rawhide report: 20081018 changes

[Fedora compose checker]

Compose started at Sat Oct 18 06:01:17 UTC 2008

New package MegaMek
        A portable, network-enabled BattleTech engine
New package ipmitool
        Utility for IPMI control
New package system-switch-java
        A tool for changing the default Java toolset
Removed package packagekit-qt
Updated Packages:

OpenIPMI-2.0.14-3.fc10
----------------------
* Thu Oct 16 18:00:00 2008 Jan Safranek <jsafrane(a)redhat.com&gt; - 2.0.14-3
- split ipmitool to separate package
- add 'reload' functionality to init script
- add seraparate -gui subpackage


alacarte-0.11.6-4.fc10
----------------------
* Fri Oct 17 18:00:00 2008 Matthias Clasen <mclasen(a)redhat.com&gt; - 0.11.6-4
- Make undoing of deletion work


anaconda-11.4.1.50-1
--------------------
* Fri Oct 17 18:00:00 2008 Chris Lumens <clumens(a)redhat.com&gt; - 11.4.1.50-1
- Update several font package names that we were missing. (clumens)
- Only bring up the netconfig dialog if the repo requires networking.
  (clumens)
- cmdline.py: Fix a small typo in a message (rh 467338) (hansg)
- Enable CCW devices used for installation (#253075) (dcantrell)
- I don't know what trq.map.trq-map is, but let's not include it. (clumens)
- If networking is needed for yum repos, bring it up before fetching
  repodata. (clumens)
- Force DPI to 96 when launching X. (#458738) (ajax)
- Lots of translation updates.


bluez-4.14-2.fc10
-----------------

cairo-dock-1.6.3-0.2.svn1357_trunk.fc10
---------------------------------------
* Sat Oct 18 18:00:00 2008 Mamoru Tasaka <mtasaka(a)ioa.s.u-tokyo.ac.jp&gt;
- rev 1357


eclipse-3.4.1-4.fc10
--------------------
* Fri Oct 17 18:00:00 2008 Andrew Overholt <overholt(a)redhat.com&gt; 3.4.1-4
- Fix quoting bug in pdebuild.sh again.


eclipse-pydev-1.3.20-2.fc10
---------------------------
* Fri Oct 17 18:00:00 2008 Alexander Kurtakov <akurtako(a)redhat.com&gt; 1:1.3.20-2
- Drop gcj support.
- Split mydev support to subpackage to correct dependencies.
- Drop eclipse-mylyn-ide BR - it is merged into eclipse-mylyn.


ecryptfs-utils-59-2.fc10
------------------------
* Fri Oct 17 18:00:00 2008 Eric Sandeen <sandeen(a)redhat.com&gt; 59-2
- Remove duplicate doc files from rpm


filezilla-3.1.4.1-1.fc10
------------------------
* Fri Oct 17 18:00:00 2008 kwizart < kwizart at gmail.com > - 3.1.4.1-1
- Update to 3.1.4.1


fmtools-1.0.2-4.fc10
--------------------
* Fri Oct 17 18:00:00 2008 kwizart < kwizart at gmail.com > - 1.0.2-4
- Rebuild for F-10


geos-3.0.1-1.fc10
-----------------
* Fri Oct 17 18:00:00 2008 Balint Cristian <rezso(a)rdsor.ro&gt; - 3.0.1-1
- new stable bugfix
- fix another gcc43 header


ghostscript-8.63-4.fc10
-----------------------
* Fri Oct 17 18:00:00 2008 Tim Waugh <twaugh(a)redhat.com&gt; 8.63-4
- Try out a work-around for bug #465311.


gtk2-2.14.4-1.fc10
------------------
* Fri Oct 17 18:00:00 2008 Matthias Clasen <mclasen(a)redhat.com&gt; - 2.14.4-1
- Update to 2.14.4


hunspell-1.2.7-4.fc10
---------------------
* Fri Oct 17 18:00:00 2008 Caolan McNamara <caolanm(a)redhat.com&gt; - 1.2.7-4
- make wordlist2hunspell remove blank lines


ibus-table-0.1.1.20081014-1.fc10
--------------------------------
* Tue Oct 14 18:00:00 2008 Peng Huang <shawn.p.huang(a)gmail.com&gt; - 0.1.1.20081014-1
- Update to 0.1.1.20081014.


