martin f krafft wrote:
Previously, the CWD was just left at the top-level, right?
Therefore, I assume that the commands all already took precautions
to ensure that they operate on the right subdir, e.g. by setting
GIT_DIR.
No, mr chdirs before running lots of stuff. And doesn't chdir before
Package: alpine
Version: 2.11+dfsg1-3
Severity: normal
A user reported that alpine had suddenly started saying things about
rsh to IMAP server, and prompting for a password but not accepting any
input... Rather than opening her inbox.
Investigation showed that alpine was running:
26465 pts/9
Package: offlineimap
Version: 6.5.5-0.1
Severity: normal
FOLDER FILTERING AND NAMETRANS
OfflineImap offers flexible (and complex) ways of filtering
andtransformingfoldernames.Pleasesee the
docs/dev-docs-src/folderfilters.rst document about details how
Package: rss2email
Version: 1:3.8-1
Severity: normal
joey@kite:~r2e add electoral-vote http://www.electoral-vote.com/index.rss
joey@kite:~r2e run electoral-vote
processing error: application/octet-stream is not an XML media type:
electoral-vote (http://www.electoral-vote.com/index.rss -
Package: rss2email
Version: 1:3.8-1
Severity: normal
joey@kite:~/.configr2e add test http://test:t...@www.kitenet.net/index.rss
joey@kite:~/.configr2e run test
processing error: 'str' does not support the buffer interface: test
(http://test:t...@www.kitenet.net/index.rss - j...@kitenet.net)
Here's another one.
processing error: text/html; charset=UTF-8 is not an XML media type: feed-76
(http://www.lightspeedmagazine.com/rss-2/ - j...@kitenet.net)
--
see shy jo
signature.asc
Description: Digital signature
W. Trevor King wrote:
On Tue, Sep 09, 2014 at 10:35:50AM -0400, Joey Hess wrote:
joey@kite:~r2e add electoral-vote http://www.electoral-vote.com/index.rss
joey@kite:~r2e run electoral-vote
processing error: application/octet-stream is not an XML media type:
electoral-vote (http
Package: rss2email
Version: 1:3.8-1
Severity: normal
Each time it is used to update list of feeds, r2e rewrites
rss2email.cfg, and neglects to preserve its file permissions when doing
so.
So, any rss2email user whose home directory is world readable, but who
has locked down the permissions of
W. Trevor King wrote:
On Tue, Sep 09, 2014 at 10:45:17AM -0400, Joey Hess wrote:
joey@kite:~/.configr2e add test http://test:t...@www.kitenet.net/index.rss
joey@kite:~/.configr2e run test
processing error: 'str' does not support the buffer interface: test
(http://test:t
Package: spamassassin
Version:
I am getting cron mails about spamassassin. I have configured CRON=1.
As far as I can tell from the mail, nothing is broken, it's just
random output.
Please avoid sending cron mails unless something went wrong.
- Forwarded message from Cron Daemon
martin f krafft wrote:
Make mr try to chdir() to the dirname() of any .mrconfig files it's
supposed to load. If the chdir() is successful, then the filename is
truncated to basename(). If the chdir() is unsuccessful, then processing
happens as before.
This seems to chdir and then not chdir
martin f krafft wrote:
So let's say ~/some/path/.mrconfig specifies
[subdir]
checkout = …
chain = true
and ~/some/path/subdir/.mrconfig specifies
[another]
checkout = …
then, with the status quo, running mr in ~/some/path will work on
two repos, as it should.
This is
martin f krafft wrote:
I think this is an inconsistency. When run in bottom-up fashion,
then either ~/.mrconfig should also be ignored, or all parent config
files need to be included.
If I had a vote, I'd say that all parent config files should be
included.
This would mean that if you
martin f krafft wrote:
The loadconfig() function does something prefixed by a comment:
# copy in defaults from first parent
in which it walks up the tree
You make it sound like it's walking the filesystem tree, but it's not,
it's messing around in its data structures from the configs
I am trying to give team members a way to add local aliases etc.
without having to modify .mrconfig
Isn't this what ~/.mrconfig's [DEFAULT] is for?
