bad. :/
On the other hand, it make me glad to have yet another example of the
value of the less popular architectures. :)
Does this mean that running the autopkgtest scripts on hurd or kfreebsd
would expose these problems? Do we have any idea how widespread the
problem is?
--
Happy hacking
Petter Reinholdtsen
ave started one myself.
I assume other distributions are affected too. What are they doing to
handle it?
--
Happy hacking
Petter Reinholdtsen
value. It agrees that gethostid() have unclear
sematics. :)
--
Happy hacking
Petter Reinholdtsen
e manual page how
to set it for those that want to modify an existing installation.
--
Happy hacking
Petter Reinholdtsen
/2308ff3051ed55cc/packages.json | jq -r
'.Packages[]'
--
Happy hacking
Petter Reinholdtsen
ique and fixed across IP changes and
reboots.
[1] http://pubs.opengroup.org/onlinepubs/9699919799/functions/gethostid.html
--
Happy hacking
Petter Reinholdtsen
=/dev/urandom bs=1 count=4 of=/etc/hostid 2>/dev/null
fi
fi
We need to figure out how to transform the UUID to a 32 bit integer, of course.
--
Happy hacking
Petter Reinholdtsen
file a pu reqest right
now like the one I got in #706762, and use it to coordinate the upload
to stable. I have not done it myself for a long time, so I am not
quite sure, but one of the release managers said it was the correct
first step.
--
Happy hacking
Petter Reinholdtsen
--
To UNSUBSCRIBE,
will make it into Wheezy soon. :)
--
Happy hacking
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to debian-glibc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20130513231413.gn13...@ulrik.uio.no
ible to look up the local host name
and always get a working IP address, return an IPv6 address that can't
be used to ping, connect or pass on.
--
Happy hacking
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to debian-glibc-requ...@lists.debian.org
with a subject of "unsubscribe&q
sist
http://repo.or.cz/w/glibc.git/blob/HEAD:/localedata/locales/nb_NO >,
I see that this too is lacking correct transliteration for æøå.
Do you have an idea how I can get correct transliteration for æøå into
eglibc?
--
Happy hacking
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to debia
6 +127,21 @@
translit_start
include "translit_combining";""
+
+% Norwegian transliteration
+% LATIN CAPITAL LETTER A WITH RING ABOVE -> "Aa"
+ "";""
+% LATIN SMALL LETTER A WITH RING ABOVE -> "aa"
+ "";""
e? Seem a bit overkill to create a new package
for a simple perl script, but it might be the best way to do this.
Happy hacking,
--
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to debian-glibc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@
PAM and NSS out of the box. :)
Happy hacking,
--
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to debian-glibc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20100529155840.gc5...@login1.uio.no
d into Debian, but its WNPP
request (#579593) is still not acted on. Besides, I have problems
getting sssd to work, but suspect that is only because I do not yet
know how to configure it properly.
Happy hacking,
--
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to debian-glibc-requ...@lists.debi
[Petter Reinholdtsen]
> If increased timeout is left out from the default configuration,
> please implement support for sourcing /etc/default/nscd and allow a
> policy compliant mechanism to replace the configuration file used by
> nscd.
Something like this would make that possible
ffline operation, I suspect the "reload-count unlimited" setting
is the most important one.
If increased timeout is left out from the default configuration,
please implement support for sourcing /etc/default/nscd and allow a
policy compliant mechanism to replace the configuration file
ect to LDAP, and never disappears when the
laptop is unable to connect to LDAP.
I do not understand all aspects of the setting for roaming laptops, so
I very much welcome input on my proposed settings, which were based on
a working setup for disconnected operation.
Happy hacking,
--
Petter Reinho
ces211
check-files servicesyes
Happy hacking,
--
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to debian-glibc-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.or
d-Stop: $remote_fs
+# Required-Start:$remote_fs $syslog
+# Required-Stop: $remote_fs $syslog
# Default-Start: 2 3 4 5
# Default-Stop: 0 1 6
# Short-Description: Starts the Name Service Cache Daemon
Happy hacking,
--
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to debian-glib
[Petter Reinholdtsen]
> One approach would be to change the default configuration in
> /etc/nscd.conf to use a longer timeout for the cache values. For
> example these values to get 30 days timeout:
>
> positive-time-to-live passwd 2592000
> positive-time-to-live group 25
ile.
Happy hacking,
--
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
ote_fs
+# Required-Stop: $remote_fs
# Default-Start: 2 3 4 5
# Default-Stop: 0 1 6
# Short-Description: Starts the Name Service Cache Daemon
Happy hacking,
--
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[Clint Adams]
> Please try the attached file and see if that does what you want.
It does. I tested, and both the start, stop and restart functions
worked as I want. :)
I did not test any error reporting.
Happy hacking,
--
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to [EMAIL PROTEC
/lsb/init-functions to feed the text into the
graphical boot screen, and thus depend on the use of the lsb output
functions.
Please change the init.d script to use the lsb output functions.
Happy hacking,
--
Petter Reinholdtsen
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of
Package: libc6
Version: 2.3.5-5
Severity: serious
The script /etc/init.d/glibc.sh can not be sourced, as it contain
'exit 0' at the end of the script. This is against policy, specifying
that all .sh scripts in /etc/rcS.d/ will be sourced. I discovered
this while fixing sysv-rc (bug #339955), as
Package: nscd1
Version: 2.3.2.ds1-22
Severity: wishlist
To be able to check boot script order, and also to be able to start
boot scripts in parallel, it is important to know the dependencies of
the various boot scripts. The Linux Software Base specifies a init.d
header file format useful for t
Package: libc6
Version: 2.3.5-7
Severity: wishlist
To be able to check boot script order, and also to be able to start
boot scripts in parallel, it is important to know the dependencies of
the various boot scripts. The Linux Software Base specifies a init.d
header file format useful for this p
This problem affect clanlib as well. I'm trying to prepare an NMU to
get it to use the new C++ ABI, and ran into this compile error:
Compiling Sources/Display/Input/X11/joystick_linux.cpp
In file included from Sources/Display/Input/X11/joystick_linux.h:45,
from Sources/Dis
[Denis Barbier]
> I will be very glad to share this presentation with anyone involved
> in glibc locales, please drop me a note if you are interested.
I'm interested. Should we tell this to libc-locales@ as well, in case
someone there want to join?
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED
Followup-For: Bug #231748
Package: libc6-dev
Version: 2.3.2.ds1-20
I ran into the same problem when trying to compile mapserver with
-pedantic using gcc 3.3.5. Not sure how to solve the problem, but
this bug make it hard to convince the mapserver developers to use
-pendantic. I want to convince
[GOTO Masanori]
> I got the report from Pasi, he said [EMAIL PROTECTED] locale could be choosed
> via (I guess) language-chooser (or country-chooser?) in
> debian-installer. However, debian glibc locales packages does not
> have [EMAIL PROTECTED] So when he upgraded locales package, installation
Package: libc6
Version: 2.3.2.ds1-16
Tags: patch
Severity: wishlist
The current DNS resolver in glibc only read the content of
/etc/resolv.conf when a program starts (or during the first call to
one of the resolver function). This is a problem for long-lasting
programs on for example lapto
A patch to fix it is available from
http://sources.redhat.com/ml/libc-hacker/2004-07/msg00042.html>.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
A patch to fix it is available from
http://sources.redhat.com/ml/libc-hacker/2004-07/msg00042.html>.
This problem is now reported upstream into glibc bugzilla as
http://sources.redhat.com/bugzilla/show_bug.cgi?id=274 >.
This problem is now reported upstream into glibc bugzilla as
http://sources.redhat.com/bugzilla/show_bug.cgi?id=274 >.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
s to end up as 42.0, not 0.0. Am I
mistaken?
Here is the source.
/*
* Demonstrate bug in strtold() in glibc. The last two number strings
* converts to 0, not 42 as it should.
* Based on code from Ivano Primi and modified by Petter Reinholdtsen.
*/
#include
#include
#include
extern int errno;
ex
up as 42.0, not 0.0. Am I
mistaken?
Here is the source.
/*
* Demonstrate bug in strtold() in glibc. The last two number strings
* converts to 0, not 42 as it should.
* Based on code from Ivano Primi and modified by Petter Reinholdtsen.
*/
#include
#include
#include
extern int errno;
ex
[GOTO Masanori]
> At Mon, 10 May 2004 22:33:51 +0200,
> Sebastien Bacher wrote:
> > $ grep ^thous /usr/share/i18n/locales/fr_FR
> > thousands_sep ""
> >
> > But the french separator for thousands is a space and not empty
>
> Petter, Denis, should this be fixed? Note that glibc uses f
[GOTO Masanori]
> At Mon, 10 May 2004 22:33:51 +0200,
> Sebastien Bacher wrote:
> > $ grep ^thous /usr/share/i18n/locales/fr_FR
> > thousands_sep ""
> >
> > But the french separator for thousands is a space and not empty
>
> Petter, Denis, should this be fixed? Note that glibc uses f
[Denis Barbier]
> to locale.alias, but IIRC Petter would like real locales and not
> aliases because otherwise gettext won't process no.po files.
The missing no_NO ISO-8859-1 is a bug. I want to have both no_NO and
nb_NO listed while we complete the move to nb_NO.
% grep n._NO /usr/share/i18n/
[GOTO Masanori]
>> glibc CVS together with the files in localedata/.
>
> What does this mean?
I am thinking about the script used to generate 11_cvs_locales.dpatch.
It is probably a good idea include locale/iso-3166.def as well.
[Denis Barbier]
> to locale.alias, but IIRC Petter would like real locales and not
> aliases because otherwise gettext won't process no.po files.
The missing no_NO ISO-8859-1 is a bug. I want to have both no_NO and
nb_NO listed while we complete the move to nb_NO.
% grep n._NO /usr/share/i18n/
[GOTO Masanori]
>> glibc CVS together with the files in localedata/.
>
> What does this mean?
I am thinking about the script used to generate 11_cvs_locales.dpatch.
It is probably a good idea include locale/iso-3166.def as well.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "
It is probably a good idea to update locale/iso-3166.def from the
glibc CVS together with the files in localedata/.
It is probably a good idea to update locale/iso-3166.def from the
glibc CVS together with the files in localedata/.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[Denis Barbier]
> The attached dpatch was used to close this bug. But its comment is not
> accurate, this dpatch adds a et_EE.ISO-8859-15 locale (and is then
> identical to upstream fix) whereas bug log and dpatch comment seem
> to imply that et_EE charset was changed from ISO-8859-1 to ISO-8859-
[Denis Barbier]
> The attached dpatch was used to close this bug. But its comment is not
> accurate, this dpatch adds a et_EE.ISO-8859-15 locale (and is then
> identical to upstream fix) whereas bug log and dpatch comment seem
> to imply that et_EE charset was changed from ISO-8859-1 to ISO-8859-
tags 207199 + patch
thanks
Here is a patch to fix this problem. It teaches the script to ignore
empty lines, and to write an error if the line don't have at least two
words.
This is the output when I tested it
Generating locales...
error: Bad entry 'en_US.ISO-8859-1 '
error: Bad entry 'e
tags 207199 + patch
thanks
Here is a patch to fix this problem. It teaches the script to ignore
empty lines, and to write an error if the line don't have at least two
words.
This is the output when I tested it
Generating locales...
error: Bad entry 'en_US.ISO-8859-1 '
error: Bad entry 'e
rfmon() formatting
# DP: Related bugs: @226047
# DP: Author: Petter Reinholdtsen
# DP: Upstream status: In CVS
# DP: Status Details:
# DP: Date: 2004-03-21
# DP: 2004-01-01 Petter Reinholdtsen <[EMAIL PROTECTED]>
# DP:
# DP: * stdlib/strfmon.c: Make formatting of left-justified currency
# DP: v
Here is more info on the problem.
The issue is that 'struct user' in on s390 need 'struct
user_regs_struct' from , and this need type
's390_fp_regs' which need type 'freq_t' which need type '__u64' which
is unavailable when compiling with -ansi. The reason the '__u64' type
is mising is that ANS
I had a look at this on raptor.debian.org (dchroot unstable). The
triggering code seem to be in in /usr/include/asm/types.h:
#ifndef __s390x__
#if defined(__GNUC__) && !defined(__STRICT_ANSI__)
typedef __signed__ long long __s64;
typedef unsigned long long __u64;
#endif
#else /* __s3
Here is more info on the problem.
The issue is that 'struct user' in on s390 need 'struct
user_regs_struct' from , and this need type
's390_fp_regs' which need type 'freq_t' which need type '__u64' which
is unavailable when compiling with -ansi. The reason the '__u64' type
is mising is that ANS
I had a look at this on raptor.debian.org (dchroot unstable). The
triggering code seem to be in in /usr/include/asm/types.h:
#ifndef __s390x__
#if defined(__GNUC__) && !defined(__STRICT_ANSI__)
typedef __signed__ long long __s64;
typedef unsigned long long __u64;
#endif
#else /* __s3
Package: libc6
Version: 2.3.2.ds1-10
Severity: normal
The current strfmon() in glibc do not format currency correctly and in
accordence with the POSIX standard. This has been fixed in glibc CVS.
These are the relevant patches:
(libc/)
2004-01-01 Petter Reinholdtsen <[EMAIL PROTEC
Package: libc6
Version: 2.3.2.ds1-10
Severity: normal
The current strfmon() in glibc do not format currency correctly and in
accordence with the POSIX standard. This has been fixed in glibc CVS.
These are the relevant patches:
(libc/)
2004-01-01 Petter Reinholdtsen <[EMAIL PROTEC
This patch was commited to the glibc CVS 2003-11-04. A newer version
of the glibc source should include the fix.
This patch was commited to the glibc CVS 2003-11-04. A newer version
of the glibc source should include the fix.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
to
Indrek, and Indrek verifies the change. Here is a patch to do the
change.
Another alternative is to add '[EMAIL PROTECTED]/ISO-8859-15 to the SUPPORTED
file. I'm not sure if that is better then just changing the charset.
2004-01-01 Petter Reinholdtsen <[EMAIL PROTECTED]>
to
Indrek, and Indrek verifies the change. Here is a patch to do the
change.
Another alternative is to add '[EMAIL PROTECTED]/ISO-8859-15 to the SUPPORTED
file. I'm not sure if that is better then just changing the charset.
2004-01-01 Petter Reinholdtsen <[EMAIL PROTECTED]>
The patch for week- and workday in fi_FI was commited to upstream CVS
2003-11-26.
This problem should thus be fixed in the next release of glibc.
The patch for week- and workday in fi_FI was commited to upstream CVS
2003-11-26.
This problem should thus be fixed in the next release of glibc.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
The nb_NO locale was just added to the glibc CVS. This is good. :)
Unfortunately, it was added by renaming the no_NO locale. Renaming
the locale is the most painful way to fix this, as this will make all
translations using the 'no' language code inaccessable by gettext.
Because of this, I urge
When I test the same using woody with locales version 2.2.5-11.5, it
work as it should.
minerva:~# grep en_US /etc/locale.gen
en_US UTF-8
minerva:~# locale-gen
Generating locales...
[...]
en_US.UTF-8... done
[...]
Generation complete.
minerva:~# echo '[' | LANG=en_US.UTF-8
Package: locales
Version: 2.3.2-9
Severity: minor
Tags: patch
I forgot a minor thing in my patch to add nb_NO as a real locale. It
should also be listed in the SUPPORTED file, to make it possible to
select it at install time.
Here is a patch to fix it.
--- SUPPORTED.orig 2003-10-31
This bug is fixed in upstream CVS. This is the changelog entry from
libc/localedata/:
2003-09-21 Jordi Mallach <[EMAIL PROTECTED]>
* locales/ca_ES: Fix mon_grouping, int_frac_digits and frac_digits
values.
* locales/eu_ES: Fix int_frac_digits and frac_digits values.
[Harry Edmon]
> It interprets it as a host name (see the code for __ivaliduser).
I checked the current glibc source for this function, and was able to
locate it in inet/rcmd.c:
/*
* XXX
* Don't make static, used by lpd(8).
*
* This function is not used anymore. It is only present because lp
Package: libc6-pic
Version: 2.3.2-3
Severity: serious
After upgrading my Sid chroot, the cdrom floppy built by d-i is unable
to boot properly.
These are the last messages printed when booting the floppy:
RAMDISK: Compressed image found at block 0
Freeing initrd memory: 1441k freed
VFS: Mo
Package: locales
Version: 2.3.2-3
Severity: wishlist
Tags: patch
The locale for Norwegian Bokmål should be nb_NO, and translation files
and programs should use this instead of the old and inaccurate no_NO.
A patch to fix this is available from
http://sources.redhat.com/ml/libc-alpha/2003-0
A locale for Aragonese was added as an_ES to the glibc CVS 2003-06-16.
It will be included in the next release of glibc.
A locale for Aragonese was added as an_ES to the glibc CVS 2003-06-16.
It will be included in the next release of glibc.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[Jeff Bailey]
> Please provide a testcase to show that this bug still occurs.
The problem still exists:
The letter ø (Octal 370) is not a valid UTF-8 sequence:
% echo -e 'hei \370 foo' | /usr/bin/iconv -f utf-8 -t iso-8859-1//TRANSLIT
hei /usr/bin/iconv: illegal input sequence at position
The abday values for de_DE is changed in the latest CVS version of
glibc. CVS was changed 2003-05-06, and the corrected version should
be included in the next glibc version.
The bug is present in Woody, glibc 2.2.5-11.2, and is fixed in Sid,
glibc 2.3.1-16. I'm not sure which version was the first fixed
version.
The locales ar_SD and sr_YU (and es_EC and [EMAIL PROTECTED]) are now
fixed in the GNU libc CVS. This bug should be fixed in the next
release of glibc (the one after v2.3.3).
This is the changelog entries from localedata/ChangeLog:
2003-04-15 Petter Reinholdtsen <[EMAIL PROTEC
The glibc developer Ulrich Drepper claims that this security problem
is fixed in later versions of glibc. See
http://sources.redhat.com/ml/libc-alpha/2003-04/msg00124.html>.
I'm not sure in which version this was fixed.
Package: libc6
Version: 2.2.5-11.2
Severity: wishlist
On solaris, it is possible to use the 'compat' mode in NSS with LDAP
directories. Now that the libnss-ldap package start to support
netgroups, the compat module should support using netgroups and user
info from LDAP. See
http://www.sunmanage
It is safer to test using 'date', as this do not require external
translation files.
# locale-gen
Generating locales...
en_US.ISO-8859-1... done
en_US.UTF-8... done
es_ES.ISO-8859-1... done
[EMAIL PROTECTED] done
Generation complete.
# date; LC_ALL=en_US date; LC_ALL=es_E
The problem with 'locale -a' listing nonexisting locales was fixed in
upstream CVS 2002-08-26. It made into release 2.3.1.
[David B Harris]
> Hey ho :) Following log should cover it, I think:
> Setting up locales (2.2.5-7) ...
> Generating locales...
> en_CA.ISO8859-1.../usr/share/i18n/locales/en_CA:27:
> LC_IDENTIFICATION: unknown character in field `address'
> done
> Generation complete.
It works if I use 'e
It is safer to test using 'date', as this do not require external
translation files.
# locale-gen
Generating locales...
en_US.ISO-8859-1... done
en_US.UTF-8... done
es_ES.ISO-8859-1... done
[EMAIL PROTECTED] done
Generation complete.
# date; LC_ALL=en_US date; LC_ALL=es_E
The problem with 'locale -a' listing nonexisting locales was fixed in
upstream CVS 2002-08-26. It made into release 2.3.1.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[David B Harris]
> Hey ho :) Following log should cover it, I think:
> Setting up locales (2.2.5-7) ...
> Generating locales...
> en_CA.ISO8859-1.../usr/share/i18n/locales/en_CA:27:
> LC_IDENTIFICATION: unknown character in field `address'
> done
> Generation complete.
It works if I use 'e
Package: locales
Version: 2.2.5-14
At the moment it is impossible to preconfigure fthe locales package
for the first time install.
I would like to supply a debconf value in the
locales/locales_to_be_generated variable, and get the locales package
to create /etc/locale.gen based on the content of
Package: locales
Version: 2.2.5-14
At the moment it is impossible to preconfigure fthe locales package
for the first time install.
I would like to supply a debconf value in the
locales/locales_to_be_generated variable, and get the locales package
to create /etc/locale.gen based on the content of
87 matches
Mail list logo
