devscripts (2.23.6)
[PTS] [DDPO]
NEW: VCS has unreleased changes: 2.23.7 > 2.23.6
- Git: https://salsa.debian.org/debian/devscripts.git
-
- Branch: master
- Path: debian/changelog
- Repo size: 36511744
- Browser: https://salsa.debian.org/debian/devscripts
- Last scan: 2023-09-29 10:44:58+00
- Next scan: 2023-10-05 12:03:00+00
- Merge requests: 20
- CI pipeline status: success
- Debian changelog in Git:
devscripts (2.23.7) UNRELEASED; urgency=medium
* Team upload
[ Nicholas D Steeves ]
* debsign: Override the gpg2 default to update trustdb at what may be an
inconvenient time, because an up-to-date trustdb is not needed for this
operation (Closes: Bug#1050387)
[ Niels Thykier ]
* wrap-and-sort:
+ Add a --no-keep-first option
+ Make --keep-first the default
+ Support wrap-and-sort'ing d/control.in template files
+ Code better with deb822 parse errors
* README:
+ Remove obsolete section about setuid wrappers
+ Remove `fakeroot` from `debclean` and `debuild` dependencies
* debuild.pl: Clarity that the `fakeroot` default comes from dpkg-buildpackage
* dpkg-genbuilddeps.sh: Remove explicit `-rfakeroot` from error message
* d/control: Remove `fakeroot` from Depends. (Closes: #1041150)
[ g0t mi1k ]
* salsa:
+ Split edit_project -> edit_project_multipart
+ Only set avatar if told too
+ Check if project is public for remote avatar
+ Set user-agent when remotely checking md5
+ List members including inherited
+ Improve some salsa messages
+ Drop --request-acc, duplicate of --request-access
[ Sean Whitton ]
* debchange: For --lts uploads, use urgency=high.
-- Yadd <yadd@debian.org> Tue, 19 Sep 2023 05:43:52 +0400
- This branch is 31 commits ahead of tag v2.23.6
- Git log:
commit ffc6f2ecade6fe9339cf7fde647203f1c22ca68c
Author: Sean Whitton <spwhitton@spwhitton.name>
Date: Fri Sep 29 11:28:34 2023 +0100
debchange: add FIXME
commit 7757fec7b4fb1ac06cb29e13ad911aee04578ea5
Author: Sean Whitton <spwhitton@spwhitton.name>
Date: Fri Sep 29 11:27:31 2023 +0100
debchange: For --lts uploads, use urgency=high
commit 6abd4471233609e7c8ae00f7edfa004b324cdfde
Author: Yadd <yadd@debian.org>
Date: Tue Sep 19 05:53:48 2023 +0400
Update d/ch
commit 03deab1b32c3b0e4367bb8ac7885e9292f8c3be9
Merge: fb1bdb15 7768ec81
Author: Yadd <yadd@debian.org>
Date: Tue Sep 19 05:40:25 2023 +0400
Merge remote-tracking branch 'tmp/_check_config'
commit fb1bdb15fedd85586177509161e016d1cb3a76e0
Merge: a62c1d57 e5a431f8
Author: Yadd <yadd@debian.org>
Date: Tue Sep 19 05:40:04 2023 +0400
Merge remote-tracking branch 'tmp/consistency'
commit 7768ec81eb41060620c3ed42b58cac37e981859b
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Tue Sep 5 17:53:10 2023 +0100
Explain boolean flags
commit 02f161590b988b54bea603d08283aad962dc7fb5
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Tue Sep 5 17:40:04 2023 +0100
Use _check_config function
commit 076301d6ff2b5f544bd9cae0d20514b51be7ff2e
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Tue Sep 5 17:38:41 2023 +0100
Fix comparing for/if statement
Has always been returning true
commit cb48648e2564f5f1e7b06c0815af808f9ad30973
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Tue Sep 5 17:06:22 2023 +0100
Remove duplicate flag: SALSA_ENABLE_REQUEST_ACC
SALSA_ENABLE_REQUEST_ACC is the same as SALSA_REQUEST_ACCESS
