Scribit Laurent Fousse dies 04/10/2007 hora 23:40:
> I find the name quite generic for something that's CL only. Maybe
> `cl-mt19937' would be appropriate?
That's already the name of binary package, FWIW.
I try to name the source package as closely as possible to the upstream
package. So "MT19937
Hi Steve,
I see there's no sign of activty since August 2006 on the
pkg-cvs-maintainers list, and only commits from you sinc October 2006.
As the RFH bug is still open, do you still need help on it?
As I'm willing to become a DD and I'm interested in version control, I'd
like to work on the CVS p
Package: reportbug-ng
Version: 0.2007.08.20
Severity: normal
After my DNS server was accidentally shut down, I invoked Rng and it
tried to get the list of bugs for a package and felt silently to resolve
the address of bugs.d.o, without telling anything else in the GUI than
the fact that it was dow
Package: wnpp
Severity: wishlist
Owner: Pierre THIERRY <[EMAIL PROTECTED]>
* Package name: bordeaux-threads
Version : 0.0.2
Upstream Author : Greg Pfeil
* URL : http://common-lisp.net/project/bordeaux-threads/
* License : MIT
Programming Lang: Commo
Package: wnpp
Severity: wishlist
Owner: Pierre THIERRY <[EMAIL PROTECTED]>
* Package name: cl-vectors
Version : 0.1.3
Upstream Author : Frédéric Jolliton <[EMAIL PROTECTED]>
* URL : http://projects.tuxee.net/cl-vectors/
* License : LLGPL
Prog
Package: wnpp
Severity: wishlist
Owner: Pierre THIERRY <[EMAIL PROTECTED]>
* Package name: mt19937
Version : 1.1
Upstream Author : Douglas T. Crosher and Raymond Toy
* URL : http://www.cliki.net/MT19937
* License : Public Domain
Programming Lang: Commo
Package: wnpp
Severity: wishlist
Owner: Pierre THIERRY <[EMAIL PROTECTED]>
* Package name: cl-salza-png
Version : 1.0
Upstream Author : Zachary Beane <[EMAIL PROTECTED]>
* URL : http://www.xach.com/lisp/salza-png.tgz
* License : BSD
Programming
Package: wnpp
Severity: wishlist
Owner: Pierre THIERRY <[EMAIL PROTECTED]>
* Package name: vecto
Version : 1.0.1
Upstream Author : Zachary Beane <[EMAIL PROTECTED]>
* URL : http://www.xach.com/lisp/vecto/
* License : BSD
Programming Lang:
Scribit Andreas Krüger dies 25/09/2007 hora 17:12:
> However, the following line, executed in your $HOME, produces a file
> name that reliably triggers the problem (Lisp running under an UTF-8
> regime).
>
> perl -e '$w = ">weird-name-" . (pack "C", 196) . "M"; open F, $w'
I confirm this. And I
Scribit Andreas Krüger dies 24/09/2007 hora 09:31:
> > I think I know what causes your problem: it is caused by having a
> > file in your home directory with a name that cannot be encoded with
> > the locale you're trying to use.
> Bingo! That was the ticket!
Not so sure it's a bug in clisp, thoug
Scribit Andreas Krüger dies 22/09/2007 hora 02:55:
> I cannot run clisp without getting this error, see below.
On my Debian system, 'dpkg-reconfigure locales' doesn't show any such
locale ([EMAIL PROTECTED]), so I tried running clisp with de_DE.UTF-8
instead, which worked fine.
When trying with [
Package: sitecopy
Version: 1:0.16.3-11
Severity: important
When trying to create directory during synchronization of a frshly
fetched site, sitecopy tries to create directories but begins with the
deepest ones in the hierarchy, so you actually have to run synch as many
times as there are levels in
Package: ffmpeg2theora
Version: 0.19-1
Severity: wishlist
When transcoding a video with ffmpeg2theora, one has to know which audio
and video quality or bitrate to choose to get neither a degraded nor
oversized video, though it should be trivial for the software to select
optimal values itself.
-
Package: amule
Version: 2.1.3-3+b1
Severity: minor
I had previously setup amule with 8 and 32 kbit/s for UL and DL. I
changed the DL bandwidth to be 128, but noticed after a while that amule
stalled around the same total DL bandwidth as previously.