imsettings-0.105.0-3.fc10
-------------------------
* Fri Oct 17 18:00:00 2008 Than Ngo <than(a)redhat.com&gt; 0.105.0-3
- readd the workaround for KDE


inkscape-0.46-6.fc10
--------------------
* Fri Oct 17 18:00:00 2008 Lubomir Rintel <lkundrak(a)v3.sk&gt; - 0.46-6
- Fix color sliders with recent GTK (#467431)


ipsec-tools-0.7.1-5.fc10
------------------------
* Fri Oct 17 18:00:00 2008 Tomas Mraz <tmraz(a)redhat.com&gt; - 0.7.1-5
- fix CVE-2008-3652 (memory leak DoS)
- compile racoon as PIE
- another fix for teardown of the IPSEC SAs on DPD in some circumstances


java-1.6.0-openjdk-1.6.0.0-0.23.b12.fc10
----------------------------------------
* Tue Oct 14 18:00:00 2008 Lillian Angel <langel(a)redhat.com&gt; - 1:1.6.0-0.23.b12
- Updated icedteasnapshot to nil.
- Updated release.
- Added jakarta-commons-logging requirement for visualvm.

* Thu Oct  2 18:00:00 2008 Lillian Angel <langel(a)redhat.com&gt; - 1:1.6.0-0.22.b12
- Enabled building of jvisualvm tool.


logwatch-7.3.6-29.fc10
----------------------
* Fri Oct 17 18:00:00 2008 Ivana Varekova <varekova(a)redhat.com&gt; 7.3.6-29
- fix postfix script again (#462174)


mfiler3-2.1.1-1.fc10
--------------------
* Sat Oct 18 18:00:00 2008 Mamoru Tasaka <mtasaka(a)ioa.s.u-tokyo.ac.jp&gt; - 2.1.1-1
- 2.1.1


mod_perl-2.0.4-7
----------------
* Fri Oct 17 18:00:00 2008 Joe Orton <jorton(a)redhat.com&gt; 2.0.4-7
- fix inline abuse (#459351)


nspluginwrapper-1.1.2-2.fc10
----------------------------
* Fri Oct 17 18:00:00 2008 Martin Stransky <stransky(a)redhat.com&gt; 1.1.2-2
- added umask to plugin config (#463736)


nss_ldap-261-4.fc10
-------------------
* Fri Oct 17 18:00:00 2008 Nalin Dahyabhai <nalin(a)redhat.com&gt; - 261-4
- add missing local-to-network conversion on port numbers when looking up
  services (#450634)

* Mon Sep 15 18:00:00 2008 Nalin Dahyabhai <nalin(a)redhat.com&gt;
- return 0 (fail) instead of 1 (success) when setnetgrent() is called for
  a netgroup which doesn't actually exist or which has no members (#445972,
  upstream #374)


nted-1.4.2-1.fc10
-----------------
* Fri Oct 17 18:00:00 2008 Hans Ulrich Niedermann <hun(a)n-dimensional.de&gt; - 1.4.2-1
- Update to upstream's 1.4.2 release.


obex-data-server-0.4-1.fc10
---------------------------
* Fri Oct 17 18:00:00 2008 - Bastien Nocera <bnocera(a)redhat.com&gt; - 0.4-1
- Update to 0.4


obexd-0.6-1.fc10
----------------
* Fri Oct 17 18:00:00 2008 - Bastien Nocera <bnocera(a)redhat.com&gt; - 0.6-1
- Update to 0.6


ocsinventory-1.02-0.9.rc3.fc10
------------------------------
* Fri Oct 17 18:00:00 2008 Remi Collet <Fedora(a)famillecollet.com&gt; 1.02-0.9.rc3
- upstream r1423 patch - migration script


openssh-5.1p1-3.fc10
--------------------
* Fri Oct 17 18:00:00 2008 Tomas Mraz <tmraz(a)redhat.com&gt; - 5.1p1-3
- fix compatibility with some servers (#466818)


perl-AnyEvent-4.300-1.fc10
--------------------------
* Fri Oct 17 18:00:00 2008 kwizart < kwizart at gmail.com > - 4.300-1
- Update to 4.3   (rpm version : 4.300 )


perl-XML-Grove-0.46alpha-37.fc10
--------------------------------
* Fri Oct 17 18:00:00 2008 Stepan Kasal <skasal(a)redhat.com&gt; - 0.46alpha-37
- avoid packing DOM-ecmascript.pod twice


plymouth-0.6.0-0.2008.10.17.1.fc10
----------------------------------
* Fri Oct 17 18:00:00 2008 Ray Strode <rstrode(a)redhat.com&gt; 0.5.0-0.2008.10.17.1
- Clear screen in details plugin when it's done
- Make plymouth-update-initrd a small wrapper around mkinitrd instead
  of the broken monstrosity it was before.

