I recreated the guest OS from scratch (netinst etch -> upgrade lenny)
and, until now, I can still boot this kernel. So, it may have been an
issue with the guest OS, not with kvm, may be bug #475923 (I was in
sid and the upgrade of the console failed).
I'll do more tests tomorrow and report them h
Package: kvm
Version: 60+dfsg-1
Severity: normal
My kvm host is a Dell with an Opteron processor. The host is a Debian
sid. Installation (of etch via netinst) went fine but, after upgrading
the host to sid, the kernel 2.6.18-6-amd64 boots but cannot mount the
root. Here is the complete log:
Boot
Package: kvm
Version: 60+dfsg-1
Severity: normal
My kvm host is a Dell with an Opteron processor. The host is a Debian
sid. Installation (of etch via netinst) went fine but, after upgrading
the host to sid, the kernel 2.6.24-1-amd64 boots but cannot mount the
root. Here is the complete log:
Init
On Fri, Apr 11, 2008 at 05:05:14PM +0200,
Julien Cristau <[EMAIL PROTECTED]> wrote
a message of 16 lines which said:
> Apparently it's --no-ssh-cm.
Same problem, even after 'aptitude dist-upgrade' this morning.
% darcs pull --no-ssh-cm
%
(There *are* patches to pull.)
--
To UNSUBSCRIBE,
On Fri, Apr 11, 2008 at 08:01:14AM -0700,
David Roundy <[EMAIL PROTECTED]> wrote
a message of 12 lines which said:
> I see, it's perhaps been renamed to --no-ssh-cm?
Indeed, but it doesn't help:
% darcs pull --no-ssh-cm
%
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject
On Fri, Apr 11, 2008 at 07:29:28AM -0700,
David Roundy <[EMAIL PROTECTED]> wrote
a message of 22 lines which said:
> Have you tried running with --disable-ssh-cm?
% darcs pull --disable-ssh-cm
darcs failed: unrecognized option `--disable-ssh-cm'
--
To UNSUBSCRIBE, email to [EMAIL PROTECT
On Fri, Apr 11, 2008 at 02:54:43PM +0200,
Stephane Bortzmeyer <[EMAIL PROTECTED]> wrote
a message of 44 lines which said:
> Every network-using command (pull, push, get) behaves the same.
>
> darcs changes, revert, record or get/pull/push with local repositories
> work fine
Package: darcs
Version: 1.0.9-1.1
Severity: grave
Justification: renders package unusable
% darcs get [EMAIL PROTECTED]:Blog
%
Every network-using command (pull, push, get) behaves the same.
darcs changes, revert, record or get/pull/push with local repositories
work fine.
ssh is working proper
Package: gdc
Version: 4.1.2-12
Severity: important
myriam:~/Programmation/D % cat hello.d
import std.stdio;
int main ()
{
writefln("Hello world!");
return 0;
}
myriam:~/Programmation/D % gdc hello.d
hello.d:1: module stdio cannot read file 'std/stdio.d'
And, indeed:
myriam:~/Programma
Package: curl
Severity: grave
Justification: renders package unusable
On a sid machine, just after an aptitude update this morning:
% sudo aptitude install curl
Reading package lists... Done
Building dependency tree
Reading state information... Done
Reading extended state informati
Package: echoping
Version: 6.0.2-1
Severity: minor
The echoping 6 package has a lot of dependencies that many users will
not need. For instance, not everyone will use the LDAP plugin and the
non-users will not be happy to have to install the LDAP client libs.
I suggest to address this problem. T
On Wed, May 30, 2007 at 06:16:56PM -0700,
Russ Allbery <[EMAIL PROTECTED]> wrote
a message of 37 lines which said:
> The behavior of -y and of `cat file` is different. -y uses every
> byte of the file as the password, including any trailing newline,
> whereas backticks will strip any trailing
Package: libbind-dev
Version: 1:9.4.1-1
Severity: normal
Unlike the old bind-dev package for BIND8, it seems libbind-dev does
no ship libbind.* library files (I don't talk about libbind9.* but
about libbind.*, they have a different content.)