In the Preferences, I discovered the bandwidth h
# Automatically generated email from bts, devscripts version 2.10.6
# I think reportbug-ng didn't show 435278 when I filed 435749
merge 435278 435749
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Package: droidbattles
Version: 1.0.6-4
Severity: normal
Configuration and bots are saved in ~/droidbattles/, which clutters the
home directory. At least it could be ~/droidbattles/, but preferably it
could be made XDG Basedir compliant, and store
- config in $XDG_CONFIG_HOME
- data in $XDG_DATA
Package: warzone2100
Version: 2.1.0~0.svn1436-1
Severity: grave
When I load a game, the application crashes from a segfault. I was
unable to get valuable information from the core:
Core was generated by `warzone2100'.
Program terminated with signal 11, Segmentation fault.
#0 0x0813c06f in ?? ()
Package: warzone2100
Version: 2.1.0~0.svn1436-1
Severity: minor
The user-specific config and data (saved games) are all gathered in
~/.warzone2100, instead of XDG Basedir directories.
- config could go to $XDG_CONFIG_HOME/warzone2100/
- data could go to $XDG_DATA_HOME/warzone2100/
--- System
Package: funnyboat
Version: 1.5-1
Severity: normal
The resolution of the game cannot be changed. In my configuration
(1280x1024), that makes it either small (windowed) or ugly (fullscreen).
--- System information. ---
Architecture: i386
Kernel: Linux 2.6.17-2-k7
Debian Release: lenny/sid
Package: amule
Version: 2.1.3-1
Severity: important
When invoking amule, starting a single search and then closing it, amule
crashes. Here is the console output in GDB and the complete backtrace:
[New Thread -1514484848 (LW
Package: kbibtex
Version: 0.1.5-5
Severity: minor
When kbibtex considers there's no modification to save and the opened
BibTeX file is modified from outside, kbibtex won't notice the change
and still won't make it possible to save the file.
--- System information. ---
Architecture: i386
Kernel:
close 423263
thanks
Scribit Jörg Sommer dies 11/05/2007 hora 12:16:
> From man bootchartd:
>
> [...]
>
> Did you observe any problems with this? I don't understand your
> question.
Sorry, I remembered having read its manpage but I didn't. From reading
bootchartd's source, it was not obvious it
Package: bootchart
Version: 0.9-4
Severity: wishlist
It would be great if bootchartd could be cloned or adapted to be used
when the system is up and running, to produce the same kind of visual
rendering of system load.
--- System information. ---
Architecture: i386
Kernel: Linux 2.6.17-2-k
Scribit Debian Bug Tracking System dies 24/04/2007 hora 09:54:
> BTW I don't consider this as a bug, since it's perfectly possible to
> report bugs about a package you've not installed.
Of course I would not consider it a bug if the package was not
installed! But it's a bug when the version doesn'
Package: mozilla-plugin-vlc
Version: 0.8.6.a.debian-6
Severity: normal
VLC's plugin for Mozilla has absolutely no UI except for the view of the
video, which makes it exceptionnaly irritating. When a page contains an
embedded video, there's absolutely no way to stop it, either you bear it
or you cl
Package: psi
Version: 0.10-2
Severity: normal
I made the psi window adapts its size automatically, initially offline
contacts are not showed, the contact window is at the bottom right edge
of the desktop and when minimizing, it goes to the tray.
When doing these steps:
- I enable "Show offline c
retitle 421194 Psi doesn't send anything when a PGP key is selected
thanks
I discovered that Psi still works as expected after I unselect my PGP
key. If I reselect it after Psi was launched, go offline then online,
I'll only get the icon of an ongoing connection, but I won't get online.
That is,
Package: psi
Version: 0.10-2
Severity: important
Psi starts in Available state, and any following state change to
something else than Offline will not really change the state:
- nothing shows up in XML console,
- state of the contact of the relevant account remains Available
but the state as sho
Package: dpkg
Version: 1.13.25
Severity: important
When the /var/lib/dpkg/available file is full of zeroes, some dpkg
incoations fail segfaulting. I attach informations from gdb running
dpkg-query that permitted to find why dpkg segfaults.
In such a case, dpkg should die segfaulting, but handle t
Package: reportbug
Version: 3.35
Severity: normal
When dpkg fails to inform reportbug of the version of an installed
package, reportbug just leaves the version field empty and won't notify
the user of the problem.
--- System information. ---
Architecture: i386
Kernel: Linux 2.6.17-2-k7
Deb
Package: reportbug-ng
Version: 0.2007.04.13
Severity: normal
When dpkg fails to inform reportbug-ng of the version of an installed
package, reportbug-ng just leaves the version field empty and won't
notify the user of the problem.
--- System information. ---
Architecture: i386
Kernel: Linux
Writing in /proc will not persist reboot, so the bug may well hit the
user after installation again, for any network activity.
Taking the suggested values, here should be the "fix" for the installed
system to be able to do TCP networking:
# cat <> /etc/sysctl.conf
net.ipv4.tcp_rmem = 4096 65536 6
The previous package was not quite good as far as quality matters, but
it was usable nonetheless.
In the meantime, Elephant developpers are preparing the next release,
0.6.1. I'll package this version and try to have it uploaded in the
archive.
Quickly,
Pierre
--
[EMAIL PROTECTED]
OpenPGP 0xD9D5
Package: ktorrent
Version: 2.0.3+dfsg1-2.2
Severity: wishlist
Command-line utilities, like bt{download,launchmany}{,curses}, make it
possible to store a torrent in a specific directory and download the
file in that directory.
Having all files downloaded to a unique place makes it a pain to
downlo
Package: linux-image-2.6.17-2-k7
Version: 2.6.17-9
Severity: serious
After doing an LVM snapshot of my /home, I tried to mount it with the
nouuid option, as it is XFS. The mount process stopped in an
uninterruptable sleep for a long time (many minutes). It was impossible
to kill it, even with the
Package: amarok
Version: 1.4.4-4
Severity: wishlist
Quite some times, the OSD bothered me where it appeared and I wanted it
to disappear, but as it had reached its timeout, it disapperead just as
I was clicking, making me click on something I didn't want (and it's not
far from typically critical b
Scribit Michael Biebl dies 14/03/2007 hora 09:20:
> How big is your home directory? What kind of data does it contain?
Approximately 25Gb, much of it being texts documents like source code.
> Could you post the output of tracker-stats:
It could connect to trackerd, it seems, so I didn't manage t
Package: tracker
Version: 0.5.4-4
Severity: normal
I discovered today, because I found the size of my home dir supsicious,
that tracker had stored nearly 1Gb of data in a dot-dir in my $HOME,
which I find rather intrusive!
Couldn't tracker respect the XDG basedir spec and store that kind of
data
Package: beagle
Version: 0.2.12-1+b1
Severity: normal
I discovered today, because I found the size of my home dir supsicious,
that beagle had stored 3Gb of data in a dot-dir in my $HOME, which I
find rather intrusive!
Couldn't beagle respect the XDG basedir spec and store that kind of data
in the
Scribit Ana Guerrero dies 03/03/2007 hora 02:16:
> #296564 "korganizer: Alarm daemon should be configurable to start
> kontact"
Thanks for taking aold wishlist bugs in account! For that matters, yes,
the bug is still totally valid today. I still don't see any UI to
configure the daemon to do any
Package: pangzero
Version: 1.1-1
Severity: important
The fact that by default, pangzero will fire up the browser to show its
website is rather intrusive.
-- System Information:
Debian Release: 4.0
APT prefers testing
APT policy: (990, 'testing'), (502, 'stable'), (501, 'unstable'), (500,
'un
Package: qingy
Version: 0.9.3ds1-1
Severity: important
When using my normal login/password, that is perfectly working on
another tty managed by getty or KDM, I get the following in
/var/log/auth.log:
Feb 20 14:01:42 bateleur qingy[29753]: Authentication failure: wrong password
for user 'rabbit'
Package: cl-swank
Version: 1:20061201-2
Severity: normal
Tags: patch
The loop in fuzzy-completion-set has a while clause before a loop, which
is inconsistent with the loop specification. It makes loading latest
swank in ECL fail.
The following patch corrects it:
diff -r bc0d362e751c swank.lisp
Package: ecl
Version: 0.9i-2
Severity: important
Tags: patch
Justification: fails to build from source
ECL throws an error during compilation if neither MAXPATHLEN nor
PATH_MAX are defined, though POSIX explicitely permits those constants
to be undefined. Applications should test for their definit
Package: ecl
Version: 0.9i-2
Severity: important
Tags: patch
Justification: fails to build from source
Here are two patches to make the build possible on Hurd:
diff -r 045b9347e44a src/h/config.h.in
--- a/src/h/config.h.in Tue Feb 06 17:36:54 2007 +0100
+++ b/src/h/config.h.in Sat Feb 10 04:15:
Package: ecl
Version: 0.9i-2
Severity: important
Tags: patch
Justification: fails to build from source
Not all uses of functions and flags related to SA_SIGINFO where protected by the
needed #ifdef:
diff -r 0169839bd656 src/c/unixint.d
--- a/src/c/unixint.d Sat Feb 10 04:52:54 2007 +0100
+++ b
Package: ecl
Version: 0.9i-2
Severity: important
Justification: fails to build from source
When the source package is unpacked in a path that contains a space, the
configure script fails at line 37.
-- System Information:
Debian Release: 4.0
APT prefers testing
APT policy: (990, 'testing'), (
Package: zim
Version: 0.17-1
Severity: normal
When a page has only one child, it is not shown as expandable in the
side pane, and the child is invisible in the pane.
-- System Information:
Debian Release: 4.0
APT prefers testing
APT policy: (990, 'testing'), (501, 'stable'), (500, 'unstable')
Package: cl-utilities
Version: 1.2.4-1
Severity: minor
Tags: patch
The documentation for the once-only macro should be written instead of
pointing to a newsgroup discussion describing it. Here is a patch that
adds the relevant description:
Index: Source/doc/once-only.html
Is there any news on the packaging of moodbar? Is there anything that
can be made to help, like testing the package?
Curiously,
Pierre
--
[EMAIL PROTECTED]
OpenPGP 0xD9D50D8A
signature.asc
Description: Digital signature
Package: linux-igd
Version: 0.cvs20060201-1.1
Severity: important
At configure time, I got:
External interface not specified in /etc/default/upnpd
invoke-rc.d: initscript upnpd, action "start" failed.
And the package ends in failed-config state. The (post-inst?) script should
either not bother a
Package: wnpp
Severity: wishlist
* Package name: moodbar
Version : 0.1.2
Upstream Author : Joe Rabinoff <[EMAIL PROTECTED]>
* URL : http://amarok.kde.org/wiki/Moodbar
* License : GPL
Programming Lang: C
Description : Analysis program for creating a color
Package: cl-unit
Version: 1.3.1-2
Severity: normal
When trying to ASDF load the package, I got the following error:
The name "REGRESSION-TEST" does not designate any package.
[Condition of type SB-KERNEL:SIMPLE-PACKAGE-ERROR]
while loading the "clunit" ASDF source file.
-- System Information
Package: amule
Version: 2.1.3-1
Severity: normal
Several times already, upgrading amule made me lose my configuration.
Specifically, I only seem to lose what's in ~/.aMule/amule.conf. Shared
folders are still as I configured them, but the incoming and temporary
folders get back to their default, a
Package: amule
Version: 2.1.3-1
Severity: normal
When doing many searchs in a fast time range, the results often come up
in the list of the last search.
-- System Information:
Debian Release: 4.0
APT prefers testing
APT policy: (990, 'testing'), (501, 'stable'), (500, 'unstable'), (500,
'sta
tags 397310 + patch
thanks
Here is the very simple patch that should make user authentication work
with the default installation of apache2. I confirm that this bug bites
many people, and some people in #apache on Freenode seem now frankly
pissed by Debian because of it. :-/
Index: Source/debian/
Package: kdm
Version: 4:3.5.5a-2
Severity: minor
This installation of KDM gave me some information about a config file
that is now obsolete, but it was just output during dpkg's processing.
