Re: Planning new release of lintian
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
On 2011-02-06 22:38, Russ Allbery wrote:
> Raphael Geissert <geissert@debian.org> writes:
>
>> During the last months we have accumulated quite a few changes and with
>> the release of squeeze some cleanup is needed.
>
According to the BTS, so far we got 42 bugs pending[1]. As far as I can
tell, the last time Lintian closed this many bugs in a single upload was
1.23.0 (23 Apr 2004). I guess that makes "Quite a few" a record. ;)
Since 2.4.3 the new test suite also have 103 less uncovered tags and we
got 1 new uncovered tag[2]. The new uncovered tag is caused by a check I
made, which I have not made a test for... shame on me (also because
t/COVERAGE is out of date in git). Hopefully I will find 5 minutes to
make that one test before the 2.5.0 release.
>> The following is what I think is still missing to make a new release:
>> * Make debhelper 8-related tweaks
>> * Drop tags/requirement of versioned depends for versions satisfied in
>> squeeze.
>> * Fix #604047
>
You probably want to add
* fix libbaz legacy test on non-i386 architectures
That or we may end up with an RC bug when (e.g.) Lucas does one of his
archive wide rebuilds. :)
>> Given the major breakage of overrides, I propose to:
>> * Introduce a new statically-embedded-library tag for packages that are
>> clearly marked as static ($pkg =~ m/-static/ ?) and lower the severity of
>> that tag.
>> * Coordinate with ftp-masters what the list of auto-reject tags will be at
>> some point.
>> * Upload an RC to experimental
>> * Send email to d-d-a mentioning the breakage
>
>> And some time later:
>> * Upload to unstable
>> * Coordinate with ftp-masters to re-enable some of the auto rejections.
>
>> What do the others think about it?
>
> Absolutely, please go ahead! And apologies that I've had so little time
> to help out. Two months go to on my major work project....
>
Personally I am good with this as well. :)
~Niels
[1] http://bugs.debian.org/cgi-bin/pkgreport.cgi?src=lintian
http://people.debian.org/~glandium/bts/l/lintian.png
[2] Using
$ git diff 2.4.3..HEAD -- t/COVERAGE | grep -e ^-[a-z] | \
grep -v -e ^-Last | wc -l
$ git diff 2.4.3..HEAD -- t/COVERAGE | grep -e ^+[a-z] | \
grep -v -e ^+Last | wc -l
Note those expressions report 104 and 2 respectively, but that is
because the legacy test suite no longer checks bad-version-in-relation
and that causes a +1/+1 on both results.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/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=1GKm
-----END PGP SIGNATURE-----
Reply to: