I have me a bug, but before I finish weighing in with my own pros and cons, I 
thought I'd ask -dev, in particular since I know the creators of things like 
catalyst and such are on here, and the results of this particular bug 
(feature request) might be of interest to them.

A brave (perhaps bored, perhaps just frustrated) user has come up with an 
ebuild ( http://bugs.gentoo.org/show_bug.cgi?id=95038 ) that should (he says 
it does, but I haven't had a chance to set up a chroot to test it in) build 
an extremely minimal perl that's capable of some core perl functionality and 
litte else out of the box. Now for the perl herd, this would imply that we 
would need to come up with some more perl-core/* ebuilds to fill in any gaps 
that the minimal perl install has.

Cons are the addition of a bunch of ebuilds to cover the functionality that 
the minimal install lacks.

Pros, though, are pretty big, once getting past the first hurtle of pain. We 
would be able to eliminate the conflicts between perl coming with module foo 
and ebuild for foo, which is usually at or greater than the same version as 
the one that came with perl (usually greater) A lot of our collision protect 
bugs would dissipate rapidly I think. It would save space (with his use 
flags/architecture, perl went from 12,300K to 930K) on livecd's. I have a 
weak unfounded suspicion that it might help with the whole openssl->perl loop 
that expresses itself when some folks are building a stage1.

I'm completely on the fence here, so any feedback welcome :)

-- 

-----o()o---------------------------------------------
Michael Cummings   |    #gentoo-dev, #gentoo-perl
Gentoo Perl Dev    |    on irc.freenode.net 
-----o()o---------------------------------------------

Attachment: pgpdJdsKhjFhT.pgp
Description: PGP signature

Reply via email to