If I weren't watching it at this very moment, I wouldn't know. I think
Debconf has a feature for that (and m
Package: araneida
Version: 0.90.1-dfsg-1
Severity: normal
Tags: patch
When I tried to receive big POST requests in multipart/form-data MIME
type, araneida failed because it tried to parse the body. The following
patch made it work:
Index: Source/daemon.lisp
===
Package: ceferino
Version: 0.97.5-1
Severity: normal
In some keyboards layouts, the default keys of ceferino are just
unusable. Changing them would be necessary.
-- System Information:
Debian Release: testing/unstable
APT prefers testing
APT policy: (990, 'testing'), (501, 'stable'), (500, 'u
Package: wormux
Version: 0.7.2-5
Severity: normal
When the jetpack has been selected by any mean, and keyboard is used to
switch to the next weapon in that category (that is, pressing ), the
game reproducibly segfaults. Here is the backtrace, I can provide the
core dump if necessary:
#0 0x00
Package: crawl
Version: 1:4.0.0beta26-8
Severity: wishlist
It should be possible to save the game and take it back when one wants
to. For me, it was killing the fun to be killed in a silly way after
some thousands of rounds going well, and I added that facility with some
shell scripts, but it's fa
As I was playing with slime to see how different backends work, I hit
that bug, but it only occured to me with CLisp. SBCL and CMUCL just
worked.
Creating an empty /usr/share/common-lisp/source/slime/ChangeLog file was
my simple workaround for this bug.
Quickly,
Nowhere man
--
[EMAIL PROTECTED]
Package: python-2play
Version: 0.1.9-1
Severity: normal
It seems impossible to remove the package without altering it's postrm:
:~# dpkg --purge python-2play
(Reading database ... 651922 files and directories currently installed.)
Removing python-2play ...
Traceback (most recent call last):
Fil
Package: evolution
Version: 2.6.1-3
Severity: critical
Justification: causes serious data loss
I tracked down that evolution, after being launched by accident
(clicking on a mailto: URL in Firefox), not only stayed running after I
quit it, but screwed up my maildirs.
It regularily moved all my ne
Package: libgtkmm-2.4-doc
Version: 1:2.6.5-1
Severity: minor
In the class documentations, compared to the online documentation, the
image representing inheritance is missing... It is strange for the
reader because the legend is still there, though.
-- System Information:
Debian Release: testing/u
Scribit Debian Bug Tracking System dies 22/05/2006 hora 08:18:
> Comix should handle the sorting correctly, compare to the output of
> `ls` if you wan't to check it for yourself.
But isn't that a bit nonsense? Comix is made to see numbered files in a
specific order, and no locale sorts numbered fi
Package: aircrack-ng
Version: 0.3-1
Severity: normal
I have a set of .cap files, namely out-0[1234].cap. I convert each in an
.ivs file.
`aircrack-ng out*.ivs` gives me a list of networks.
Then I do
$ ivstools --merge out*.ivs merge.ivs
But the result file seems to be invalid:
$ aircrack-ng m
Package: doc-central
Version: 1.8.2
Severity: normal
Each request to view doc-central regenerate the contents frame, which a
big time loss. Especially as the output of the CGI script could be
easily cached. Instead of regenerating it, some make-like system should
check if any source file has been
Package: xserver-xorg
Version: 1:7.0.14
Severity: normal
When I use xkbsel or xkbsel-aw to select my keyboard layout, my X
server crashes. Or dies. I was running xkbsel-aw from bash in screen,
here is it's output at the very moment (X only crashes when I select a
layout):
$ xkbsel-aw
XIO: fatal
Package: shishid
Version: 0.0.23-1
Severity: grave
Justification: renders package unusable
The shishid script in /etc/init.d doesn't start the daemon in the
background, so the server won't be started after installation (you're
forced to kill it to let dpkg end it's job).
It may also prevent the s
Package: kasablanca
Severity: wishlist
Could kasablanca also support the WebDAV protocol, which has similar
goals than FTP?
-- System Information:
Debian Release: testing/unstable
APT prefers testing
APT policy: (990, 'testing'), (501, 'stable'), (500, 'unstable'), (500,
'stable')
Architectu
Package: kftpgrabber
Severity: wishlist
Could kftpgrabber also support the WebDAV protocol, which has similar
goals than FTP?