--
see shy jo
signature.asc
Description: Digital signature
martin f krafft wrote:
also sprach Joey Hess jo...@debian.org [2014-09-09 20:43 +0200]:
This seems to chdir and then not chdir back. How does it avoid
breaking existing setups?
I intentionally did it this way, because I could not see a way in
which this would actually matter, given
martin f krafft wrote:
So saying something like chain=pure is really the same, except it
would handle status/log/diff etc. as well, and implicitly. Maybe
there is also a better way to do it, e.g. chain=true and then
something like skip_all_but_checkout=true.
skip = ! lazy will work, although
Package: sshuttle
Version: 0.54-2
Severity: normal
Tags: security
With -H, sshuttle writes a ~/.sshuttle.hosts file on the remote ssh
server host. It seems to contain a lot of dns hostnames. Not just local
hostnames, but random websites visited via sshuttle, etc. The file is
left lying around
So, I see no problem with adding gnome-orca to task-xfce-desktop, given
those installation size numbers.
I sympathize with Yves in wanting to keep the xfce4-* packages
containing only actual upstream XFCE stuff; that's part of why we have
the task packages to add currently missing (and in some
There is going to be a limited amount of space in tasksel for blends,
given current debconf UI constraints.
I think that using popcon as a rough pass to select the blends makes
rather a lot of sense. The Debian Pure Blends effort has been around
for several releases and been publicised. The
Some things that seem to be missing from the cinnamon
task that are included in other desktop tasks:
* package management (eg synaptic)
* printer configuration? (system-config-printer)
* network-manager
I have not looked at cinnamon in detail, this is just looking at what's
in the other tasks
Package: obnam
Version: 1.8-1
Severity: normal
While running an obnam restore --to /tmp at root, I found I could not send
emails, since gpg was trying to talk to a socket there, and obnam had
made the directory mode 700.
sudo obnam --config /etc/obnam.conf.archive restore --to /tmp/
Package: virtualbox
Version: 4.3.14-dfsg-1
Severity: normal
The first error message in this strace is output by
/usr/lib/virtualbox/VBoxXPCOMIPCD, and points to an actual, understandable
and fixable problem:
23157 write(1, Wrong mode (42700) of '/tmp/.vbo..., 94) = 94
23157 exit_group(0)
Andrew Ayer wrote:
That sounds good to me. Would you have any objections to adding a
Depends: to a hypothetical strip-nondeterminism package that would
contain normalizer utilities for ar, gzip, jar, etc. files?
That sounds like a good approach.
--
see shy jo
signature.asc
Description:
Jérémy Bobbio wrote:
Currently, static libraries shipped in Debian package capture the time
when the package is built. As part of the “reproducible builds”
project [1], it would be great to have static libriaries normalized.
The attached patch will make `dh_strip` replace non-deterministic
+1
Do you have a plan to get packages not using dh or cdbs to use this new
command?
I can think of 2 ways.. one is to add it to some existing command,
perhaps renaming the command. I thought maybe dh_fixperms could be
renamed to dh_fixmetadata. But, I think some will use dh_fixperms -X to
Jérémy Bobbio wrote:
Joey Hess:
Do you have a plan to get packages not using dh or cdbs to use this new
command?
Its heart is a single find+xargs+touch command. I had in mind that
packages not using dh or cdbs could have their own way on how to make
the mtimes deterministic. Possibly
Jérémy Bobbio wrote:
Andrew Ayer has been working on a `dh_strip_nondeterminism` helper:
http://anonscm.debian.org/cgit/reproducible/strip-nondeterminism.git/
We can move that chunk of code to it, alongside normalizers for gzip,
zip, and jar files.
But we were kinda hoping to get that
Christian Kastner wrote:
The attached patch extends dh_installdirs, dh_install, and dh_link such
that in v10 mode, the literal string :multiarch: is substituted with the
appropriate multiarch tuple during their operation.
