New package mediawiki-imagemap The ImageMap extension for MediaWiki New package php-laconica PHP tool for microblogging New package vldocking A Java ™ docking system for JFC Swing applications Updated Packages:
NetworkManager-pptp-0.7.0-0.11.svn4027.fc10 ------------------------------------------- * Sun Oct 5 18:00:00 2008 Lubomir Rintel lkundrak@v3.sk 1:0.7.0-11.svn4027 - Add pptp dependency (#465644)
abcm2ps-5.9.1-1.fc10 -------------------- * Sun Oct 5 18:00:00 2008 Gerard Milmeister gemi@bluewin.ch - 5.9.1-1 - new release 5.9.1
expendable-0.0.4-2.fc10 ----------------------- * Sun Oct 5 18:00:00 2008 Tim Waugh twaugh@redhat.com 0.0.4-2 - 0.0.4. No longer need packing patch.
fakechroot-2.8-15.fc10 ---------------------- * Sun Oct 5 18:00:00 2008 Axel Thimm Axel.Thimm@ATrpms.net - 2.8-15 - Fix getpeername/getsockname socklen initialization.
fakeroot-1.9.7-18.fc10 ---------------------- * Fri Oct 3 18:00:00 2008 Axel Thimm Axel.Thimm@ATrpms.net - 1.9.7-18 - Update to 1.9.7.
freedink-1.08.20080920-4.fc10 ----------------------------- * Sun Oct 5 18:00:00 2008 Sylvain Beucler beuc@beuc.net - 1.08.20080920-4 - Use liberation-fonts in all distro versions
glibmm24-2.18.0-3.fc10 ---------------------- * Sun Oct 5 18:00:00 2008 Adel Gadllah adel.gadllah@gmail.com - 2.18.0-3 - Patch error.h directly rather than error.hg
* Sun Oct 5 18:00:00 2008 Adel Gadllah adel.gadllah@gmail.com - 2.18.0-2 - Backport upstream fix that resolves HOST_NOT_FOUND symbol conflicts (GNOME #529496)
gstreamer-plugins-good-0.10.10-5.fc10 ------------------------------------- * Sun Oct 5 18:00:00 2008 Hans de Goede hdegoede@redhat.com 0.10.10-5 - Fix gst-plugins-good-0.10.9-libv4l.patch to also patch config.in and Makefile.in so that the libv4l code actually gets enabled for real this time and remove plenty of egg from face for not fixing this properly in 0.10.10-2 (rh465599) - Explicitly disable pulse plugin so the spec builds even if you have the pulse devel packages installed.
hunspell-ca-0.20081005-1.fc10 ----------------------------- * Sun Oct 5 18:00:00 2008 Caolan McNamara caolanm@redhat.com - 0.20081005-1 - latest version
ibus-0.1.1.20081005-1.fc10 -------------------------- * Sun Oct 5 18:00:00 2008 Huang Peng shawn.p.huang@gmail.com - 0.1.1.20081005-1 - Update to 0.1.1.20081005.
kernel-2.6.27-0.392.rc8.git7.fc10 --------------------------------- * Sun Oct 5 18:00:00 2008 Dave Airlie airlied@redhat.com - drm-fix-drm-mode-h.patch - fix drm mode .h header
libcanberra-0.10-1.fc10 ----------------------- * Mon Oct 6 18:00:00 2008 Lennart Poettering lpoetter@redhat.com 0.10-1 - New version
linkage-0.2.0-3.fc10 -------------------- * Sun Oct 5 18:00:00 2008 Adel Gadllah adel.gadllah@gmail.com 0.2.0-3 - Rebuild
ltsp-5.1.26-1.fc10 ------------------ * Sun Oct 5 18:00:00 2008 Warren Togami wtogami@redhat.com - 5.1.26-1 - Fix copying of mkinitrd config file within chroot
* Sun Oct 5 18:00:00 2008 Warren Togami wtogami@redhat.com - 5.1.25-1 - Bug #462228 Prevent bogus mkinitrd config from landing on systems who accidentally install ltsp-client for no good reason. (Thanks John Ellson) - Remove irrelevant plugins from package
mediawiki-1.13.2-41.fc10 ------------------------ * Sun Oct 5 18:00:00 2008 Axel Thimm Axel.Thimm@ATrpms.net - 1.13.2-41 - Update to 1.13.2.