-- System Information:
Debian Release: testing/unstable
APT prefers testing
APT policy: (990, 'testing'), (501, 'stable'), (500, 'unstable'), (500,
'stable')
Architec
Package: pioneers-client
Version: 0.9.55-1
Severity: wishlist
The game selection dialog should use mDNS/DNS-SD to provide a list of
available games.
-- System Information:
Debian Release: testing/unstable
APT prefers testing
APT policy: (990, 'testing'), (501, 'stable'), (500, 'unstable'), (5
Package: pioneers-server-gtk
Version: 0.9.55-1
Severity: wishlist
When the server is started, it should publish it's service with
mDSN/DNS-SD.
-- System Information:
Debian Release: testing/unstable
APT prefers testing
APT policy: (990, 'testing'), (501, 'stable'), (500, 'unstable'), (500,
'
Package: pioneers-server-gtk
Version: 0.9.55-1
Severity: wishlist
Isn't it a bug that there is two separate binaries that act as a
pioneers server, one providing CLI and the other a GUI?
IMHO, there could be a deamon, and different administration
interfaces... Wouldn't it make sense?
-- System I
Package: update-notifier
Version: 0.41.3-1
Severity: normal
It is impossible to use update-notifier without giving the root
password, although it seems to me that most operations I understood it
is made for are easily made by an unpriviledged user.
-- System Information:
Debian Release: testing/u
Package: viruskiller
Version: 1.0-1-1
Severity: normal
When starting, the application searches the directories of the user. At
this moment, there is no way to shut down the application, and the
process can take some time if you have a very large number of
directories and files.
-- System Informat
Package: katapult
Version: 0.3.1-1
Severity: important
Clicking the menu just crashes the application. Attached is the gdb
backtrace...
-- System Information:
Debian Release: testing/unstable
APT prefers testing
APT policy: (990, 'testing'), (501, 'stable'), (500, 'unstable'), (500,
'stable'
Package: wnpp
Severity: wishlist
* Package name: bss
Version : 0.6
Upstream Author : Pierre Bertouin <[EMAIL PROTECTED]>
* URL : http://www.secuobs.com/news/05022006-bluetooth10.shtml
* License : GPL
Description : BSS (Bluetooth Stack Smasher) is a L2CAP l
Package: aptitude
Version: 0.4.1-1
Severity: wishlist
There is %v and %V for the version installed and the one that would be
installed or updated, but if there is the %t for the archive(s) of the
package version that would be installed or updates, there is no such
format string %-escape for the ar
Package: aptitude
Version: 0.4.1-1
Severity: minor
In the dependency solver (a great enhancement of aptitude!), when a
package version is the one installed, it is shown as coming from 'now'.
But the %t in a format string for the package display doesn't shwo
anything from packages only available lo
Package: python-wxglade
Version: 0.3.5.cvs20050824-0.2
Severity: wishlist
When adding a wxBoxSizer, it's slots are shown in the design window, but
not in the tree. There should be an option to make them first-class
citizens of the tree view, IMHO (to be able to remove them there, at
least):
+ App
Scribit Matej Vela dies 25/11/2005 hora 14:58:
> Well, if you have the patience, try doing a backtrace with an
> unstripped Tcl/Tk from
>
> deb http://people.debian.org/~vela/xconq-dbg/ ./
I had the patience, but not the time to. Now I have, so here's the
baktrace:
(gdb) run
Starting program
Package: libboost-doc
Version: 1.32.0-6
Severity: minor
Some hyperlinks point to the SGI web site for information about the STL,
but this needs an internet link, although the needed files are present
in the documentation package of the STL in Debian.
e.g. in HTML/libs/graph/doc/adjacency_list.htm
Package: php4-sqlite
Version: 1.0.2-9
Severity: serious
Justification: Policy 7.2
php4-sqlite doesn't depends on apache, but its postinst tries to do
something with it, thus when apache is not installed, the package is
impossible to install, and to remove...
[EMAIL PROTECTED]:~$ dpkg --purge php4
tag 340608 + patch
thanks
Scribit Michael Tautschnig dies 26/11/2005 hora 15:19:
> Are the files created by fai-setup/make-fai-nfsroot really to be
> considered "installed files", or does the policy rather talk about
> files installed by dpkg?