I think that Ubuntu had a patch for this or something similar
earlier,
Antoine Beaupré wrote:
Please provide the output of monkeysign -d, feel free to encrypt the
result to me if you are worried about leaking information.
joey@darkstar:~monkeysign -d
Usage: monkeysign [options] keyid
wrong number of arguments, use -h for full help
But it's monkeysign that is
Package: monkeysign
Version: 1.1
Severity: important
joey@darkstar:~monkeyscan
Traceback (most recent call last):
File /usr/bin/monkeyscan, line 41, in module
u.main()
File /usr/lib/python2.7/dist-packages/monkeysign/gtkui.py, line 60, in main
self.window = MonkeysignScan()
File
Nils Asmussen wrote:
The problem seems to be that
git --git-dir=/media/nils/USB/annex cat-file --batch
should be closed after completing all transfers, but is not closed. This
process is started by the git-annex assistant running
in /home/nils/sync.
I have reproduced this problem, and you're
Yaroslav Halchenko wrote:
ATM
$ git annex importfeed --help
git-annex: unrecognized option `--help'
Usage: git-annex importfeed URL ... [option ...]
--template=FORMAT template for filenames
--relaxed skip size check
To see additional options common to all commands,
Nils Asmussen wrote:
the setup is as follows: I have two computers. Using the git-annex webapp, I
set up a usb dongle to sneakernet files between them.
Whenever I connect the usb dongle to one of the computers, the files are
transferred as expected.
But: When I then try to unmount the usb
Yaroslav Halchenko wrote:
even though claims to be unrecognized, spits out help:
Well, so will any unrecognized option.
While it might be nice to make --help a known option that prints out
help for the current command, it would be unncessarily complicating to
add it to the current
Yaroslav Halchenko wrote:
Original motivation was expressed here:
https://github.com/data-git/datagit/issues/1
Through-away limited-view (i.e. some keys not present) could be useful
e.g. to test for correct operation of a given data analysis pipeline
given a subset of files (which were
Yaroslav Halchenko wrote:
yeap -- sounds like a good analysis/concerns, hinting once again to make
it optional thus maintaining original integrity and solely allowing
linked storage as user's choice -- for the throw-away clones (per my use
case) above concerns unlikely be relevant.
Another
Jonas Meier wrote:
Version: 1.20120628.1
That version is not the current version. I just tested it and it works
with the current version at least.
--
see shy jo
signature.asc
Description: Digital signature
. Caso contrário, pode causar múltiplas
instâncias do mesmo texto a ser adicionado aos scripts do mantenedor.
VEJA TAMBÉM
debhelper(7)
Este programa é parte do debhelper.
AUTOR
Joey Hess jo...@debian.org
Tradução
Américo Monteiro
Se encontrar algum erro na
Package: xmonad
Version: 0.11-8
Severity: normal
When I press the key combo to restart xmonad, it is currently failing,
displaying a xmessage that it's not in PATH. Strace shows:
22164 execve(/bin/sh, [/bin/sh, -c, if type xmonad; then xmonad
--re...], [/* 58 vars */]) = 0
22164 write(1, xmonad
Package: debootstrap
Version: 1.0.60
Severity: normal
After a foreign deboostrap, the second stage failed:
dpkg: regarding .../systemd-sysv_208-7_armhf.deb containing systemd-sysv,
pre-dependency problem:
systemd-sysv pre-depends on systemd
systemd is unpacked, but has never been configured.
Guillem Jover wrote:
Hmm, I had already pondered about this when I saw the announcement. So,
while I agree such option would make life easier for people that want
to do “almost-source-only uploads” right now (because TBH the other
alternatives are very annoying), on first thought this seemed
Raphael Hertzog wrote:
Funnily this can already be done with “dpkg-buildpackage
--changes-option=-S” but it will generate a _arch.changes file
that list only the source package (that's because the filename
is decided by dpkg-buildpackage while the content comes from
dpkg-genchanges).
