Apart of the Any::Moose indicator: last upstream commit was 9 years
ago, tiny popcon, most reverse dependencies on CPAN have not been
updated since 5+ years.
The Mouse ecosystem is essentially on life-support as the community
has been adopting other, better maintained OO frameworks. I'd rather
encourage our users to use one of those other OO frameworks instead of
having to choose among many of them.
At this point, to me this smells like a candidate for removal.
The last app we had which used the latter (libmousex-getopt-perl),
i.e. libbot-training-perl, has been migrated away from Mouse a while
ago; I just dropped the obsolete dependency we still (mistakenly) had
on libmouse* there, and uploaded it. I'll file rm-candidate proposal
bugs against libmousex-types-path-class-perl,
libmousex-configfromfile-perl, and libmousex-getopt-perl.
So I think the main blocker for removal here is that libany-moose-perl
and libtype-tiny-perl build-depend on libmousex-types-perl, because
Mouse is one of the several OO frameworks they support, and they have
tests that exercise said support. I did not check how these tests
would behave if MouseX::Types was not available at build time.
Investigation to be continued! :)
OTOH, as long as it keeps kind-of-working and we did not reach the
point when it's realistic to remove Any::Moose, I suppose it's OK to
keep libmousex-types-perl in the archive for a while.