Age | Commit message (Collapse) | Author | Files | Lines |
|
This makes it possible to publish these files so that other dpkg-related
projects can use them, instead of either embedding copies, or
duplicating the macros.
|
|
aclocal uses this to determine whether it needs to update m4 files in
local trees.
|
|
When we are building the source package, while detecting whether there
is any original upstream tarball signature to be included in the .dsc,
it's the best time to verify them, so that when the .dsc gets eventually
signed there's a certification path for the maintainer that they are
including what they expected to be there.
|
|
This is needed, for example, to verify original tarball signatures.
|
|
This is required to be able to verify original upstream tarball
signatures, as they are expected to be ASCII armored detached
signatures for the upstream tarballs.
|
|
|
|
We use cp() in Dpkg::Source::Package to preserve the file attributes,
but just move() in dpkg-buildpackage where it does not matter.
|
|
|
|
The current default in Debian and anywhere else is to use unversioned
GnuPG 2.x binaries, so there's no need anymore to try these first or at
all.
|
|
This makes it possible to specify more than one package, which might be
needed during transition periods for example.
Closes: #926669
Based-on-patch-by: Frank Schaefer <kelledin@gmail.com>
|
|
We should not clear the selections for packages that are not already
installed in the system, as they will end up disappearing on the next
operation anyway.
Closes: #927752
|
|
This change is the companion to the one in dpkg-genchanges that lists
only relevant packages, instead of all the ones listed in debian/control,
but for the .buildinfo file instead of the .changes file.
|
|
|
|
Warned-by: codespell
|
|
This code got deprecated before the spec had been finalized, but left in
just to make sure no remaining implementations were still using it.
|
|
This got introduced after a request so that we could switch our tools
to a safe subset of the bundling logic, so let's enable it now that it
is available in the minimum supported perl version.
Bundling is in general unsafe, and not future-proof, as new options
might gain arguments which would then get misinterpreted, we should
not really be using it. But bundling the value into the option is a
common pattern and should be allowed.
|
|
The Perl version in Debian stretch is 5.24.1, which is the release that
will be oldstable once 1.20.x gets uploaded to Debian unstable.
|
|
This file contains information only valid as long as the downloaded
sources are also available. So it is closer to a cache than a database,
and as such, there is no need for an uncompressed backup, which can
take quite some space.
Closes: #343578
|
|
|
|
And in follow up on debian-l10n-german
|
|
Update to 2998t.
Improvements from debian-l10n-german
|
|
|
|
|
|
|
|
|
|
Closes: #926665
Signed-off-by: Guillem Jover <guillem@debian.org>
|
|
Closes: #929664
Signed-off-by: Guillem Jover <guillem@debian.org>
|
|
|
|
Warned-by: codespell
|
|
The README was not very clear on the listed software being source
projects. So instead of making the description more complex, just
list autopoint as an unconditional dependency, and list gettext as
only necessary with --enable-nls.
Closes: #929601
|
|
Expand what the builder means. Mention the dpkg-deb option to use to set
the owner and group to root. And reword the prototyping section to be
less confusing with what is currently implemented.
Closes: #929019
|
|
Fixes: commit 438db7a31e074691e70e8ace7fff91631030624b
Closes: #929727
Stable-Candidate: 1.18.x
|
|
If the header is not valid, we should not try to normalize or even
consider it is in any sane state.
|
|
This new dependtry level will also check trigger cycles on trigger
process deferral due to unsatisfiable dependencies.
Closes: #928429
|
|
|
|
|
|
Otherwise we incorrectly override them.
Fixes: commit 52427d9075b4b062fa0db16a70d2c2bda6407857
Closes: #928671
|
|
The code checking whether the amount of errors had been reached, was
incrementing the counter after the check, which meant it was allowing
one more error than expected.
[guillem@debian.org: Move return to the next line. ]
Closes: #924886
Signed-off-by: Guillem Jover <guillem@debian.org>
|
|
|
|
|
|
|
|
Warned-by: lintian
|
|
Closes: #924776, #924777
Signed-off-by: Guillem Jover <guillem@debian.org>
|
|
Update to 2997t1f.
|
|
The find() function does not cope well with trying to traverse an
unexistent directory and generates various alarming warnings.
|
|
On source format 1.0, the default is for the debian/source/format file
not being present, which means we'll start with an empty Format field
name.
Fixes: commit d1d35a56e401181b6d15baf474a7db21d3372a65
|
|
Using -sA by default means the user might lose data on overwrite if
there is already a directory with the same name laying around.
Closes: #910737
|
|
We should use the correct errno value, otherwise the error string makes
little sense, and ends up being very confusing.
|
|
We need to assign a status value in the non end-of-tape branch,
otherwise we are using an uninitialized value, which might be anything,
and could end up making the code proceed even if the archive is
damaged/broken, etc.
|
|
When we are destroying a tar_entry we need to zero the struct, otherwise
we will end up freeing already freed memory and crashing.
|