Hmm, so
Guillem Jover wrote:
But, this has a nice side-effect, you can also do something like:
$ dpkg-buildpackage --changes-option=-g
which will perform a normal full build, but filter the generated
.changes file to only include source+arch-indep packages. Which I
think is what you'd be
Package: sunxi-tools
Version: 1.2-1
Severity: normal
The README should be included in the package.
--
see shy jo
signature.asc
Description: Digital signature
Package: unattended-upgrades
Version: 0.79.5
Severity: normal
Tags: security
/var/log/unattended-upgrades/ is readable by all, so when this package is
run on a multi-user system, non-admin users can trawl the upgrade logs
for interesting information.
I don't know what they might find.. Which is
Source: haskell-dav
Severity: normal
Tags: patch
Please apply the two attached patches, which add stuff git-annex needs
to use DAV more efficiently.
--
see shy jo
From e9cc0657853024d32eb2c5a62d8e8a44882fe018 Mon Sep 17 00:00:00 2001
From: Joey Hess j...@kitenet.net
Date: Wed, 6 Aug 2014 16:16
be left as-is, while
providing a withContentM that uses Network.HTTP.Client.withResponse.
The caller can then stream the response (eg out to a file) in constant memory.
git patch attached.
--
see shy jo
From 550b24904a8ad3e83c42aff4d1a8af9f392955fb Mon Sep 17 00:00:00 2001
From: Joey Hess j
Package: dgit
Version: 0.21
Severity: wishlist
Now that the archive finally supports .deb-less uploads, dgit should be
able to take it a step further, to .tar-less push.
I envision something like: dgit push --via somehost
That would push the git repository as usual, then go to somehost and
dgit
Ian Jackson wrote:
I envision something like: dgit push --via somehost
I don't understand what the somehost is for.
Why wouldn't dgit just generate the necessary source package files in
the usual way ?
Perhaps your --via is simply the build-host for rpush.
The use case is that it's
Ansgar pointed out that this could be simplified some if there were a
way to reproducible build the same *source* package from info in git.
The remote machine then does not need to cobble together the dsc and
changes, but can just be sent the signed dsc and changes from the build
machine, pair
Package: dpkg-dev
Version: 1.17.10
Severity: wishlist
Now that the archive supports source-only uploads, with the exception
that arch:all debs have to still be uploaded (as no autobuilder is
responsible for them), dpkg-genchanges -S is useful, but not quite
right.
It would be good to have an
I got the -passfile to work.
The file needs to exit, or alpine won't try to use it even with the
option given. An empty file works here though, no crash. Alpine writes
the password to the file and uses it next time.
--
see shy jo
signature.asc
Description: Digital signature
Package: apt
Version: 1.0.5
Severity: wishlist
Many of apt's users are human, and thus prone to making mistakes. A common
mistake seems to be running apt-get dist-upgrade and not noticing that it wants
to remove your backup system and/or terminal emulator and/or steam. Or not
noticing that it
I suspect there are multiple failure modes here. Two I have seen:
1. Progress bar stalls with bottom bar displaying network error, as
described earlier.
2. Posting immediately fails. Retrying also fails. Moving to a very fast
and good network and retrying then also fails. In this case I
Michael Prokop wrote:
It would be nice if there would be an option to skip certain
repositories. It's useful if you don't want to download very large
repositories or if a certain repository is known to cause problems.
I assume your're talking about repositories you've owned or starred?
--
Gabriel Niebler wrote:
The package would have to figure out which VCS to use, which should be
easy if there's only one, and set $VCS accordingly in etckeeper.conf
before using it.
This is problimatic because the VCS used would then be essentially
non-deteministic.
Instead, I think etckeeper
Package: xmonad
Version: 0.11-7+b2
Severity: normal
Tags: patch
xmonad floats mplayer by default. However, Debian has switched to
mplayer2, and so on upgrade, it will stop floating. This can be easily
fixed in the user's config file, but here is a patch that avoids
bothering the user with
Package: cpphs
Version: 1.18.4-1
Severity: normal
The man page is missing at least the --include option.