SALSA_REQUEST_ACCESS has been around longer, thus opt'ing to keep that
commit 71aeb1cf99b605ed7fd9373a805034990574e779
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Tue Sep 5 17:04:37 2023 +0100
request_access_enabled listed twice
commit e5a431f81336a12a1d1da9dca95787c29d5c8dda
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Sep 4 11:04:15 2023 +0100
Use symbols over grammar like other statements
commit c48ab5be4b6ccd565cb0379378ef21bb4a134828
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Sep 4 11:03:40 2023 +0100
Define the context of "private"
commit 75431f0689a422a553dc279c1bf9fcc7972f3274
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Sep 4 11:01:51 2023 +0100
Drop : - no arguments to follow
commit bc3b92be711e991da905e9660532bc10a57c978f
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Jan 23 07:23:17 2023 +0000
Fix consistency with what variable being used
commit 30be4e8948a15860380716f04dc1d63c30603d5c
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Jan 23 07:22:48 2023 +0000
Bug fix when project doesn't exists
name may not yet have been defined (if project doesn't exists)
commit a62c1d5734cf4340533a72bba4dfca4de6fc2ce3
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Sep 4 10:31:22 2023 +0100
List members including inherited
group_member -> all_group_members
commit 8584d45be7c5ec8ade5b91f274892b36eb2fc3ad
Author: Niels Thykier <niels@thykier.net>
Date: Sat Sep 16 18:00:57 2023 +0000
wrap-and-sort: Code better with deb822 parse errors
Previously, we would dump a stack trace and terminate. Now, we print
a one-line warning and ignore the file.
Signed-off-by: Niels Thykier <niels@thykier.net>
commit 81bb2e5ac1f66608bb2e7c132034b3a401f70d26
Author: Niels Thykier <niels@thykier.net>
Date: Sat Sep 16 17:02:51 2023 +0000
wrap-and-sort: Support wrap-and-sort'ing d/control.in template files
Signed-off-by: Niels Thykier <niels@thykier.net>
commit 50ef3626476be59ff05c39a1c20aa652ec7d279f
Author: Niels Thykier <niels@thykier.net>
Date: Sat Sep 16 13:13:20 2023 +0000
d/control: Remove `fakeroot` from Depends.
Nothing in devscripts actively uses fakeroot. The place where
dependency relevant is when interacting with dpkg-buildpackage. The
`dpkg-dev` provides the relevant optional dependency, so there is no
point in us having it as well.
Closes: #1041150
Signed-off-by: Niels Thykier <niels@thykier.net>
commit e7e8d0930cdf455f8b9fa39810b2d2081c25f5a2
Author: Niels Thykier <niels@thykier.net>
Date: Sat Sep 16 13:09:24 2023 +0000
README: Remove `fakeroot` from `debclean` and `debuild` dependencies
They do not actually use `fakeroot` themselves anymore. They delegate
to `dpkg-buildpackage` and have it manage the default except when
configured to provide a different default to `dpkg-buildpackage`.
Accordingly, it is a bit of a stretch to claim that these tool
"Depend" on `fakeroot`. Especially when combined with the fact that
half the archive no longer require fakeroot or real root.
Signed-off-by: Niels Thykier <niels@thykier.net>
commit a11a7433cee12dd519b474b654db52eb891efdde
Author: Niels Thykier <niels@thykier.net>
Date: Sat Sep 16 13:07:00 2023 +0000
dpkg-genbuilddeps.sh: Remove explicit `-rfakeroot` from error message
The command in question invokes `dpkg-buildpackage` and these days
`-rfakeroot` is the default for packages that need `-r`.
