| Commit message (Collapse) | Author | Age |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
This patch provides code and documentation changes to accomodate
Debian's '/etc/mailname'. The patch includes related modifications,
for example not to add a domain to the hostname when 'defaultdomain'
is empty, as it might reasonably be on a Debian system. The
presently also disables tests related to 'me'. It is not clear how
the '/etc/mailname' behavior is tested best.
Author: Felix Lechner <felix.lechner@lease-up.com>
Bug-Debian: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=504184
Last-Update: 2017-08-31
Gbp-Pq: Name 0005-Provide-for-etc-mailname.patch
|
| |
|
|
|
|
|
|
|
|
| |
This patch allows unqualified domain names for senders and recipients,
because are required in some use cases (Debian bug #504184). The
associated post-build tests were also disabled.
Author: Felix Lechner <felix.lechner@lease-up.com>
Bug-Debian: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=504184
Last-Update: 2017-09-07
Gbp-Pq: Name 0004-Allow-unqualified-sender-address.patch
|
| |
|
|
|
|
|
|
| |
Sort files when using mergelib to create libnullmailer.a, to get
reproducible build
Author: Alexis Bienvenüe <pado@passoire.fr>
Gbp-Pq: Name 0004-Sort-files-in-archive-reproducible-builds.patch
|
| |
|
|
|
|
|
|
|
|
|
| |
Testing the qmqp protocol fails because the testing server cannot
acquire the lock. This is solved by waiting for the previous
server to exit gracefully and clear the lock. Similarly, wait
before removing temporary directories.
Forwarded: no
Last-Update: 2017-10-07
Gbp-Pq: Name 0001-Remove-race-conditions-from-tests.patch
|
| |\
| |
| |
| |
| |
| | |
First breakwater merge.
[git-debrebase anchor: declare upstream]
|
| | |
| |
| |
| |
| |
| | |
Delete debian/patches, as part of converting to git-debrebase format.
[git-debrebase convert-from-gbp: drop patches from tree]
|
| | |
| |
| |
| | |
Not sure if the patch was corrupted, or there is some bug with base64 encoding
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| | |
| |
| |
| | |
This is pretty much untested, hence the upload to experimental
|
| | | |
|
| |\ \ |
|
| | | |
| | |
| | |
| | |
| | | |
Re-enabled patch avoid race conditions in tests
Changed watch file to Github
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | | |
Patch from Benda Xu mid:87wp0io6r7.fsf@gentoo.org
Apparently the --daemon switch is uneeded, and logging can be provided
by the (essential) program logger.
|
| |/ / |
|
| | | |
|
| | | |
|
| | | |
|
| | |
| |
| |
| | |
IMHO most nullmailer users don't need it.
|
| | | |
|
| | |
| |
| |
| |
| | |
This is a bit experimental; it's not clear it will work without also
forcing logging via syslog.
|
| | |
| |
| |
| | |
depend on systemd-sysv
|
| | | |
|
| | |
| |
| |
| | |
apparently this is what dh_installchangelogs wants?
|
| | |
| |
| |
| | |
just documentation changes.
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| |\| |
|
| | | |
|
| | |\
| | |
| | |
| | |
| | |
| | | |
Version 2.1
Leave deleted files deleted.
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
command line arguments to 'dsn' are not processed correctly. The
(void) pointer passed to lib/cli++/main.cc is cast to
(unsigned int) instead of (ulong). On 64-bit little endian
architectures such as amd64 this is not readily apparent as long
as the higher bits were initialized, but it causes grave errors
on big-endian architectures such as Debian's powerpc or s390x.
This patch processes the timestamps given on the command line
correctly.
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | |
| | | |
If no address could be used to bind the source address, the resulting
error code from tcpconnect would be incorrect (temporary gethostbyname
failure). This rewrites it to a more sensible failure to bind source.
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