-- System Information:
Debian Release: jessie/sid
APT prefers unstable
APT policy: (500, 'unstable'), (500, 'testing')
Architecture: i386 (i686)
Kernel: Linux 3.13-1-686-pae (SMP w/2 CPU
Niko Tyni wrote:
There's a pending Perl policy change discussed in #750017 that requires
the handful of packages in the archive that install nonbinary Perl modules
into /usr/lib/perl5 (or, more generally, $Config{vendorarch} to add a
perlapi-* dependency. The attached mostly untested patch
Christoph Anton Mitterer wrote:
reopen 749795
I'm reopening this for now, even if the issue is solved from a technical
point of view (see below why).
AAICS, #749795 talked about bringing this to the security team's
attention, but they never seem to have been CCed.
So the security team may not
I finally fixed handling of non-ascii repo descriptions in version
5.20140529.
--
see shy jo
signature.asc
Description: Digital signature
martin f krafft wrote:
No, I am talking about the default symlink method. And it really
does not seem to be using git-checkout:
pull origin
ok
Well, this does a git fetch followed by a git merge. If that doesn't run
git-checkout, then I suppose git doesn't intend that hook to be run
martin f krafft wrote:
Indeed, I thought it was using plumbing to mock a checkout. But
you're right, git-merge does not call post-checkout either. The
question now is: should git-annex call a single hook to allow
touching up permissions in the worktree (one example)? If so,
I think it might
martin f krafft wrote:
Of course, but at least views seem to use checkout and somehow
I think it would be better to not expect git-annex users to have to
install two hooks or to understand why that would be necessary.
So I mainly wonder whether there's a singular hook interface for
every
martin f krafft wrote:
git-annex does a whole lot of magic using plumbing commands to
provide checkout-like behaviour, but it does not call the
post-checkout hook.
I assume you're talking about after git-annex sync in direct mode?
(Indirect mode just uses git checkout.)
--
see shy jo
Osamu Aoki wrote:
In the Bug#538978 report at #47, Piotr Ozarowski mentioned as:
python3 package in experimental (3.3.0-2) contains pybuild tool ...
https://bugs.debian.org/538978#47
This pybuild command is now a part of his new package dh-python available
in testing/unstable which also
Cameron Norman wrote:
diff --git a/autoscripts/postrm-init b/autoscripts/postrm-init
index 6f5bb09..b5cecb8 100644
--- a/autoscripts/postrm-init
+++ b/autoscripts/postrm-init
@@ -1,3 +1,5 @@
+update-rc.d #SCRIPT# disable /dev/null
+
if [ $1 = purge ] ; then
update-rc.d #SCRIPT#
Package: bash
Version: 4.3-7
Severity: normal
joey@wren:~/tmp/ghc-7.6.3q!
-bash: q!: command not found
malloc: /usr/homes/chet/src/bash/src/parse.y:2314: assertion botched
realloc: start and end chunk sizes differ
Aborting...Connection to kite closed.
This seems reproducible, at least in my
Package: docker.io, coreutils
Severity: minor
docker mounts stuff under a directory that normal users cannot read,
which makes df full of errors. Example below from a system with 4
running containers.
I feel this is a polish/integration issue that would be worth finding
some fix for. At least so
Package: docker.io
Version: 0.11.1~dfsg1-1
Severity: normal
Docker, at least as currently available in Debian, seems not to
provide any form of user id isolation between containers and the host
system.
For example, if I install apache inside a docker container, and also
install it in the host
Pádraig Brady wrote:
I feel this is a polish/integration issue that would be worth finding
some fix for. At least so far, I have not seen any value in df showing
these mount points to root, so perhaps they should be completly omitted
for clarity?
Does `df -a` not show them for root?