Signed-off-by: Niels Thykier <niels@thykier.net>
commit e760102fc3ba86bf403d4bb5310ea0c13c34e586
Author: Niels Thykier <niels@thykier.net>
Date: Sat Sep 16 13:06:05 2023 +0000
debuild.pl: Clarity that the `fakeroot` default comes from dpkg-buildpackage
No where in `debuild.pl` do we state `fakeroot` explicitly anymore.
We almost certainly did historically but these days, dpkg-buildpackage
manages this.
Signed-off-by: Niels Thykier <niels@thykier.net>
commit 82dd47a45d95074d5eb0d26789c8f1019f4d00cd
Author: Niels Thykier <niels@thykier.net>
Date: Sat Sep 16 12:59:39 2023 +0000
README: Remove obsolete section about setuid wrappers
The `debpkg` program was removed in 2.23.5 and so anything related to
that command is obsolete on account of the command no longer existing.
The remaining wrapper `debuild` still exists. However, the section
focuses heavily on using `debuild` as a setuid wrapper or with
fakeroot, which is not really relevant for `debuild` anymore. We no
longer recommend using setuid with `debuild` and the `fakeroot` usage
default is in fact moved to `dpkg-buildpackage` (and even then, by now
over 50% of the archive do not even need fakeroot anymore).
With that in mind, I opted for just purging the entire section rather
than trying to salvage a line or two and try to retrofit that
somewhere else.
Signed-off-by: Niels Thykier <niels@thykier.net>
commit f3e2ccef1e6ea9b1853ce89fd1cf4bbaa10987a7
Author: Niels Thykier <niels@thykier.net>
Date: Thu Aug 17 18:55:21 2023 +0000
wrap-and-sort: Make --keep-first the default
The previous behaviour was "unsafe-by-default" because it could break
packages using debhelper by changing semantics of `debian/install` (by
making it apply to a different package). This change makes
`--keep-first` the default, which is a much "safer" default as it
cannot cause semantic changes in any wide-spread packaging tool.
Signed-off-by: Niels Thykier <niels@thykier.net>
commit 962a1a31ccd0d00ea473a4111388e897387b9344
Author: Niels Thykier <niels@thykier.net>
Date: Thu Aug 17 18:53:54 2023 +0000
wrap-and-sort: Add a --no-keep-first option
Signed-off-by: Niels Thykier <niels@thykier.net>
commit c20be072758a22bfa2629e51c0a7a9dde114d9cf
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Sep 4 10:57:07 2023 +0100
Set user-agent when remotely checking md5
Matches what's in: ./scripts/debsnap
commit ce6eda266f0891478f0beecdddb157b16f78b0d7
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Sep 4 10:52:33 2023 +0100
Check if project is public for remote avatar
Project/repo has to be pubic since GitLab 11.5.1 (November 2018)
https://gitlab.com/gitlab-org/gitlab/-/issues/25498
commit 618003ed3015b11fc223258b9a36f7f36c5d67bd
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Sep 4 10:49:21 2023 +0100
Only set avatar if told too
commit 94b60b7a37a3b174f4edc2ab175c8e2df771994f
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Jan 23 07:22:38 2023 +0000
More verbose actions
commit 06e87cf88a40b95cf4ae0d87655a3c62ffebfa48
Author: g0t mi1k <have.you.g0tmi1k@gmail.com>
Date: Mon Jan 23 06:54:37 2023 +0000
Split edit_project -> edit_project_multipart
!311 wasn't quite ready yet to be merged until upstream API ((libgitlab-api-v4-perl) could handle the new type of request.
Due to how it generates code, was easier to split into different functions rather than use a single
commit f8bccd1b4c9feb0bd71b48a5668837174f1dfb1b
Author: Nicholas D Steeves <sten@debian.org>
Date: Fri Aug 25 21:19:11 2023 -0400
debsign: Override the gpg2 default to update trustdb at what may be…
an inconvenient time, because an up-to-date trustdb is not needed for
this operation (Closes: Bug#1050387)