* Fri Oct 17 18:00:00 2008 Ray Strode <rstrode(a)redhat.com&gt; 0.5.0-0.2008.10.15.3
- Move plymouth-set-default-plugin, plymouth-update-initrd, and
  plymouth-populate-initrd to plymouth-scripts subpackage
  (the last fix didn't actually help with bug 467356)

* Fri Oct 17 18:00:00 2008 Ray Strode <rstrode(a)redhat.com&gt; 0.5.0-0.2008.10.15.2
- Move plymouth-set-default-plugin to -libs (might help with bug 467356)
- Fix up requires, provides and postun scripts


prelude-lml-0.9.14-1.fc10
-------------------------
* Fri Oct 17 18:00:00 2008 Steve Grubb <sgrubb(a)redhat.com&gt; 0.9.14-1
- new upstream release fixing bz #463459


pyclutter-0.8.0-1.fc10
----------------------
* Wed Oct 15 18:00:00 2008 Allisson Azevedo <allisson(a)gmail.com&gt; 0.8.0-1
- Update to 0.8.0


python-gammu-0.27-1.fc10
------------------------
* Sat Oct 11 18:00:00 2008 Xavier Lamien <lxtnow(a)gmail.com&gt; - 0.27-1
- Update release.


python-ply-2.5-1.fc10
---------------------
* Fri Oct 17 18:00:00 2008 Tom "spot" Callaway <tcallawa(a)redhat.com&gt;
2.5-1
- update to 2.5


selinux-policy-3.5.13-1.fc10
----------------------------
* Fri Oct 17 18:00:00 2008 Dan Walsh <dwalsh(a)redhat.com&gt; 3.5.13-1
- Policy cleanup


slim-1.3.1-2.fc10
-----------------
* Wed Oct 15 18:00:00 2008 Marco Pesenti Gritti <mpg(a)redhat.com&gt;  1.3.1-2
- Enable pam_selinux


system-config-printer-1.0.8-7.fc10
----------------------------------
* Fri Oct 17 18:00:00 2008 Tim Waugh <twaugh(a)redhat.com&gt; 1.0.8-7
- Added patch for pysmbc changes in git to prevent getdents crashing
  (bug #465975).


varnish-2.0.1-1.fc10
--------------------
* Fri Oct 17 18:00:00 2008 Ingvar Hagelund <ingvar(a)linpro.no&gt; - 2.0.1-1
- 2.0.1 released, a bugfix release. New upstream sources


Summary:
Added Packages: 3
Removed Packages: 1
Modified Packages: 40
Broken deps for i386
----------------------------------------------------------
	python-basemap-0.99-2.fc10.i386 requires libgeos-3.0.0.so
	qgis-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so
	qgis-grass-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so
	qgis-python-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so
	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



Broken deps for x86_64
----------------------------------------------------------
	python-basemap-0.99-2.fc10.x86_64 requires libgeos-3.0.0.so()(64bit)
	qgis-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so
	qgis-0.11.0-2.fc10.x86_64 requires libgeos-3.0.0.so()(64bit)
	qgis-grass-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so
	qgis-grass-0.11.0-2.fc10.x86_64 requires libgeos-3.0.0.so()(64bit)
	qgis-python-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so
	qgis-python-0.11.0-2.fc10.x86_64 requires libgeos-3.0.0.so()(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)



Broken deps for ppc
----------------------------------------------------------
	python-basemap-0.99-2.fc10.ppc requires libgeos-3.0.0.so
	qgis-0.11.0-2.fc10.ppc requires libgeos-3.0.0.so
	qgis-0.11.0-2.fc10.ppc64 requires libgeos-3.0.0.so()(64bit)
	qgis-grass-0.11.0-2.fc10.ppc requires libgeos-3.0.0.so
	qgis-grass-0.11.0-2.fc10.ppc64 requires libgeos-3.0.0.so()(64bit)
	qgis-python-0.11.0-2.fc10.ppc requires libgeos-3.0.0.so
	qgis-python-0.11.0-2.fc10.ppc64 requires libgeos-3.0.0.so()(64bit)
	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



