Control: clone -1 -2 -3
Control: retitle -1 lintian: mark tag as renamed to missing-tests-control
Control: retitle -2 lintian: fix override usage detection for renamed tags
Control: retitle -3 lintian: Wide character in say at Lintian::Output line 156
Hi Thorsten,
On Sat, Oct 24, 2020 at 4:39 PM Thorsten Glaser wrote:
>
> I’ve been recompiling musescore2_2.3.2+dfsg3-10.dsc (currently
> in sid) on latest sid
On Debian stable. this package actually triggers three Lintian bugs:
1. The tag testsuite-autopkgtest-missing was renamed to
missing-tests-control, the old name was not recorded properly. This is
remedied quite easily and will result in a 'renamed-tag' warning.
2. The override comes up as unused, but that does not seem to be true.
Your sources do not declare a d/tests/control.
3. At least on stable your override comment for cute-field, "what the
ever…", which contains the Unicode codepoint U+2026 (e2 80 a6)
HORIZONTAL ELLIPSIS, triggers the following warning:
Wide character in say at
/lcl/lechner/lintian/git/bin/../lib/Lintian/Output.pm line 156.
Lintian::Output::msg(Lintian::Output::EWI=HASH(0x5639082d1c60),
"oh really?! what the ever\x{2026}") called at
/lcl/lechner/lintian/git/bin/../lib/Lintian/Output/EWI.pm line 170
Lintian::Output::EWI::print_tag(Lintian::Output::EWI=HASH(0x5639082d1c60),
Lintian::Tag::Standard=HASH(0x56390e6ec1d0)) called at
/lcl/lechner/lintian/git/bin/../lib/Lintian/Output/EWI.pm line 113
Lintian::Output::EWI::issue_tags(Lintian::Output::EWI=HASH(0x5639082d1c60),
ARRAY(0x56391071feb8)) called at
/lcl/lechner/lintian/git/bin/../lib/Lintian/Pool.pm line 286
Lintian::Pool::process(Lintian::Pool=HASH(0x5639082d2378),
Lintian::Profile=HASH(0x56390955ace0), SCALAR(0x563909616fc0),
HASH(0x56390960a388), GLOB(0x563909692948),
Lintian::Output::EWI=HASH(0x5639082d1c60)) called at bin/lintian line
758
N: oh really?! what the ever…
O: musescore2 source: cute-field debian/control@source VCS-git vs Vcs-Git
We are unsure about the last bug, especially because you did not
report it in unstable (and it would have been hard to miss). The Perl
folks are telling us it is probably a bug in Perl. The output layer
"encoding:uft-8" is active on the handle being printed to. We are at a
loss for now.
This bug is being cloned and retitled to deal with each of the
described conditions. Thanks for bringing them to our attention.
> lintian output has
> become less reliable and more hard to parse in the last few
> months…
I disagree with that broad condemnation. As a community project that
touches the lives of many contributors, Lintian accepted bug fixes in
many wrong places. The ripples you see stem from a careful but
difficult reshuffling of decades of hard work into a modern code base.
Lintian is no longer multi-threaded, and when that is accounted for,
the software runs better and faster than ever. It will also be much
easier to maintain (and contribute to) going forward.
Kind regards
Felix Lechner