I think it is a reasonable expecting that the behavio
reopen 335313
found 335313 2.0.0-2
tag 335313 + patch
thanks
There are still two oodraw2 commands:
diff -U 0 debian/openoffice.org-draw.mime debian.new/openoffice.org-draw.mime
--- debian/openoffice.org-draw.mime 2005-11-22 19:00:28.573330664 +0100
+++ debian.new/openoffice.org-draw.mime 2005
Scribit Steve Langasek dies 26/11/2005 hora 03:47:
> No, because this data *is* both shareable and read-only; it is written
> to only by certain admin operations.
I don't see how you can still consider data that is sometimes modified
by priviledged users read-only... Data only modified by the core
Scribit Steve Langasek dies 25/11/2005 hora 18:42:
> I certainly agree that it's desirable to never have anything written
> to /usr except by the package management system and to be able to keep
> it read-only otherwise, but I don't find that the FHS mandates this.
I found, indeed:
``/usr is sh
Scribit Thomas Lange dies 25/11/2005 hora 17:35:
> I like to skip the move to /var/lib/fai, and wait until I can finally
> move to /srv.
But this is still a bug, and a policy violation. Users applying Debian
security guidelines will still encounter this bug with the default
configuration...
Prag
Scribit Stephen Gran dies 25/11/2005 hora 15:19:
> My understanding is that while /srv is the right place for this kind
> of data, it would be incorrect for Debian packages to dump stuff
> there. /srv is the domain of the local admin.
This is precisely why it should be put there by fai-setup. fai
Scribit Thomas Lange dies 25/11/2005 hora 15:34:
> My future plans are to move it to /srv, but the question is, if it's
> really a FHS violation.
nfsroot can be updated, regenerated, modified to fit the user's needs,
and so on. I don't see how it can really be seen read-only. So it can't
be in /us
Scribit Steve Langasek dies 25/11/2005 hora 00:56:
> > According to the FHS, ``/usr is shareable, read-only data''. So FAI
> > should not by default try to write anything in /usr and place it's
> > nfsroot there. See #309554.
> Could you elaborate on why you believe this is an FHS violation?
My /u
Package: fai
Version: 2.8.4
Severity: serious
Justification: FHS
According to the FHS, the /mnt directory is the place where a temporary
mount should be done. At least, the / directory should not be the place
where a custom FAI directory is created. AFAIK, /media is the perfect
place.
-- System I
Package: fai
Version: 2.8.4
Severity: serious
Justification: FHS
According to the FHS, ``/usr is shareable, read-only data''. So FAI
should not by default try to write anything in /usr and place it's
nfsroot there. See #309554.
-- System Information:
Debian Release: testing/unstable
APT prefers
Scribit Steve Langasek dies 18/11/2005 hora 18:19:
> While it is desirable to have sed preserve EAs and ACLs when used with
> -i, I think this severity is overinflated and the security tag is
> incorrect.
I won't argue on the severity (I was not really sure which I had to
choose), but the bug inde
Package: sed
Version: 4.1.2-8
Severity: grave
Tags: security
Justification: user security hole
When doing in-place editing, sed creates a new file without copying ACLs
and user-defined EA. It's not only a loss of maybe precious data
(user-defined EA) but a security hole, because dropping the ACLs
There is a problem in your patch, at least according to the gpg manpage:
RETURN VALUE
The program returns 0 if everything was fine, 1 if at least a
signature was bad, and other error codes for fatal errors.
I have the same problem, and obviously those errors are not fatal as gpg
contin
Scribit Pierre THIERRY dies 09/11/2005 hora 18:26:
> we could use rel="prefetch"
You'll maybe notice it seems in contradiction with my comment on
#210697, but I think it is in fact in it's spirit: if we use rel="next"
for both the image and the page, one browser
Scribit Martin Michlmayr dies 04/11/2005 hora 19:51:
> Does anyone know whether two "link rel" elements are allowed
It does not make sense IRT the HTML spec, but as here the purpose is
specifically that the image be prefetched, we could use rel="prefetch".
I heard it is working on Mozilla, I don't
101 - 200 of 306 matches
Mail list logo