Michael Stone wrote:
On Fri, May 23, 2014 at 11:13:32AM -0400, Joey Hess wrote:
df:
‘/var/lib/docker/aufs/mnt/81abbf65e1935a6754e11276c044b94012082304f247be1b70306fa7dfda3650’:
Permission denied
df:
‘/var/lib/docker/containers
Package: dgit
Version: 0.21
Severity: normal
joey@darkstar:~/tmp/git-remote-gcrypt#siddgit fetch wheezy-backports
downloading
http://backports.debian.org/debian-backports//pool/main/g/git-remote-gcrypt/git-remote-gcrypt_0.20130908-5~bpo70+1.dsc...
no version available from the archive
Warning:
Michael Stone wrote:
Ok, scratch that idea. So it looks like there's a combination of bind mounts
and union mounts. The bind mounts should all magically disappear because of
the deduplication logic, but the union mounts are something that probably
shouldn't be entirely hidden by default; I
Frank Lin PIAT wrote:
Please document that dh_installdirs is only meant to be used when you need to
ship an empty directory [1].
[1]
https://wiki.debian.org/Multiarch/Implementation#Dynamic_debian.2F.2A_files
This wiki page is wrong; dh_installdirs is meant to be used in any
situaion where
Jérémy Lal wrote:
As far as i understand it, having less events can save battery life ?
When the git-annex assistant is idle, it blocks indefinitely in an
epoll() call, and so generates no load at all.
If the git-annex assistant is busy syncing files, well, it will use as
much CPU then as
Michael Biebl wrote:
This bug has been open for a while and is now blocking the removal of
hal [0]. What's your plan regarding sleepd? Do you still want to port it
to some newer/other interfaces or should the package be removed from the
archive?
I have no plans for sleepd.
--
see shy jo
gregor herrmann wrote:
This is already implemented since 1.28 but probably not well
documented ... Ah, here it is:
http://kgb.alioth.debian.org/App/KGB/Client/Git.html -- the squash_*
git-config parameters.
Now the question is why the default didn't kick in in your case ...
Are you running
Source: kgb-bot
Version: 1.31-1
Severity: normal
Suppose I am developing on branch A, pushing as I go, with kgb
mentioning each change on IRC. After a long time, I reach a point where
I merge A into branch B. When I push B, kgb will flood IRC with all the
commits, which it has already mentioned
Package: kexec-tools
Severity: wishlist
Low-end VPS hosting providers often do not provide a way to run a
distribution provided kernel, leaving users stranded with no good way to
get security fixes from their distribution. See for example Digital Ocean:
Lars Wirzenius wrote:
I'm proposing to fix this by making Obnam create restored files with
minimal permissions (0300 for directories, 0200 for everything else),
until the data has been restored, and setting the right permissions
afterwards. Does that seem correct to you, Joey?
yes
--
see
Jameson Graef Rollins wrote:
Agreed. I guess we have a couple of options here:
0) don't sign the db file at all
1) sign but don't fail on sig check invalid
2) add option to sign the db and check sig validity
3. Provide a useful error message and leave as-is :)
I added the signing and
Jameson Graef Rollins wrote:
Hey, Joey. None of the assword modules have been updated recently. The
only thing that has been updated recently that might trigger this issue
is python-gpgme.
The exception that we're seeing in this case is a by-product of a gpg
signature validation error.
Jameson Graef Rollins wrote:
This is a known issue and will be fixed in the next release. I'm
curious that you about your broken by upgrade subject, though. Was
this error prompted by an assword upgrade, or by an upgrade of something
else?