obex-data-server-0.3.99-2.fc10 ------------------------------ * Mon Oct 6 18:00:00 2008 - Bastien Nocera bnocera@redhat.com - 0.3.99-2 - Update to rev 1998 - Use gdk-pixbuf instead of ImageMagick for BIP support
perl-DBIx-Class-0.08010-8.fc10 ------------------------------ * Sun Oct 5 18:00:00 2008 Chris Weyl cweyl@alumni.drew.edu 0.08010-8 - filter all prov/req from anything under _docdir - note we still filter perl(DBD::Multi), at least until review bug bz#465690 is completed... - ...and perl(DBD::Pg) will always be filtered
pulseaudio-0.9.13-1.fc10 ------------------------ * Mon Oct 6 18:00:00 2008 Lennart Poettering lpoetter@redhat.com 0.9.13-1 - New release
renrot-1.1-0.3.rc3.fc10 ----------------------- * Mon Oct 6 18:00:00 2008 Andy Shevchenko andy@smile.org.ua - 1.1-0.3.rc3 - update to 1.1rc3 - change License to Artistic 2.0 accordingly to mainstream - update URLs - require (optional) Image::Magick
revisor-2.1.2-2.fc10 -------------------- * Sun Oct 5 18:00:00 2008 Jeroen van Meeuwen <kanarip a fedoraunity.org> 2.1.2-2 - Fix anaconda removing splittree.py - Latest rebuild - Minor bugfixes (#344 pkgorder traceback) - Add SELinux Check
scim-1.4.7-33.fc10 ------------------ * Sun Oct 5 18:00:00 2008 Huang Peng phuang@redhat.com - 1.4.7-33 - Add ICON="/usr/share/scim/icons/trademark.png" in xinput-scim
sepostgresql-8.3.4-2.1077.fc10 ------------------------------
slim-1.3.0-7.fc10 ----------------- * Sun Oct 5 18:00:00 2008 Anders F Bjorklund afb@users.sourceforge.net 1.3.0-7 - add compat req (#465631)
smart-1.1-56.fc10 ----------------- * Sun Oct 5 18:00:00 2008 Axel Thimm Axel.Thimm@ATrpms.net - 1.1-56 - Update to 1.1.
xastir-1.9.4-4.fc10 ------------------- * Sun Oct 5 18:00:00 2008 Lucian Langa cooly@gnome.eu.org - 1.9.4-4 - misc cleanups
xdoclet-1.2.3-9.4.fc10 ---------------------- * Sun Oct 5 18:00:00 2008 Matt Wringe mwringe@redhat.com - 0:1.2.3-9.4 - Update xdoclet-ant.not-required.patch to apply with fuzz=0.
