On Tue, Jan 02, 2018 at 09:33:08PM -0800, Steve Langasek wrote: > These statements are not in contradiction. Saying "this package complies > with policy version X" doesn't say anything about what version of policy the > package *should* comply with. Our tooling should absolutely be optimized > for reporting discrepancies against the current policy version, not for > second-guessing the correctness of a given Standards-Version declaration. In my opinion, we don't want a situation where packages get to choose which versions of the Policy they comply to. They should always comply to the current version. The reason we have a Policy document at all is so that all the packages work together. Some of this is simple: all manpages go under /usr/share/man, all executables go in one of /bin, /sbin, /usr/bin, /usr/sbin. Some things are more complicated. All in all there's a lot of details to get right for 50,000 packages maintained by 1000 people to form a cohesive, working whole. The Standard-Version header in debian/control is one of the mechanisms we use to keep track of where each package is, as far as compliance. lintian is another. Unfortunately, many of the details are difficult to test automatically, so can't only rely on lintian. Over the past several years, I've spent less time updating Standard-Version in my own packages than I've spent reading this email thread. I don't want to ridicule anyone's concerns, but I'm of the opinion that when it comes to S-V, the concerns seem rather overblown to me. (I would like to offer a tiny squeak of support for those willing to do work to move stuff from out of the source package, when it doesn't need to be there. Homepage, Vcs-*, Maintainer, and Upladers from debian/control, at least, and debian/watch files entirely. Possibly Section and Priority fields as well.) -- I want to build worthwhile things that might last. --joeyh
Attachment:
signature.asc
Description: PGP signature