[pmc hat off]
Jason Bacon <jtocino%gmx.com@localhost> writes:
> We discussed this a few years back and I think the consensus was it
> would be OK to add a science category, but then my work priorities
> changed and it fell off the radar. I have some things in wip that
> would go under science.
I think the key question is how many packages would end up there,
between:
- things not in pkgsrc that should be added
- things that properly would be moved, but NB we do not move as a
cleanup process because
* VCS does not yet support history across moves
* upgrade tooling does not all cope with PKGPATH changing
and therefore the churn is worse for users and developers than it
helps.
I can see that there are things that would fit, but if it's only going
to be 10, that's what misc is for. It seems that if it is 50 somebody
could easily post a list and establish this. With that established I
don't have a problem with someone creating a science category.
> There are some things already poorly categoried due to the lack of a
> science category like biology/gromacs, maybe biology/chemtool,
> math/py-scipy, etc. I'm not suggesting they should all be moved, but
> it would be nice to stop adding to the confusion (for those of us who
> care about categories).
Are you suggesting that biology/* be moved to science, in a future world
where moving doesn't hurt? I am guessing not.
FWIW I see scipy as squarely belonging in math even if there is science
despite the name, just from reading DESCR. The nature of math is that
it's generally useful, and lots of what is in scipy is reasonably usable
in engineering, rather than being specialist software for a particular
research discipline within science, and not usable in an engineering
context.
Attachment:
signature.asc
Description: PGP signature