[Bioperl-l] Splits again
Chris Fields
cjfields at uiuc.edu
Thu Jun 28 23:03:02 UTC 2007
On Jun 28, 2007, at 5:13 PM, David Messina wrote:
> Coming late to this party, I'm replying to snippets from multiple
> emails.
>
>
>> [Chris]
>> what we do about deprecated modules which linger
>> about on CPAN
>
>> [Sendu]
>> Delete them from CPAN seems appropriate.
>
> I coulda sworn this was frowned upon, but a recent thread suggests
> it's totally kosher.
>
> http://www.nntp.perl.org/group/perl.qa/2007/03/msg8473.html
As long as it doesn't show up somewhere to confuse newbies I'm okay
with it.
>> [Sendu]
>> people don't have to
>> 'install' Bioperl, they can still just uncompress the archive (or
>> check
>> out the package from svn) and point their PERL5LIB to the root dir of
>> the package.
>
> Could you elaborate a bit on how this works? How is XS code that
> needs compiling handled? Or the scripts directory? I would love to
> be able to do this.
Maybe Sendu can add to this, but the XS code is limited to bioperl-
ext AFAIK. We could keep that separate until it plays well with
bioperl itself.
Scripts and examples - maybe packaged along with a Bundle?
>> [Sendu]
>> For that reason I very much like the idea of folding the current
>> split-out packages (run, network etc.) back into the core package so
>> everything is one place. Folding them back in should obviously wait
>> until everything is in place and working with core already.
>
> From an organizational standpoint, I'm concerned that with ~900
> modules in core right now, adding all of the additional stuff from
> the split-out packages would make for a daunting directory.
>
> But as you said, this is way down the road, so this proposal
> doesn't bear on the other, closer-to-now issues on the table.
Well, the code in bioperl-db and network complement code in core, so
I agree with Sendu they belong there. They should be under the same
scrutiny as the rest anyway (code, tests, etc), but won't be bundled
unles there is an 'install everything' Bundle.
>> [Chris]
>> Okay. Maybe it's worth doing on a branch as a test run when 1.5.3
>> is ready to go. We'll still need to get thoughts on this from other
>> core devs out there, and it prob. should until everybody is
>> comfortable with the idea.
>
> If we go forward with the CPAN split plan, I like the idea of
> having a trial. We can foresee some of the issues that such a
> change may bring, and yet still more no doubt wait for us once we
> do it.
That's what branches are for; testing stuff out like this.
chris
More information about the Bioperl-l
mailing list