package dpkg tech-ctte clone 575786 -1 reassign -1 tech-ctte retitle -1 reaffirm that violating Debian Policy deserves RC bug thanks
[ dropped debian-policy from CC ] Raphael Hertzog wrote: > I don't agree with your interpretation. The packaging system doesn't refer > only to dpkg but to the whole system: apt + dpkg. What is 'packaging system'? If you talk about policy, my PDF viewer failed to find a single line about APT. And the error comes from dpkg side, this has nothing to do with high-level package managers. Without a support from dpkg side, this usage of Replaces is totally useless. > Currently the non-regression tests of dpkg ensure that doing a dpkg -i > in your situation effectively fails: > http://git.debian.org/?p=dpkg/pkg-tests.git;a=tree;f=t-conflict-provide-replace-virtual;hb=master Wrong regression tests don't change the severity of bugs. > So it's not really a mistake and I don't agree with the RC severity. Until you modified a policy to exclude that paragraph, it is. > I agree it would be nice to be able to tell dpkg somehow to take the > decision alone without having to resort to remove with --force-depends > or similar. Thus I'm keeping the bug open at severity important for now. It's not "would be nice". It's policy requirement. Obey it or change the policy and the archive. > But please understand that it's not a priority for us right now. I've already realized that after filing my first bug against dpkg. > If you > want to see this fixed, by all means please try to provide a patch and > stop raising the severity. I don't agree you have a single argument to lower the severity. I pass this question to tech-ctte. -- Eugene V. Lyubimkin aka JackYF, JID: jackyf.devel(maildog)gmail.com C++/Perl developer, Debian Developer
signature.asc
Description: OpenPGP digital signature