This is a problem for programs like milter-greylist
(h
BTW, echoping 6.0.1 (just released) installs the headers in
$PREFIX/include/echoping.
signature.asc
Description: Digital signature
Problem fixed in release 6.0.1. Many thanks for the bug report.
signature.asc
Description: Digital signature
Thanks to all those who search and reported solutions for Debian bug
#397774. I just upgraded a Subversion server (authenticating with
LDAP) to "etch" and the following config works for me:
# Disable basic file-based authentication
# Starting with Apache 2.2, everything changed :-(
# See D
tag 340310 upstream
thanks
Problem registered upstream as
http://sourceforge.net/tracker/index.php?func=detail&aid=1688939&group_id=4581&atid=104581
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
tag 407430 upstream
thanks
Problem registered upstream as
http://sourceforge.net/tracker/index.php?func=detail&aid=1688940&group_id=4581&atid=104581
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Commenting out the "free(hostname)" suffices to solve the problem. For
more study.
Here is the backtrace:
(gdb) bt
#0 0x2b44570428d5 in raise () from /lib/libc.so.6
#1 0x2b4457043b6e in abort () from /lib/libc.so.6
#2 0x2b4457078987 in ?? () from /lib/libc.so.6
#3 0x2b445707da
I cannot reproduce the bug in echoping version 6 (just shipped). I did
not fix it on purpose :-) but it seems fixed nevertheless.
I confirm the bug in 5.2. I hesitate to make a "security fix" since
there is no real security issues. But it is indeed a nasty bug.
signature.asc
Description: Digita
Package: echoping
Version: 5.2.0-2.1
Severity: wishlist
echoping 6 is out:
http://echoping.sourceforge.net/
But it is more complicated to package since it now introduces dynamic
plugins. Several plugins are shipped with echoping, for instance the
PostgreSQL one. It is probably not wise to make
Package: netbase
Version: 4.29
Severity: minor
UDP-Lite, protocol 136, described in RFC 3828, is missing in
/etc/protocols.
Although I do not think that /etc/protocols should list only the
protocols implemented in Linux (think of tcpdump, for instance),
UDP-Lite *is* in Linux:
http://lwn.net/Ar
This orphaning is also for the dnsdoctor-cgi package (same source
package).
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Package: wnpp
Severity: normal
I intend to orphan the dnsdoctor package. Because I leave Debian.
The package is complicated and strongly depends on Ruby.
It seems unmaintained upstream. May be a merge with zonecheck (same
code base) should be considered.
The package description is:
DNSdoctor
Package: wnpp
Severity: normal
I am retiring from Debian (cause: lack of activity)
I intend to orphan the echoping package.
No pending bugs. The package is quit simple and suitable for a beginner.
The package description is:
Can test if a server is listening on a remote machine
and can measu
On Wed, Jan 31, 2007 at 10:37:58AM -0800,
Sebastien Delafond <[EMAIL PROTECTED]> wrote
a message of 27 lines which said:
> So... do you think the problem is in the zonecheck code, or rather
> something to do with the way libxml-ruby1.8 is packaged ?
No idea, I'm afraid.
--
To UNSUBSCRIBE,
On Tue, Jan 23, 2007 at 05:55:35PM -0800,
Sebastien Delafond <[EMAIL PROTECTED]> wrote
a message of 54 lines which said:
> I am not seeing that on unstable as of today;
I did see the problem on "unstable", too. And I've found a solution:
apt-get remove libxml-ruby1.8
This solves the proble
A colleague had it on Fedora Core 5, too. Same mystery.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
On Tue, Jan 23, 2007 at 05:55:35PM -0800,
Sebastien Delafond <[EMAIL PROTECTED]> wrote
a message of 54 lines which said:
> Stephane, do you have other etch boxes where you can reproduce this
> problem ?
Right, another etch box does not have the problem (both have been
'aptitude dist-upgrade' t
Package: zonecheck
Version: 2.0.4-6
Severity: important
% zonecheck --ns 'NS1.DREAMHOST.COM;NS2.DREAMHOST.COM;NS3.DREAMHOST.COM'
france-japon.fr
ZONE : france-japon.fr.
NS <= : ns1.dreamhost.com. [66.33.206.206]
NS : ns2.dreamhost.com. [66.201.54.66]
NS : ns3.dreamhost.com. [66.33.216.2
Package: whois
Version: 4.7.17
Severity: normal
% whois indom.mobi
This TLD has no whois server, but you can access the whois database at
http://pc.mtld.mobi/whois/
But there is such a server, officially announced by IANA:
% whois -h whois.dotmobiregistry.net indom.
On Fri, Sep 15, 2006 at 11:21:12PM +0200,
Arjan Oosting <[EMAIL PROTECTED]> wrote
a message of 69 lines which said:
> frown - LALR(k) parser generator for Haskell 98
Before I read the documentation, how does it compare with Parsec and
Happy which are already in Debian?
--
To UNSUBSCRIBE, em
Upgrading my workstation from sarge to etch broke
wdm. Authentification is always refused "ERROR login failed".
login and xdm have no problem.
The transition described in
http://lists.debian.org/debian-devel-announce/2003/debian-devel-announce-200308/msg00012.html
has been completed a long time a
> Since this bug was in grepmail's cache code, it's possible that
> other changes in its cache (~/.grepmail-cache) have made it not
> reprodce anymore. You might try deleting your current cache,
Upgrading (with aptitude) from sarge to etch seems sufficient to
triggers the bug. Your workaround (rm
Package: dnsutils
Version: 1:9.3.2-2
Severity: minor
dnsutils now includes a man page for the program query-loc (which is
in BIND's contrib/) but not the program itself. Why?
Including the man page without the program is worse than nothing since
it conflicts with my unofficial query-loc package.
> Could you please try with a Debian kernel image
There is one? I do not find it in either etch or sid (or on
Alioth). Do I have to recompile a kernel?
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Package: xen-utils-3.0
Version: 3.0.2+hg9656-1
Severity: grave
Justification: renders package unusable
When xend starts:
[2006-05-22 22:07:18 xend] INFO (SrvDaemon:278) Xend Daemon started
[2006-05-22 22:07:18 xend] INFO (SrvDaemon:282) Xend changeset: Thu Apr 27
09:58 :50 2006 +0100
Package: graphviz
Version: 2.2.1-1sarge1
Severity: minor
When using the (documented) option 'charset = iso-8859-1', I observe
that neato does not set the proper XML encoding when it produces
SVG. Hence the output file is rejected as not well-formed.
% more sample.dot
graph network {
charset
Package: rss2email
Version: 1:2.54-2
Severity: normal
When ran against my blog, http://www.bortzmeyer.org/feed.atom,
rss2email produces emails like:
From: "Blog de =?utf-8?Q?St=E9phane?= Bortzmeyer, =?utf-8?Q?St=E9phane?=
Bortzmeyer" <[EMAIL PROTECTED]>
They are invalid: the encoding says "UTF
Package: hlfl
Version: 0.60.1-1
Severity: normal
~ % hlfl -v
zsh: segmentation fault hlfl -v
(Otherwise, it works. -v option is described in the man page.)
-- System Information:
Debian Release: 3.1
Architecture: i386 (i686)
Kernel: Linux 2.6.8-2-686
Locale: LANG=C, LC_CTYPE=fr_FR (charmap=ISO
On Mon, Jan 16, 2006 at 12:06:54PM +0100,
Andreas Rottmann <[EMAIL PROTECTED]> wrote
a message of 14 lines which said:
> JFYI: I just uploaded a darcsweb 0.14-1 to sid.
I tested it on a sarge machine and it works fine except if the commit
messages contain Latin-1 characters (darcs does not han
Package: wnpp
Severity: wishlist
* Package name: darcsweb
Version : 0.13
Upstream Author : Alberto Bertogli <[EMAIL PROTECTED]>
* URL : http://users.auriga.wearlab.de/~alb/darcsweb/
* License : BOLA (see
http://lists.debian.org/debian-legal/2006/01/msg00098.ht
Package: wnpp
Severity: wishlist
* Package name: rnv
Version : 1.7
Upstream Author : David Tolpin, Davidashen <[EMAIL PROTECTED]>
* URL : http://ftp.davidashen.net/PreTI/RNV/
* License : BSD
Description : Relax NG Compact Syntax Validator in C
rnv is a v
I just tried dillo on a sid machine (0.8.5-4) and the problem is now
the opposite (same Web site, http://www.bortzmeyer.org/). UTF-8
characters are OK but characters coded as entities are not displayed
at all.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Troubl
Package: dillo
Version: 0.8.3-1
Severity: normal
On a Web site like http://www.bortzmeyer.org/, Unicode characters
represented as numeric entities (é) are displayed fine but when
they are direclty in UTF-8, Dillo displays the bytes uninterpreted.
AFAIK, content-type in is correct, and HTTP head
forwarded 339433 Sourceforge BTS
quit
Forwarded upstream as #1366369. See
https://sourceforge.net/tracker/index.php?func=detail&aid=1366369&group_id=87005&atid=581684
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Package: liferea
Version: 0.9.1-1
Severity: normal
For ATOM feeds (like http://www.bortzmeyer.org/feed.atom, which is a
valid and tested feed) no date is displayed by Liferea. (RSS feeds are
OK.)
-- System Information:
Debian Release: 3.1
Architecture: i386 (i586)
Kernel: Linux 2.4.27-2-k6
Local
tag 333490 upstream
thanks
Upstream bug is against feedparser.org (#1312408 at Sourceforge). You
can follow it here:
http://sourceforge.net/tracker/index.php?func=detail&aid=1312408&group_id=112328&atid=661937
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Troub
tag 333490 patch
thanks
Here is a patch against fedparser.py which seems to work. The
namespace is the official namespace of Atom 1.0 (see the
Internet-Draft, soon to be RFC).
--- feedparser.py.orig 2005-10-12 11:33:40.0 +0200
+++ feedparser.py 2005-10-12 11:44:15.0 +0200
@
Package: rss2email
Version: 1:2.54-2
Severity: normal
When you "add" an Atom field (one which uses the final Atom standard,
specified in Internet-Draft draft-ietf-atompub-format-11, approved by
IESG and which will be a RFC soon), "r2e run" fails:
=== SEND THE FOLLOWING TO [EMAIL PROTECTED] ===
E
On Tue, Sep 13, 2005 at 06:51:56AM +0530,
Ganesan Rajagopal <[EMAIL PROTECTED]> wrote
a message of 44 lines which said:
> 1.0.3 is already in testing.
Oops, I run "sarge" and did not test properly that it exists on "sid"
(I tried also on a machine which was supposed to run sid but which was
"
Package: python2.3-clientcookie
Version: 0.4.19-1
Severity: wishlist
There is, for more than a year, a 1.0.3 on the upstream site
(http://wwwsearch.sourceforge.net/ClientCookie/) and interesting
programs like mechanize (http://wwwsearch.sourceforge.net/mechanize/)
require it.
Here, installing me
Unlike the man page, online documentation
(http://www.haskell.org/happy/doc/html/sec-invoking.html) says that
you *must* use -a with -d (something which is unfortunately not
checked by Happy). Indeed, it seems to solve the problem.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of
Package: happy
Version: 1.15-1
Severity: minor
Tags: patch
"man happy" says:
-d, --debug
Generate a parser that will print debugging information to
stderr at run-time, including all the shifts, reductions, state
transitions and token inputs
Package: ipmasq
Version: 4.0.2
Severity: important
Tags: patch
ipmasq-kmod cannot load important modules (like ip_conntrack_ftp)
because it looks only for "*.o" files under /lib/modules while 2.6
kernels use "*.ko".
Being unable to work with 2.6 is enough to deserve an "important"
severity.
--
Yes, you simply have to recompile the current source package, without
a change (tested on i386 and sparc, with the current up-to-date sid).
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
#309648 appears as fixed while the bug is still in sarge. sarge being
frozen, the bug does *not* appear as RC
(http://bugs.debian.org/release-critical/debian/all.html).
I believe we should reopen the bug (with its current tags, including
"sarge").
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
w
Package: python2.3-epydoc
Version: 2.1-8
Severity: normal
epydoc --html ignores the encoding of the Python source file (PEP 263)
and always put a:
which probably comes from my locale (it should come from the encoding
specified as '# -*- coding: utf-8 -*-').
-- System Information:
Debian Relea
Package: wnpp
Severity: wishlist
* Package name: vampire
Version : 1.5
Upstream Author : Graham Dumpleton <[EMAIL PROTECTED]>
* URL : http://www.dscpl.com.au/projects/vampire/
* License : BSD (http://www.dscpl.com.au/projects/vampire/license.html)
Description
On Fri, Apr 08, 2005 at 02:26:26AM -0700,
Steve Langasek <[EMAIL PROTECTED]> wrote
a message of 44 lines which said:
> > The lack of a copyright file (the licence is there, but not the
> > mandatory copyright file) seems a violation of
> > http://www.debian.org/doc/debian-policy/ch-docs.html#s-
Package: graphviz
Version: 2.2-2
Severity: serious
Justification: Policy 12.5
The lack of a copyright file (the licence is there, but not the
mandatory copyright file) seems a violation of
http://www.debian.org/doc/debian-policy/ch-docs.html#s-copyrightfile.
-- System Information:
Debian Release
Package: mydns-pgsql
Version: N/A; reported 2005-03-15
Severity: minor
Description says:
>It is primarily designed for organizations with many zones and/or
>resource records who desire the ability to perform real-time dynamic
>updates on their DNS data via MySQL.
The last reference should be to
On Wed, Mar 09, 2005 at 01:43:53PM +0100,
Torsten Landschoff <[EMAIL PROTECTED]> wrote
a message of 37 lines which said:
> Now that 2.1.30 is in sarge, could you test it? I still think it is
> a problem with the gcrypt/gnutls stuff.
Indeed, its seems to work now on the UltraSparc.
--
To UNS
Package: latex-beamer
Version: 3.01-1
Severity: normal
I can produce DVI files but I cannot see them with xdvi:
gs: Error: /undefined in pgfo
gs: Operand stack:
gs:
gs: Execution stack:
gs:%interp_exit .runexec2 --nostringval-- --nostringval--
--nostringval--
gs:2 %stopped_pu
tag 290830 upstream
quit
On Mon, Jan 17, 2005 at 02:38:58AM +0100,
Adrian Zaugg <[EMAIL PROTECTED]> wrote
a message of 56 lines which said:
> Fatal
> Root-servers list is not consistent with ICANN
Do note that you can change the test result for a simple warning. (In
default.profile, edit to
Package: dpkg
Version: 1.10.25
Severity: wishlist
Tags: patch
dpkg-source -i ignores files from many revision control systems but
not from darcs (http://www.darcs.net/, the Debian package is
"darcs"). Here is a small patch:
--- dpkg-source.pl.orig 2005-01-10 22:23:10.0 +0100
+++ dpkg-sou
101 - 166 of 166 matches
Mail list logo
