Most likely people don't want to support perl less than 5.8, or feel that they are justified in insisting that people upgrade to a newer perl if they want to do serious development.I'd consider such reasons abusing one's power as maintainer, I sure hope noone is doing such things, and as far as I understand debian contract prevents should prevent people from doing such bad things. This must be something else.
... and *not* doing such things is an abuse as well, only of the maintainers rather than of power. abusing maintainers (with hellishly complex support load) is not a good way to retain them.
This is a pretty sane attitude for a package maintainer to have, to be honest. Whether it is "true" or not - whether the package CAN be builtusually such behaviour is being struck down - for example marking package as buildable only on x86, while it can be made to work on wider range of architectures.
depends. accuracy is good, but sanity helps too. see the other comment in this thread re: debian-specific patches to core perl bits.
I'd consider defining smallest possible set of dependencies good practice, maybe I should learn something?with an older perl or not - is *completely* irrelevant.
the default perl for sarge is going to be 5.8.x, isn't it? it makes very little sense to support ancient perl installations in any other way than "please upgrade your packages via this supported upgrade path, install reasonably current versions of all problematic packages, and then try again."
if people want to run ancient crappy software (and i mean REALLY ancient, not ancient-as-in-debian-stable ancient..) they are welcome to install crufty historical versions of perl modules from cpan and enjoy themselves... but i don't think anyone is going to say that that's a good idea.
this whole thread is a bit trollish. i think you are probably talking to the wrong audience with your comments.