New package dbh Disk based hash library
New package xfce-mcs-manager Multi channel settings manager
New package xfce-mcs-plugins Plugins for multi channel settings manager
New package xfce-utils Utilities for the XFce Desktop Environment
New package xfce4-panel Next generation panel for xfce
New package xfdesktop Desktop manager for the XFce Desktop Environment
New package xffm Next generation filemanager and SMB network navigator for XFce4
New package xffm-icons Icons for xffm file manager
New package xfwm4 Next generation window manager for xfce
New package xfwm4-themes Additionnal themes for xfwm4
Removed package raidtools
Updated Packages:
Omni-0.9.1-1 ------------ * Tue Jan 06 2004 Tim Waugh twaugh@redhat.com 0.9.1-1
- 0.9.1. - No longer need pkgdatadir, acinit or validate patches.
* Thu Nov 27 2003 Tim Waugh twaugh@redhat.com
- Build requires glib-devel (bug #110745).
authconfig-4.6-1 ---------------- * Thu Jan 08 2004 Nalin Dahyabhai nalin@redhat.com 4.6-1
- authconfig-gtk.py: require rhpl, which is required by the script (#104209) - both: require usermode (authconfig-gtk transitively), else leave a dangling symlink (#104209) - the great redhat-config-authentication/system-config-authentication renaming, as was foretold in the fedora-config-list archives
* Wed Jan 07 2004 Nalin Dahyabhai nalin@redhat.com
- preserve "compat" if it's used in /etc/nsswitch.conf
checkpolicy-1.4-2 ----------------- * Wed Jan 07 2004 Dan Walsh dwalsh@redhat.com 1.4-2
- Remove check for roles transition
cvs-1.11.11-1 ------------- * Wed Jan 07 2004 Nalin Dahyabhai nalin@redhat.com 1.11.11-1
- turn kserver, which people shouldn't use any more, back on
* Tue Dec 30 2003 Nalin Dahyabhai nalin@redhat.com
- update to 1.11.11
* Thu Dec 18 2003 Nalin Dahyabhai nalin@redhat.com 1.11.10-1
- update to 1.11.10
cyrus-sasl-2.1.17-1 ------------------- * Wed Jan 07 2004 Nalin Dahyabhai nalin@redhat.com 2.1.17-1
- forcibly disable otp and sql plugins at compile-time
* Fri Dec 19 2003 Nalin Dahyabhai nalin@redhat.com
- update to 2.1.17, forcing the gssapi plugin to be shared now, as before - use a bundled libdb (#112215) - build static-with-all-plugins and normal-shared libsasl versions - add sasl2-{shared,static}-mechlist for very basic sanity checking - make inclusion of sasl1 stuffs conditional, because it's so going away
* Sat Dec 13 2003 Jeff Johnson jbj@jbj.org 2.1.15-7
- rebuild against db-4.2.52.
diffutils-2.8.1-10 ------------------ * Thu Jan 08 2004 Tim Waugh twaugh@redhat.com 2.8.1-10
- Fix mistaken use of '|' instead of '||'.
e2fsprogs-1.35-4 ---------------- * Thu Jan 08 2004 Florian La Roche Florian.LaRoche@redhat.de
- add patch from Dave Jones
* Sun Dec 14 2003 Florian La Roche Florian.LaRoche@redhat.de
- update to 1.35-WIP-1207
httpd-2.0.48-5 -------------- * Sun Jan 04 2004 Joe Orton jorton@redhat.com 2.0.48-5
- use graceful restart in logrotate - bump default MaxRequestsPerChild for prefork to 4000 - move vendor string for Server header into spec file - include mod_status extension hook and use it in mod_ssl to include SSL session cache statistics in server-status output.
iptables-1.2.9-1.1 ------------------ * Wed Jan 07 2004 Thomas Woerner twoerner@redhat.com 1.2.9-1.1
- rebuild
jwhois-3.2.2-2 -------------- * Wed Jan 07 2004 Nalin Dahyabhai nalin@redhat.com 3.2.2-2
- fix typos in jwhois.conf (#113012)
libgnomeprint22-2.4.2-1.1 ------------------------- * Wed Jan 07 2004 Owen Taylor otaylor@redhat.com 2.4.2-1.1
- Upgrade to 2.4.2
mtools-3.9.9-7 -------------- * Thu Jan 08 2004 Tim Waugh twaugh@redhat.com 3.9.9-7
- Fix mistaken use of '&' instead of '&&'.
ncpfs-2.2.3-2 ------------- * Wed Jan 07 2004 Florian La Roche Florian.LaRoche@redhat.de
- add patch from Dave Jones
netatalk-1.6.4-1 ---------------- * Wed Jan 07 2004 Florian La Roche Florian.LaRoche@redhat.de
- 1.6.4
* Fri Aug 01 2003 Charlie Bennett ccb@redhat.com
- Update with 1.6.3 upstream sources
nkf-2.04-1 ---------- * Wed Jan 07 2004 Akira TAGOH tagoh@redhat.com 2.04-1
- New upstream release.
openldap-2.1.25-1 ----------------- * Thu Jan 08 2004 Nalin Dahyabhai nalin@redhat.com 2.1.25-1
- change logging facility used from daemon to local4 (#112730, reversing #11047) BEHAVIOR CHANGE - SHOULD BE MENTIONED IN THE RELEASE NOTES.
* Wed Jan 07 2004 Nalin Dahyabhai nalin@redhat.com
- incorporate fix for logic quasi-bug in slapd's SASL auxprop code (Dave Jones)
* Thu Dec 18 2003 Nalin Dahyabhai nalin@redhat.com
- update to 2.1.25, now marked STABLE
policy-1.4.4-2 -------------- * Wed Jan 07 2004 Dan Walsh dwalsh@redhat.com 1.4.4-2
- Add run daemon patch
* Mon Jan 05 2004 Dan Walsh dwalsh@redhat.com 1.4.4-1
- Many changes added, access to random device - add directories ownership
* Tue Dec 23 2003 Dan Walsh dwalsh@redhat.com 1.4.2-2
- Fix bluetooth
policycoreutils-1.4-3 --------------------- * Wed Jan 07 2004 Dan Walsh dwalsh@redhat.com 1.4-3
- fix command parsing restorecon
rpmdb-fedora-1.90-0.20040108 ----------------------------
sed-4.0.8-3 ----------- * Wed Jan 07 2004 Jakub Jelinek jakub@redhat.com 4.0.8-3
- if not -n, print current buffer after N command on the last line unless POSIXLY_CORRECT (#112952) - adjust XFAIL_TESTS for the improved glibc regex implementation (#112642)
shadow-utils-4.0.3-17 --------------------- * Wed Jan 07 2004 Nalin Dahyabhai nalin@redhat.com 4.0.3-17
- fix use of uninitialized memory in useradd (#89145)
spamassassin-2.61-2 ------------------- * Wed Dec 31 2003 Dan Walsh dwalsh@redhat.com 2.61-2
- Change sysconfdir to not use full path
system-config-language-1.1.3-1 ------------------------------ * Wed Jan 07 2004 Than Ngo than@redhat.com 1.1.3-1
- make changes for Python2.3
Once upon a time, Build System buildsys@redhat.com said:
cyrus-sasl-2.1.17-1
- Fri Dec 19 2003 Nalin Dahyabhai nalin@redhat.com
- use a bundled libdb (#112215)
Question: why add to the list of packages with a private copy of libdb? If the system libdb is broken, it should be fixed, right? I'd like to see _fewer_ (or even better, zero) packages with their own private copy of normally shared libraries (especially libdb).
On Thu, Jan 08, 2004 at 08:54:44AM -0600, Chris Adams wrote:
Once upon a time, Build System buildsys@redhat.com said:
cyrus-sasl-2.1.17-1
- Fri Dec 19 2003 Nalin Dahyabhai nalin@redhat.com
- use a bundled libdb (#112215)
Question: why add to the list of packages with a private copy of libdb? If the system libdb is broken, it should be fixed, right? I'd like to see _fewer_ (or even better, zero) packages with their own private copy of normally shared libraries (especially libdb).
You have to watch out that shared libraries don't mix. Consider an application which links with libdb-4.2.so. If libsasl2 links against libdb-4.1.so (even if it's just because it hasn't been recompiled to use libdb-4.2.so), and the application uses libsasl2, you can wind up with more than one version of the library loaded, *providing an overlapping set of symbols*. The situation is the same if the application uses an older version of libdb than libsasl2 does. Mismatches break apps.
It gets scarier when you consider something like nss_db, where if you're using it, every application on the system could end up pulling in whatever shared libdb it links with, even applications which were built on earlier releases using an older version of that same library.
Newer versions of Berkeley DB have a compile-time argument you can use to build a static library which won't conflict in this way with a library which wasn't built with that switch. Using it adds a measure of safety. Because the sasldb code in libsasl2 is the only user of the one file which it reads and writes, possible version skew versus the system copy of libdb doesn't create a problem here.
Cheers,
Nalin
cmadams@hiwaay.net (Chris Adams) writes:
Question: why add to the list of packages with a private copy of libdb? If the system libdb is broken, it should be fixed, right?
db4 is broken since RHL9, the fix is easy but nobody did the correction yet. Instead of, it is suggested that every package ships and maintains an internal copy of db4.
Enrico https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=91933
On Thu, Jan 08, 2004 at 08:58:46AM -0500, Build System wrote:
openldap-2.1.25-1
- Thu Jan 08 2004 Nalin Dahyabhai nalin@redhat.com 2.1.25-1
- change logging facility used from daemon to local4 (#112730, reversing #11047)
Rationale? OpenLDAP logging to local4 is IMHO fundamentally broken, as it was outlined in 11047. Why is this being reverted?
Regards, Daniel
On Thu, Jan 08, 2004 at 04:01:55PM +0100, Daniel Roesen wrote:
On Thu, Jan 08, 2004 at 08:58:46AM -0500, Build System wrote:
openldap-2.1.25-1
- Thu Jan 08 2004 Nalin Dahyabhai nalin@redhat.com 2.1.25-1
- change logging facility used from daemon to local4 (#112730, reversing #11047)
Rationale? OpenLDAP logging to local4 is IMHO fundamentally broken, as it was outlined in 11047. Why is this being reverted?
a. It surprises people who read official OpenLDAP documentation. b. It's run-time configurable now, so you can set SLAPD_OPTIONS in /etc/sysconfig/ldap and forget about the whole thing.
Nalin
Build System buildsys@redhat.com writes:
Removed package raidtools
Won't this break RAID devices mounted during boot (I didn't spot a change to initscripts)?
Alex Kiernan wrote:
Build System buildsys@redhat.com writes:
Removed package raidtools
Won't this break RAID devices mounted during boot (I didn't spot a change to initscripts)?
I assume this means mdadm will finally become the standard in RH. This is very welcome news to me.
Warren
On Sun, 11 Jan 2004, Warren Togami wrote:
I assume this means mdadm will finally become the standard in RH. This is very welcome news to me.
I haven't had time to look into my problems very much yet, but I'm not very enthusiastic about mdadm on non-x86. On Linux/x86 it works okay, but on Linux/ultrasparc I'm having major issues with the two arrays I converted from raidtools for grins. It keeps destroying the disk labels on some of the disks....
later, chris
On Mon, 2004-01-12 at 08:58, Chris Ricker wrote:
On Sun, 11 Jan 2004, Warren Togami wrote:
I assume this means mdadm will finally become the standard in RH. This is very welcome news to me.
I haven't had time to look into my problems very much yet, but I'm not very enthusiastic about mdadm on non-x86. On Linux/x86 it works okay, but on Linux/ultrasparc I'm having major issues with the two arrays I converted from raidtools for grins. It keeps destroying the disk labels on some of the disks....
Have you reported the problems to the linux-raid list and/or neil brown? He seems fairly receptive to problem reports about mdadm.
-sv
On Mon, 2004-01-12 at 07:58, Chris Ricker wrote:
On Sun, 11 Jan 2004, Warren Togami wrote:
I assume this means mdadm will finally become the standard in RH. This is very welcome news to me.
I haven't had time to look into my problems very much yet, but I'm not very enthusiastic about mdadm on non-x86. On Linux/x86 it works okay, but on Linux/ultrasparc I'm having major issues with the two arrays I converted from raidtools for grins. It keeps destroying the disk labels on some of the disks....
Dunno if it helps, but this is the conversion process that Cal documented from a while back:
https://lists.dulug.duke.edu/pipermail/aurora-sparc-devel/2002-May/000784.ht...
mdadm should work properly on sparc, neil is well aware of the "special" needs of that arch. bugs should definitely go in his direction if you find any.
~spot --- Tom "spot" Callaway <tcallawa(a)redhat*com> LCA, RHCE Red Hat Sales Engineer || Aurora SPARC Linux Project Leader
"The author's mathematical treatment of the conception of purpose is novel and highly ingenious, but heretical and, so far as the present social order is concerned, dangerous and potentially subversive. Not to be published." -- Aldous Huxley's "Brave New World"
On Mon, 12 Jan 2004, Tom 'spot' Callaway wrote:
Dunno if it helps, but this is the conversion process that Cal documented from a while back:
https://lists.dulug.duke.edu/pipermail/aurora-sparc-devel/2002-May/000784.ht...
mdadm should work properly on sparc, neil is well aware of the "special" needs of that arch. bugs should definitely go in his direction if you find any.
The conversion went fine. My problem is that the VTOC was getting corrupted on some of the disks, so every time I rebooted I had to relabel before my arrays would come up.... I'm still trying to hunt down why / how it's getting corrupted
later, chris
On Sun, 2004-01-11 at 01:55 -1000, Warren Togami wrote:
Alex Kiernan wrote:
Build System buildsys@redhat.com writes:
Removed package raidtools
Won't this break RAID devices mounted during boot (I didn't spot a change to initscripts)?
This was done prematurely and it's been undeprecated (for now)
I assume this means mdadm will finally become the standard in RH. This is very welcome news to me.
The plan is to move to doing this, but until the code exists across the board to do so (especially installer and initscrips), removing raidtools is premature.
Jeremy