* Fri Aug 29 18:00:00 2008 Michael Schwendt mschwendt@fedoraproject.org - 0:1.2.3-9.3 - include /usr/share/doc/xdoclet-1.2.3 directory
Summary: Added Packages: 3 Removed Packages: 0 Modified Packages: 26 Broken deps for i386 ---------------------------------------------------------- asterisk-mobile-1.6.0-0.21.beta9.fc10.i386 requires libbluetooth.so.2 gnuradio-3.1.2-2.fc9.i386 requires libcppunit-1.12.so.0 gtkmozembedmm-1.4.2.cvs20060817-21.fc10.i386 requires gecko-libs = 0:1.9.0.1 lvm2-cluster-2.02.39-3.fc10.i386 requires libcman.so.2 lvm2-cluster-2.02.39-3.fc10.i386 requires libdlm.so.2 1:openoffice.org-langpack-ta_IN-3.0.0-9.2.fc10.i386 requires lohit-fonts-tamil publican-0.37-0.fc10.noarch requires lohit-fonts-tamil pyclutter-0.6.2-2.fc9.i386 requires libclutter-glx-0.6.so.0 pyclutter-cairo-0.6.2-2.fc9.i386 requires libclutter-cairo-0.6.so.0 pyclutter-cairo-0.6.2-2.fc9.i386 requires libclutter-glx-0.6.so.0 pyclutter-gst-0.6.2-2.fc9.i386 requires libclutter-gst-0.6.so.0 pyclutter-gst-0.6.2-2.fc9.i386 requires libclutter-glx-0.6.so.0 pyclutter-gtk-0.6.2-2.fc9.i386 requires libclutter-glx-0.6.so.0 pyclutter-gtk-0.6.2-2.fc9.i386 requires libclutter-gtk-0.6.so.0 ruby-rpm-1.2.3-4.fc9.i386 requires librpmdb-4.4.so ruby-rpm-1.2.3-4.fc9.i386 requires librpm-4.4.so ruby-rpm-1.2.3-4.fc9.i386 requires librpmio-4.4.so ruby-rpm-1.2.3-4.fc9.i386 requires librpmbuild-4.4.so syncevolution-0.7-3.fc10.i386 requires libcppunit-1.12.so.0
Broken deps for x86_64 ---------------------------------------------------------- asterisk-mobile-1.6.0-0.21.beta9.fc10.x86_64 requires libbluetooth.so.2()(64bit) gnuradio-3.1.2-2.fc9.i386 requires libcppunit-1.12.so.0 gnuradio-3.1.2-2.fc9.x86_64 requires libcppunit-1.12.so.0()(64bit) gtkmozembedmm-1.4.2.cvs20060817-21.fc10.i386 requires gecko-libs = 0:1.9.0.1 gtkmozembedmm-1.4.2.cvs20060817-21.fc10.x86_64 requires gecko-libs = 0:1.9.0.1 lvm2-cluster-2.02.39-3.fc10.x86_64 requires libcman.so.2()(64bit) lvm2-cluster-2.02.39-3.fc10.x86_64 requires libdlm.so.2()(64bit) pyclutter-0.6.2-2.fc9.x86_64 requires libclutter-glx-0.6.so.0()(64bit) pyclutter-cairo-0.6.2-2.fc9.x86_64 requires libclutter-glx-0.6.so.0()(64bit) pyclutter-cairo-0.6.2-2.fc9.x86_64 requires libclutter-cairo-0.6.so.0()(64bit) pyclutter-gst-0.6.2-2.fc9.x86_64 requires libclutter-glx-0.6.so.0()(64bit) pyclutter-gst-0.6.2-2.fc9.x86_64 requires libclutter-gst-0.6.so.0()(64bit) pyclutter-gtk-0.6.2-2.fc9.x86_64 requires libclutter-glx-0.6.so.0()(64bit) pyclutter-gtk-0.6.2-2.fc9.x86_64 requires libclutter-gtk-0.6.so.0()(64bit) ruby-rpm-1.2.3-4.fc9.x86_64 requires librpmio-4.4.so()(64bit) ruby-rpm-1.2.3-4.fc9.x86_64 requires librpm-4.4.so()(64bit) ruby-rpm-1.2.3-4.fc9.x86_64 requires librpmdb-4.4.so()(64bit) ruby-rpm-1.2.3-4.fc9.x86_64 requires librpmbuild-4.4.so()(64bit) syncevolution-0.7-3.fc10.x86_64 requires libcppunit-1.12.so.0()(64bit)
Broken deps for ppc ---------------------------------------------------------- asterisk-mobile-1.6.0-0.21.beta9.fc10.ppc requires libbluetooth.so.2 gnuradio-3.1.2-2.fc9.ppc requires libcppunit-1.12.so.0 gnuradio-3.1.2-2.fc9.ppc64 requires libcppunit-1.12.so.0()(64bit) gtkmozembedmm-1.4.2.cvs20060817-21.fc10.ppc requires gecko-libs = 0:1.9.0.1 gtkmozembedmm-1.4.2.cvs20060817-21.fc10.ppc64 requires gecko-libs = 0:1.9.0.1 lvm2-cluster-2.02.39-3.fc10.ppc requires libcman.so.2 lvm2-cluster-2.02.39-3.fc10.ppc requires libdlm.so.2 pyclutter-0.6.2-2.fc9.ppc requires libclutter-glx-0.6.so.0 pyclutter-cairo-0.6.2-2.fc9.ppc requires libclutter-cairo-0.6.so.0 pyclutter-cairo-0.6.2-2.fc9.ppc requires libclutter-glx-0.6.so.0 pyclutter-gst-0.6.2-2.fc9.ppc requires libclutter-gst-0.6.so.0 pyclutter-gst-0.6.2-2.fc9.ppc requires libclutter-glx-0.6.so.0 pyclutter-gtk-0.6.2-2.fc9.ppc requires libclutter-glx-0.6.so.0 pyclutter-gtk-0.6.2-2.fc9.ppc requires libclutter-gtk-0.6.so.0 ruby-rpm-1.2.3-4.fc9.ppc requires librpmdb-4.4.so ruby-rpm-1.2.3-4.fc9.ppc requires librpm-4.4.so ruby-rpm-1.2.3-4.fc9.ppc requires librpmio-4.4.so ruby-rpm-1.2.3-4.fc9.ppc requires librpmbuild-4.4.so stapitrace-1.0.0-12.20080622cvs_alpha.fc10.ppc requires libbfd-2.18.50.0.9-1.fc10.so stapitrace-1.0.0-12.20080622cvs_alpha.fc10.ppc requires libopcodes-2.18.50.0.9-1.fc10.so syncevolution-0.7-3.fc10.ppc requires libcppunit-1.12.so.0
Broken deps for ppc64 ---------------------------------------------------------- TurboGears-1.0.7-2.fc10.noarch requires python-decoratortools >= 0:1.4 asterisk-mobile-1.6.0-0.21.beta9.fc10.ppc64 requires libbluetooth.so.2()(64bit) gnuradio-3.1.2-2.fc9.ppc64 requires libcppunit-1.12.so.0()(64bit) gtkmozembedmm-1.4.2.cvs20060817-21.fc10.ppc64 requires gecko-libs = 0:1.9.0.1 livecd-tools-018-1.fc10.ppc64 requires yaboot lvm2-cluster-2.02.39-3.fc10.ppc64 requires libcman.so.2()(64bit) lvm2-cluster-2.02.39-3.fc10.ppc64 requires libdlm.so.2()(64bit) mock-0.9.11-1.fc10.noarch requires python-decoratortools pyclutter-0.6.2-2.fc9.ppc64 requires libclutter-glx-0.6.so.0()(64bit) pyclutter-cairo-0.6.2-2.fc9.ppc64 requires libclutter-glx-0.6.so.0()(64bit) pyclutter-cairo-0.6.2-2.fc9.ppc64 requires libclutter-cairo-0.6.so.0()(64bit) pyclutter-gst-0.6.2-2.fc9.ppc64 requires libclutter-glx-0.6.so.0()(64bit) pyclutter-gst-0.6.2-2.fc9.ppc64 requires libclutter-gst-0.6.so.0()(64bit) pyclutter-gtk-0.6.2-2.fc9.ppc64 requires libclutter-glx-0.6.so.0()(64bit) pyclutter-gtk-0.6.2-2.fc9.ppc64 requires libclutter-gtk-0.6.so.0()(64bit) python-peak-rules-0.5a1.dev-0.2569.fc10.noarch requires python-decoratortools >= 0:1.7 python-peak-util-addons-0.6-1.fc10.noarch requires python-decoratortools >= 0:1.5 python-peak-util-assembler-0.5-1.fc10.noarch requires python-decoratortools >= 0:1.2 rt3-3.8.1-1.fc10.noarch requires perl(Locale::Maketext::Fuzzy) ruby-rpm-1.2.3-4.fc9.ppc64 requires librpmio-4.4.so()(64bit) ruby-rpm-1.2.3-4.fc9.ppc64 requires librpm-4.4.so()(64bit) ruby-rpm-1.2.3-4.fc9.ppc64 requires librpmdb-4.4.so()(64bit) ruby-rpm-1.2.3-4.fc9.ppc64 requires librpmbuild-4.4.so()(64bit) syncevolution-0.7-3.fc10.ppc64 requires libcppunit-1.12.so.0()(64bit)
On Mon, 2008-10-06 at 11:51 +0000, Rawhide Report wrote:
Broken deps for ppc64
rt3-3.8.1-1.fc10.noarch requires perl(Locale::Maketext::Fuzzy)
What is going on here?
