[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#536364: RFC: Make some results non overridable

Mike Hommey <mh+reportbug@glandium.org> writes:

> Quoting http://blog.technologeek.org/2009/07/09/227:
>    How to not use lintian overrides
>    Look what just got added to all the packages maintained by the
>    Debian Forensics team:
>    --- md5deep-3.4.orig/debian/source.lintian-overrides
>    +++ md5deep-3.4/debian/source.lintian-overrides
>    @@ -0,0 +1,3 @@
>    +# Avoid warnings if non-uploaders to uploads.
>    +md5deep source: changelog-should-mention-nmu
>    +md5deep source: source-nmu-has-incorrect-version-number
>    You’re doing it wrong.
> Reading this, i'm wondering if some lintian results shouldn't be made
> non-overridable.
> What do you think?

It's the problem that we've tried to address by both showing a summary
of the count overridden tags in the Lintian output if there are no
special flags and adding the --show-overrides option to Lintian.  I know
that ftpmaster does NEW reviews using the latter.

I'm a little nervous of getting in the way of maintainers more than
that.  It's always possible that something will change about how Debian
does things when the Lintian maintainers aren't available to make a
quick update (it's happened before), and overrides are an okay bridge.
Also, Lintian does get used for private packages (we use it to check all
our Stanford-local packages, for example), and in those cases some
overrides that are unreasonable from a Debian perspective make sense.

Russ Allbery (rra@debian.org)               <http://www.eyrie.org/~eagle/>

Reply to: