Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
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.
|
|
There was a bug in the killproc function implementation that made it not
pass the --name argument, so it would end up with just the --pidfile.
The new security requirements from s-s-d made that in some cases, when
the usage was not secure, it would bail out with an error.
Closes: #923861
|
|
These fields have been there mostly to showcase their usage, which
should instead be dynamically added at build-time, and as a reminder
that this should be done.
But there is already a bug filed as a reminder (see #511160), they
are also now documented in the deb-src-control(5) man page, they are
actually wrong in many cases when present on a downstream, and they
generate nagging lintian warnings to boot.
|
|
|
|
This new public module centralizes the handling of debian/source/format
so that other projects can reuse it instead of having to reimplement it.
Prompted-by: Mattia Rizzolo <mattia@debian.org>
Ref: https://salsa.debian.org/debian/devscripts/merge_requests/63
|
|
|
|
|
|
The behavior we are breaking is implemented in the shared library, not
the standalone binary, which affects other frontends, such as aptitude.
The affected behavior was introduced in apt 1.4~beta1, and the last ABI
change of libapt-pkg was in apt 1.1~exp9, so earlier libapt-pkg*
versions are not affected.
Analysis-by: Sven Joachim <svenjoac@gmx.de>
Closes: #909959
|
|
apt was expecting duplicate entries, and was producing confusing
warnings. We add a versioned Breaks to avoid confusing users.
|
|
|
|
The debootstrap package in Debian stable (stretch) has been fixed so we
can rerevert this change.
|
|
The Dpkg::Source::Package::V2 module will try to use sensible-editor if
it's present. Add the Suggests to make the relationship explicit.
Warned-by: lintian
|
|
Thanks-to: Niels Thykier <niels@thykier.net> (for typo in dpkg(1))
|
|
The Alioth site has shut down, let's move to the new dpkg.org hosting.
|
|
Fixes: commit 114121d649b57bea9236b807d7c57bc918c67723
Thanks-to: Sven Joachim <svenjoac@gmx.de>
|
|
This is part of the ncurses 6 transition, and the previous packages
are now transitional, let's use the current one and leave the old for
backwards compatibility.
|
|
The functional test suite under t-func/ uses these compression tools
to unpack a test .deb package. This has not been a problem up-to-now
because dpkg-dev, which gets installed on the build systems, depends
also on these packages. But the dependencies are really missing.
Prompted-by: Bálint Réczey <balint.reczey@canonical.com>
|
|
This reverts commit ba52826a3429b9037e1876ba63c51afb211f151d.
R³ is supported fully now by debhelper, so we can use it when building
ourselves.
|
|
|
|
Older dgit versions assumed that Dpkg::Compression::Process was
available, via implicit import from Dpkg::Source::Package.
Reported-by: Ian Jackson <ijackson@chiark.greenend.org.uk>
|
|
We want to make debhelper users that will use R³ support not having to
add a versioned Depends on the relevant dpkg-dev package.
|
|
If we use a dpkg that fully supports the field, with a debhelper that
does not, this will produce broken builds. Disable for now, until such
debhelper has been released, and then we'll need to add some kind of
relationship to avoid the problem.
|
|
That package is using private modules with no API guarantees, and broke
due to recent changes in 1.19.0.
Closes: #878919
|
|
This makes it possible to use the pkexec tool to execute u-a from GUI
programs.
We need to bump the gettext build-dependency to at least 0.19.7 which is
the first version introducing ITS support and GETTEXTDATADIR variable
support (GETTEXTDATADIRS support was added in 0.19.8).
Proposed-by: Boyuan Yang <073plan@gmail.com>
|
|
The Dpkg::Source::Package::V3 source formats can use them.
Closes: #877688
|
|
|
|
|
|
|
|
|
|
We can also remove the versioned dpkg-dev Build-Depends now that we
do not use any build profile anymore.
|
|
The code supports optionally calling this program. Listing it in the
Suggests will also help people that are looking for a maintained and
supported .deb signing tooling.
Prompted-by: Stuart Prescott <stuart@debian.org>
|
|
Even though this was a correct change, it currently breaks debootstrap in
stable (jessie), testing (stretch) and unstable, which does not support
arch-qualified dependencies.
We'll have to live with the Multi-Arch hinter being unhappy for now.
|
|
The perl policy states that lib*-perl packages have to depend on the
perl package. Depending on a perl-modules packages is currently not
possible. And dh_perl will not emit a perl:any dependency for us, so
the current best option is to just hardcode it ourselves.
Reported-by: Sven Joachim <svenjoac@gmx.de>
|
|
Because tying the modules to a specific interpreter is wrong.
|
|
|
|
This will allow other projects to use the same whitelist as dpkg does.
Requested-by: Johannes Schauer <josch@debian.org>
|
|
Commit 9d7ba99cc3ff84fc553ed39da9d2e4f4008d35b6 renamed the
triplettable file to tupletable and added the necessary dpkg
dependency to libdpkg-perl, but did not forbid the other now broken
combination of old libdpkg-perl and new dpkg.
Closes: #843429
Signed-off-by: Guillem Jover <guillem@debian.org>
|
|
The scripts use modules not present in perl-base, so to be entirely
correct we require the perl package, even though it gets pulled anyway
by the libdpkg-perl package.
Warned-by: dpkg-gencontrol
|
|
This allows to detangle the libc used from the calling conventions.
|
|
debhelper assigns special meaning by considering it the main package.
|
|
|
|
There are very few packages now using the hardcoded paths, with bugs and
patches. It's about time to clean this up.
|
|
All these are satisfied in Debian oldstable.
|
|
|
|
The Dpkg::Gettext module makes conditional usage of this module, so
we should Recommend it to give the user a chance to get it installed.
|
|
We are using --porefs wrap option which was introduced in 0.43.
|
|
Clamp the mtime of the source files when building the tarball to the
first defined value from the source_date option, SOURCE_DATE_EPOCH
environment variable or time().
This makes sure the generated source tarballs always contain the same
mtime for files or directories that have been modified during the build.
|
|
This will guarantee that no file in binary packages has an mtime later
than the specified time. Which will be required to make binary packages
reproducible.
The option was officially added in GNU tar 1.29, but in Debian it was
introduced as a vendor patch in 1.28, so on Debian we depend on the
latter instead of the former version.
Closes: #759886
|
|
|