perl-Locale-Maketext-Fuzzy is a noarch package and should be available on all architectures.
It also is present in: http://download.fedora.redhat.com/pub/fedora/linux/development/ppc64/os/Pack...
Examining the contents also shows this package is providing what the report is complaining above:
# rpm -q --provides -p perl-Locale-Maketext-Fuzzy-0.10-2.fc9.noarch.rpm warning: perl-Locale-Maketext-Fuzzy-0.10-2.fc9.noarch.rpm: Header V3 DSA signature: NOKEY, key ID 0b86274e perl(Locale::Maketext::Fuzzy) = 0.10 perl-Locale-Maketext-Fuzzy = 0.10-2.fc9
Ralf
On Mon, 06 Oct 2008 14:15:10 +0200, Ralf Corsepius wrote:
On Mon, 2008-10-06 at 11:51 +0000, Rawhide Report wrote:
Broken deps for ppc64
rt3-3.8.1-1.fc10.noarch requires perl(Locale::Maketext::Fuzzy)
What is going on here?
perl-Locale-Maketext-Fuzzy is a noarch package and should be available on all architectures.
It also is present in: http://download.fedora.redhat.com/pub/fedora/linux/development/ppc64/os/Pack...
Compare with the 20081005 report, where this was not listed as a broken dep. Package is not listed as removed and wasn't updated in this report either. First I thought probably some other package introduced bad Obsoletes, but "python-decoratortools" is also reported as breaking deps although it is present:
TurboGears-1.0.7-2.fc10.noarch requires python-decoratortools >= 0:1.4 mock-0.9.11-1.fc10.noarch requires python-decoratortools python-peak-rules-0.5a1.dev-0.2569.fc10.noarch requires python-decoratortools >= 0:1.7 python-peak-util-addons-0.6-1.fc10.noarch requires python-decoratortools >= 0:1.5 python-peak-util-assembler-0.5-1.fc10.noarch requires python-decoratortools >= 0:1.2
Package is there:
http://download.fedora.redhat.com/pub/fedora/linux/development/ppc64/os/Pack...
On Mon, 2008-10-06 at 16:22 +0200, Michael Schwendt wrote:
Compare with the 20081005 report, where this was not listed as a broken dep. Package is not listed as removed and wasn't updated in this report either. First I thought probably some other package introduced bad Obsoletes, but "python-decoratortools" is also reported as breaking deps although it is present:
TurboGears-1.0.7-2.fc10.noarch requires python-decoratortools >= 0:1.4 mock-0.9.11-1.fc10.noarch requires python-decoratortools python-peak-rules-0.5a1.dev-0.2569.fc10.noarch requires python-decoratortools >= 0:1.7 python-peak-util-addons-0.6-1.fc10.noarch requires python-decoratortools >= 0:1.5 python-peak-util-assembler-0.5-1.fc10.noarch requires python-decoratortools >= 0:1.2
Package is there:
http://download.fedora.redhat.com/pub/fedora/linux/development/ppc64/os/Pack...
This seems to be from a bug we've been chasing in createrepo for a while, at least in the way that we use it to make rawhide.
Error Packages/perl-Locale-Maketext-Fuzzy-0.10-2.fc9.noarch.rpm: No Package ID found for package perl-Locale-Maketext-Fuzzy-0.10-2.fc9.noarch, not going to add it Error Packages/lohit-fonts-tamil-2.3.1-1.fc10.noarch.rpm: No Package ID found for package lohit-fonts-tamil-2.3.1-1.fc10.noarch, not going to add it Error Packages/python-decoratortools-1.7-1.fc9.noarch.rpm: No Package ID found for package python-decoratortools-1.7-1.fc9.noarch, not going to add it
Those are the error messages we get from createrepo, so those packages just don't make it into the repodata, which trips the spam-o-matic script. The problem is that it's not easily reproducible, it's not always the same packages, and it doesn't always happen.
On Mon, 06 Oct 2008 09:20:41 -0700, Jesse Keating wrote:
This seems to be from a bug we've been chasing in createrepo for a while, at least in the way that we use it to make rawhide.
Error Packages/perl-Locale-Maketext-Fuzzy-0.10-2.fc9.noarch.rpm: No Package ID found for package perl-Locale-Maketext-Fuzzy-0.10-2.fc9.noarch, not going to add it Error Packages/lohit-fonts-tamil-2.3.1-1.fc10.noarch.rpm: No Package ID found for package lohit-fonts-tamil-2.3.1-1.fc10.noarch, not going to add it Error Packages/python-decoratortools-1.7-1.fc9.noarch.rpm: No Package ID found for package python-decoratortools-1.7-1.fc9.noarch, not going to add it
Those are the error messages we get from createrepo, so those packages just don't make it into the repodata, which trips the spam-o-matic script. The problem is that it's not easily reproducible, it's not always the same packages, and it doesn't always happen.
Hmmm, an error message should be very good for somebody who is familiar with the createrepo/Yum code:
if po.checksum in (None, ""): raise MDError, "No Package ID found for package %s, not going to add it" % po
I wanted to have a brief look at where that checksum value comes from (here with F9), but the Yum package objects use self._checksum instead. That's weird and doesn't make sense to me yet.
On Mon, 6 Oct 2008 19:10:01 +0200, Michael Schwendt wrote:
if po.checksum in (None, ""): raise MDError, "No Package ID found for package %s, not going to add it" % po
I wanted to have a brief look at where that checksum value comes from (here with F9), but the Yum package objects use self._checksum instead. That's weird and doesn't make sense to me yet.
OMG, it's using properties further down in the same file. =:-O So much about readability...
On Mon, 2008-10-06 at 19:10 +0200, Michael Schwendt wrote:
Hmmm, an error message should be very good for somebody who is familiar with the createrepo/Yum code:
if po.checksum in (None, ""): raise MDError, "No Package ID found for package %s, not
going to add it" % po
I wanted to have a brief look at where that checksum value comes from (here with F9), but the Yum package objects use self._checksum instead. That's weird and doesn't make sense to me yet.
The yum folks are still looking into it. The really bizarre thing is that there is no problem with the package itself, in fact it's seen in other arches, and in those other arch directories it's a hardlink back to the actual file somewhere else on the filesystem. That's the way it is for each of those arch trees. It's a real headscratcher of a problem.
On Mon, 06 Oct 2008 10:19:46 -0700, Jesse Keating wrote:
The yum folks are still looking into it. The really bizarre thing is that there is no problem with the package itself, in fact it's seen in other arches, and in those other arch directories it's a hardlink back to the actual file somewhere else on the filesystem. That's the way it is for each of those arch trees. It's a real headscratcher of a problem.
Unless I'm misreading the code, it's a race-condition. mash creates the metadata for all architectures in parallel. One process per arch. The createrepo checksum cache file I/O [1] is not protected against concurrent access. With odd timing, it can happen that one process reads an empty checksum file that has just been opened for writing by another process. Give it a try. Add a safety check after reading a checksum file.
On Mon, 2008-10-06 at 20:33 +0200, Michael Schwendt wrote:
On Mon, 06 Oct 2008 10:19:46 -0700, Jesse Keating wrote:
The yum folks are still looking into it. The really bizarre thing is that there is no problem with the package itself, in fact it's seen in other arches, and in those other arch directories it's a hardlink back to the actual file somewhere else on the filesystem. That's the way it is for each of those arch trees. It's a real headscratcher of a problem.
Unless I'm misreading the code, it's a race-condition. mash creates the metadata for all architectures in parallel. One process per arch. The createrepo checksum cache file I/O [1] is not protected against concurrent access. With odd timing, it can happen that one process reads an empty checksum file that has just been opened for writing by another process. Give it a try. Add a safety check after reading a checksum file.
That tickles a part of my brain, I think I speculated that this was the problem while in the shower or something, and then completely forgot to pursue it once I got ready to work. It is a possibility.
On Mon, 2008-10-06 at 11:41 -0700, Jesse Keating wrote:
On Mon, 2008-10-06 at 20:33 +0200, Michael Schwendt wrote:
On Mon, 06 Oct 2008 10:19:46 -0700, Jesse Keating wrote:
The yum folks are still looking into it. The really bizarre thing is that there is no problem with the package itself, in fact it's seen in other arches, and in those other arch directories it's a hardlink back to the actual file somewhere else on the filesystem. That's the way it is for each of those arch trees. It's a real headscratcher of a problem.
Unless I'm misreading the code, it's a race-condition. mash creates the metadata for all architectures in parallel. One process per arch. The createrepo checksum cache file I/O [1] is not protected against concurrent access. With odd timing, it can happen that one process reads an empty checksum file that has just been opened for writing by another process. Give it a try. Add a safety check after reading a checksum file.
That tickles a part of my brain, I think I speculated that this was the problem while in the shower or something, and then completely forgot to pursue it once I got ready to work. It is a possibility.
This should fix it:
http://koji.fedoraproject.org/koji/taskinfo?taskID=868762
...the patch being basically:
- csumo = open(csumfile, 'w') + + # This is atomic cache creation via. rename, so we can have two + # tasks using the same cachedir ... mash does this. + csumo = tempfile.NamedTemporaryFile(mode='w', dir=self.crp_cachedir csumo.write(checksum) csumo.close() + os.rename(csumo.name, csumfile)
...it'll get upstream as soon as I can push :).
On Wed, 08 Oct 2008 12:44:35 -0400, James Antill wrote:
[race in mash+createrepo]
This should fix it:
http://koji.fedoraproject.org/koji/taskinfo?taskID=868762
...the patch being basically:
csumo = open(csumfile, 'w')
# This is atomic cache creation via. rename, so we can have two
# tasks using the same cachedir ... mash does this.
csumo = tempfile.NamedTemporaryFile(mode='w', dir=self.crp_cachedir csumo.write(checksum) csumo.close()
os.rename(csumo.name, csumfile)
...it'll get upstream as soon as I can push :).
Looks good. It's the most you can do in createrepo without adding locks. A small penalty remains, because the same checksum may be generated more than once in a race-condition. Neglectable IMO.
[...]
Quick update on broken deps investigation with latest rawhide:
Broken deps for i386
asterisk-mobile-1.6.0-0.21.beta9.fc10.i386 requires libbluetooth.so.2
Already a response on:
https://bugzilla.redhat.com/show_bug.cgi?id=465661
looks like some new packages need a review before asterisk 1.6.0 will build properly.
gnuradio-3.1.2-2.fc9.i386 requires libcppunit-1.12.so.0
I took the liberty of rebuilding this package for broken deps because of libcppunit soname bump:
https://koji.fedoraproject.org/koji/buildinfo?buildID=52252
it had been rebuilt for F-10 before, but for some reason the build was cancelled:
https://koji.fedoraproject.org/koji/buildinfo?buildID=52252
gtkmozembedmm-1.4.2.cvs20060817-21.fc10.i386 requires gecko-libs = 0:1.9.0.1
Opened up a new bug requesting clarification:
https://bugzilla.redhat.com/show_bug.cgi?id=465910
I suspect the maintainer simply wanted to orphan the package but didn't really know how.
lvm2-cluster-2.02.39-3.fc10.i386 requires libcman.so.2 lvm2-cluster-2.02.39-3.fc10.i386 requires libdlm.so.2
Looking at a my own scratch build:
https://koji.fedoraproject.org/koji/taskinfo?taskID=865866
this actually is a simple spec file mistake, the build error is:
error: File not found: /builddir/build/BUILDROOT/lvm2-1.02.27-5.fc10.x86_64/usr/sbin/lvmconf
and the spec file says:
%files cluster %defattr(-,root,root,-) %attr(755,root,root) /usr/sbin/clvmd %{_sbindir}/lvmconf %{_mandir}/man8/clvmd.8.gz /etc/rc.d/init.d/clvmd
but the %install part of the spec file is placing lvmconf in /usr/sbin:
install -m 0755 scripts/lvmconf.sh $RPM_BUILD_ROOT/sbin/lvmconf
I would fix this, but there are ACLs.
1:openoffice.org-langpack-ta_IN-3.0.0-9.2.fc10.i386 requires lohit-fonts-tamil publican-0.37-0.fc10.noarch requires lohit-fonts-tamil
Not sure about these, was lohit-fonts-tamil removed from rawhide?
pyclutter-0.6.2-2.fc9.i386 requires libclutter-glx-0.6.so.0 pyclutter-cairo-0.6.2-2.fc9.i386 requires libclutter-cairo-0.6.so.0 pyclutter-cairo-0.6.2-2.fc9.i386 requires libclutter-glx-0.6.so.0 pyclutter-gst-0.6.2-2.fc9.i386 requires libclutter-gst-0.6.so.0 pyclutter-gst-0.6.2-2.fc9.i386 requires libclutter-glx-0.6.so.0 pyclutter-gtk-0.6.2-2.fc9.i386 requires libclutter-glx-0.6.so.0 pyclutter-gtk-0.6.2-2.fc9.i386 requires libclutter-gtk-0.6.so.0
https://bugzilla.redhat.com/show_bug.cgi?id=465096
ruby-rpm-1.2.3-4.fc9.i386 requires librpmdb-4.4.so ruby-rpm-1.2.3-4.fc9.i386 requires librpm-4.4.so ruby-rpm-1.2.3-4.fc9.i386 requires librpmio-4.4.so ruby-rpm-1.2.3-4.fc9.i386 requires librpmbuild-4.4.so
https://bugzilla.redhat.com/show_bug.cgi?id=465103
syncevolution-0.7-3.fc10.i386 requires libcppunit-1.12.so.0
Again, just rebuild for new libcppunit:
On Mon, Oct 06, 2008 at 09:33:29PM -0700, Alex Lancaster wrote:
gnuradio-3.1.2-2.fc9.i386 requires libcppunit-1.12.so.0
I took the liberty of rebuilding this package for broken deps because of libcppunit soname bump:
Umh, I didn't announced cppunit soname bump because I thought that nothing ended up being linked at runtime with cppunit... Looks like I was wrong: # repoquery --whatrequires libcppunit-1.12.so.0 syncevolution-0:0.7-3.fc10.i386 gnuradio-0:3.1.2-2.fc9.i386
However these may be spurious deps, I'll have a look.
-- Pat