's
Module::Dependency can do this very easily (I've put a PNG or SVG up
on the wiki for Bio::SeqIO):
It appears to be partly broken, though; I'm not getting any external
deps, just blank nodes. Should be easy to fix and patch, but Tim's
longer maintaining it, though he's open to a co-maintainer (any
> We never really decided on splits based on that, beyond the obvious
Well, we had a very general idea but never really acted on it, partly
due to time but primarily for the same reason I've been pointing out;
we have never specifically defined what is core and what isn't (hard
to take initiative when faced with that). I'm veering towards a very
small core, myself, Bio::Root* only, using bundling to create larger
distributions (something we raised in the past for Bundle::Bioperl and
that Robert brought up via a Task::BioPerl).
Anyway, the plan on the wiki was along the lines of wanting to break a
deadlock between 'core is everything' and 'core is very small', and
initiate the process by splitting the package into a core-main-dev
hierarchy (see the Talk page for that).
> There's a wiki page somewhere that was keeping track of what we
> the best splits might be. Was in mentioned in this thread already?
> If not,
> someone please post it.
Already mentioned, though easily lost in the discussion. As it's
moving a bit beyond core I've renamed it:
I've also add a direct link to the main wiki page. I can plan a new
1.6.x release around all this, but my feeling is that past that future
releases in 1.6.x will likely be dead to focus on this (we can go
through an extensive alpha release). Viva la 1.7?
More information about the Bioperl-l