The following commit has been merged in the master branch:
commit 61e353f65b60cc5d58fc883ac150f90de244dfd2
Author: Guillem Jover guil...@debian.org
Date: Fri Nov 6 22:25:11 2009 +0100
libdpkg: Namespace include guards with LIBDPKG instead of DPKG
diff --git a/lib/dpkg/buffer.h
The following commit has been merged in the master branch:
commit f5c904d58627ec1422173a596f2a2ef8584485c2
Author: Guillem Jover guil...@debian.org
Date: Fri Nov 6 23:03:01 2009 +0100
libdpkg: Rename and namespace subproc related functions
diff --git a/dpkg-deb/build.c b/dpkg-deb/build.c
The following commit has been merged in the master branch:
commit d5388296797726e34abe13031dcbdd302319b908
Author: Guillem Jover guil...@debian.org
Date: Sat Nov 7 00:07:02 2009 +0100
libdpkg: Split process waiting from subproc_wait_check into subproc_wait
diff --git a/lib/dpkg/subproc.c
The following commit has been merged in the master branch:
commit a720b6aec5ab71211067cfd0250b57efa0aa18f9
Author: Guillem Jover guil...@debian.org
Date: Sat Nov 7 00:30:16 2009 +0100
Use subproc_wait instead of directly calling waitpid
diff --git a/dselect/method.cc b/dselect/method.cc
The following commit has been merged in the master branch:
commit 0814d1b1eb42c08809d31980d9a7f2b921c713e5
Author: Guillem Jover guil...@debian.org
Date: Sat Nov 7 14:04:41 2009 +0100
Use find with -L instead of deprecated -follow
diff --git a/dselect/methods/disk/install
]] Joey Hess
| I support changing 3.0 (git) to use a bundle. Besides closing the bugs
| mentioned in this thread, a bundle consists of a simple header + a
| standard git pack. Since git packs are used as a wire format, this
| provides better assurance that future versions of git will retain
|
Hi,
A few quick thoughts and questions about conffiledb_automerge:
Sean Finney wrote:
This functions by performing a 3-way diff using the new conffile, the
currently-installed conffile, and the pristine version of the conffile
shipped in the currently installed package (in the conffile
]] Goswin von Brederlow
| Tollef Fog Heen tfh...@err.no writes:
|
| ]] Goswin von Brederlow
|
| | Remove 3.0 (qit) format completly. A full history bloats the source
| | and a cut down history has no advantage over a plain source.
|
| I think this would be a mistake. I'm increasingly
Tollef Fog Heen tfh...@err.no writes:
]] Goswin von Brederlow
| My feeling is that 3.0 (git) format adds bloat to the source packages
| that hardly anyone ever uses, makes it that much harder for any
| non-git user to edit the source and is of little extra value when the
| maintainers git
Goswin von Brederlow wrote:
Weigh that against the bandwidth spend for mirrors and for people that
do not need or want the history and the extra cost in terms of needing
more CD/DVD images to contain a source snapshot. Also the cost for
snapshot.debian.org having to have the extra bloat for
Jonathan Nieder wrote:
Until then, let me compare the analogous moment in the current
configure phase. (deferred_configure_conffile(), in configure.c)
1. if .dpkg-new file is missing, skip to the next conffile
2. compute .dpkg-new file hash
3. prompt the user for action
Free filename buffers when returning early from
deferred_configure_conffile().
While we’re at it, speed up the initialization of the buffer
cdr2. The current code scans cdr for its end unnecessarily,
twice, and then uses that length to calculate a buffer size about
40 bytes too large.
Also let
When backing up the .dpkg-new file, make a hardlink backup and
then fall through to the deleting .dpkg-new because it is not to
be used case. This is messier than a true rename, but it
creates some flexibility by allowing us to perform some
operations between the two steps in the future.
Split up the rename(conffile.dpkg-new, conffile) into
link(conffile.dpkg-new, conffile);
unlink(conffile.dpkg-new);
in preparation for an operation in between. Since link() does
not actually atomically replace its target, we must simulate it
by linking to a temporary file and
If deferred_configure_conffile() is interrupted after the
.dpkg-new file is removed but before its hash is recorded as the
distributed conffile hash, that hash is lost forever, and if it
is different from the previous one then the conffile will appear
to have been modified by the user the next
Tollef Fog Heen tfh...@err.no writes:
]] Joey Hess
| I support changing 3.0 (git) to use a bundle. Besides closing the bugs
| mentioned in this thread, a bundle consists of a simple header + a
| standard git pack. Since git packs are used as a wire format, this
| provides better assurance
Tollef Fog Heen tfh...@err.no writes:
Whether git.d.o is up or not does not help me if I'm at the end of tiny
link. Also, if we go with my claim that source (in the «preferred form
for modification» sense) includes history, it means you need to ship the
history somehow. Saying «it's on
There’s a typo in this one:
Jonathan Nieder wrote:
--- a/src/configure.c
+++ b/src/configure.c
@@ -220,11 +216,6 @@ deferred_configure_conffile(struct pkginfo *pkg, struct
conffile *conff)
if (rename(cdr2.buf, cdr.buf))
ohshite(_(unable to install
Currently, dpkg lets interested packages know when saving a
backup (.dpkg-dist file) for an updated conffile that is not
going to used.
This is, strictly speaking, harmless, since file triggers were
never a guarantee that a file was actually modified. But on one
hand, the triggered action can be
Package: dpkg
Version: 1.14.25
Severity: normal
Hello,
tspc doesn't work anymore. So when I tried to remove it, dpkg delivered
the following:
dpkg -r tspc
(Reading database ... 175350 files and directories currently installed.)
Removing tspc ...
Shutting down IPv6 tunnel: invoke-rc.d:
20 matches
Mail list logo
