Package: cruft Version: 0.9.6-0.15 Followup-For: Bug #383364
I thought about it, and although I originally thought that filter files make sense for broken symlinks, maybe this is the kind of thing that should be scripted. For instance, broken symlinks for module-init-tools might only be broken if modutils isn't installed. So a script could manage this kind of conditional checking. However, I don't see any value to scripts explaining that a symlink "must be" broken. Instead, I would give the scripts "may be broken" semantics. Just to tie it into my recommendation for Bug #373708, there could be in /etc/cruft: must_exist_scripts/ may_exist_scripts/ <- globbing could be allowed here maybe? may_exist_filters/ may_be_broken_scripts/ <- and here too? may_be_broken_filters/ These directories would allow the end user a lot of control over the output of cruft's report (especially if coupled with your CAPITALS-naming convention idea). Let me know if there are any real problems with these ideas in theory. If it's just a matter of time and priority then maybe some time later I'll have a closer look at cruft's source. Also, I don't know, but are any features in dpkg 2.0 supposed to obsolete cruft? I'll have to look into that too. - Sukant -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