I think it was an upgrade of python modules that
Package: analog
Version: 2:6.0-20
Severity: normal
joey@wren:~echo | analog -
analog: analog version 6.0/Unix
analog: Warning D: Turning all pie charts off because OUTFILE is stdout and
CHARTDIR is unset
(For help on all errors and warnings, see
/usr/share/doc/analog/docs/errors.html)
Package: assword
Version: 0.7-1
Severity: important
joey@darkstar:~assword gui
Traceback (most recent call last):
File /usr/bin/assword, line 267, in module
gui(sys.argv[2:], method=method)
File /usr/bin/assword, line 218, in gui
db = open_db(keyid)
File /usr/bin/assword, line 95,
Package: dh-python
Version: 1.20140128-1
Severity: normal
dh_python3
E: dh_python3 tools:93: fix_shebang (debian/debconf/usr/sbin/dpkg-reconfigure):
doesn't look like a shebang: #!/usr/bin/perl -w
E: dh_python3 tools:93: fix_shebang
(debian/debconf/usr/sbin/dpkg-preconfigure): doesn't look
martin f krafft wrote:
Yes, rsync/cp are enforcing the mode/ownership/group membership/ACL
data from the source onto the destination, which I don't think they
should.
git-annex lets rsync/cp preserve as much of the file metadata is they
can, which is generally considered a good thing, and then
martin f krafft wrote:
also sprach Joey Hess jo...@debian.org [2014-04-20 19:37 +0200]:
AFAIK, git-annex has good reasons to use the permissions it uses for
objects directories.
They seem to be according to umask, not according to
core.sharedRepository, which is inconsistent if git-annex
martin f krafft wrote:
In a team environment with a centralised repository, we have now
encountered multiple times a situation by which someone committed
new annex data with git annex sync, but without --content. As
a result, the central repository now knew about the existence of the
new
martin f krafft wrote:
With views, filenames seem to be created as a function of their
original path relative to the repo root. For instance, a file
2013.08.19/DSC_2764.JPG
tagged gallery becomes DSC_2764_%2013.08.19%.JPG when the
tag==gallery view is in effect.
It would be great to
Blocked on http://git-annex.branchable.com/design/caching_database/
--
see shy jo
signature.asc
Description: Digital signature
% setfacl -R -m mask::rwX,group:pm:rwX shared.git
% setfacl -R -d -m mask::rwX,group:pm:rwX shared.git
I don't understand all this ACL stuff[1]. Is this bug report about file
permissions, or about ACLs? AFAIK, git's core.sharedRepository does
not involve ACLs.
I am resorting to reading the tea
Joey Hess wrote:
I think that the level it would make sense to support such a thing in
git-annex is some form of query like `git annex findref $gitref`
Now implemented:
joey@darkstar:~/tmp/test/foo.gitgit annex findref master --in here
file
joey@darkstar:~/tmp/test/foo.gitgit annex findref
Package: obnam
Version: 1.7.4-1
Severity: normal
Tags: security
Imagine obnam restore running on a live system. Perhaps root is
kindly restoring to /home/user/deleted-files-restored-from-yesterday/
on a multi-user system, where /home/user does not have locked down
permissions, so other users can
Simon McVittie wrote:
Recent versions of Automake have a new test driver that records test
output in log files rather than printing it to stdout/stderr. It is
not typically possible to retrieve these log files from the buildds.
If the environment variable VERBOSE is set to a non-empty value
Stephen Kitt wrote:
I don't think so, because in Joey's case, simply creating a new user on the
same system fixes things.
joey@darkstar:~steam
Running Steam on debian 64-bit
STEAM_RUNTIME is enabled automatically
[2014-04-14 15:06:16] Startup - updater built Feb 25 2014 13:44:43
[2014-04-14
Simon McVittie wrote:
Apache 2.4 has forward secrecy, stronger DH/ECDH, and other nice things.
Unfortunately, its Debian packaging is not 100% compatible with 2.2:
sites in /etc/apache2/sites-available/ now have to end with .conf.
And they made sure to maximize pain by making 2.2 not accept
Package: pumpa
Version: 0.8.3-1
Severity: normal
pumpa sometimes double-posts comments. I'm using an identi.ca account.
I have mostly seen this on dialup, but just observed it while on
broadband. Clicked post and the dialog box wet away and came back less
than 1 second later. Reloaded the
501 - 600 of 6096 matches
Mail list logo