Broken deps for ppc64
----------------------------------------------------------
	livecd-tools-019-1.fc10.ppc64 requires yaboot
	python-basemap-0.99-2.fc10.ppc64 requires libgeos-3.0.0.so()(64bit)
	qgis-0.11.0-2.fc10.ppc64 requires libgeos-3.0.0.so()(64bit)
	qgis-grass-0.11.0-2.fc10.ppc64 requires libgeos-3.0.0.so()(64bit)
	qgis-python-0.11.0-2.fc10.ppc64 requires libgeos-3.0.0.so()(64bit)
	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)

[Alex Lancaster]

> >>>> Rawhide Report  writes: 
[...]

Balint,

Can you please announce any soname bumps here on the list *before*
rebuilding to give other package maintainers a chance to rebuild their
dependent packages in time for the next rawhide compose?  

>  geos-3.0.1-1.fc10 ----------------- 
>  * Fri Oct 17 18:00:00 2008
>  Balint Cristian <rezso(a)rdsor.ro&gt; - 3.0.1-1 - new stable bugfix - fix
>  another gcc43 header 
At the very least it gives people a heads-up that some things might
break in the next rawhide.  Especially when those dependent packages
aren't owned by you such as qgis and python-basemap:

>  Broken deps for i386
>  ----------------------------------------------------------
>  python-basemap-0.99-2.fc10.i386 requires libgeos-3.0.0.so
>  qgis-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so
>  qgis-grass-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so
>  qgis-python-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so 
Advance announcements of soname bumps on fedora-devel-list is
especially important when we are getting close to the final freeze.

Thanks,
Alex

[Alex Lancaster]

> >>>> "AL" == Alex Lancaster  writes: 
[...]

> > Broken deps for i386
> > ----------------------------------------------------------
> > python-basemap-0.99-2.fc10.i386 requires libgeos-3.0.0.so
> > qgis-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so
> > qgis-grass-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so
> > qgis-python-0.11.0-2.fc10.i386 requires libgeos-3.0.0.so 
AL> Advance announcements of soname bumps on fedora-devel-list is
AL> especially important when we are getting close to the final
AL> freeze.

I took the liberty of rebuilding these two packages to fix the broken
deps:

python-basemap:
https://koji.fedoraproject.org/koji/buildinfo?buildID=66750

qgis:
https://koji.fedoraproject.org/koji/buildinfo?buildID=66749

Alex

[Jef Spaleta]

On Sat, Oct 18, 2008 at 3:17 AM, Alex Lancaster
<alexl(a)users.sourceforge.net&gt; wrote:
> > geos-3.0.1-1.fc10 -----------------
> > * Fri Oct 17 18:00:00 2008
> > Balint Cristian <rezso(a)rdsor.ro&gt; - 3.0.1-1 - new stable bugfix - fix
> > another gcc43 header
>
>  At the very least it gives people a heads-up that some things might
>  break in the next rawhide.  Especially when those dependent packages
>  aren't owned by you such as qgis and python-basemap: 

Hey guess what the geos bump is an API change.  You quick fix of the
build script for python-basemap doesn't actually work the code checks
the geos version via a library call. So your "fix" gets basemap
building...but it will not import.  Please, please, please test
patches in runtime before merging them.

-jef"Geos moving to 3.0 at this late date is a problem for me. Not
cool. This should have come in MUCH earlier in devel cycle"spaleta

[Jef Spaleta]

On Wed, Oct 22, 2008 at 5:04 PM, Jeff Spaleta <jspaleta(a)gmail.com&gt; wrote:
>  -jef"Geos moving to 3.0 at this late date is a problem for me.
> Not
>  cool. This should have come in MUCH earlier in devel cycle"spaleta 


So yeah... not the packagers fault.... clearly upstream "fixed" their
code to report version 3.0 instead of 2.3..well after the 3.0 release.

Alex if you have the time, please hunt down and patch the runtime
version query that is keeping this from importing.  I'm not gonna be
able to get to this till the weekend it seems.

-jef

[Kevin Kofler]

Jeff Spaleta <jspaleta <at> gmail.com> writes:
>  Alex if you have the time, please hunt down and patch the runtime
>  version query that is keeping this from importing.  I'm not gonna be
>  able to get to this till the weekend it seems. 
Well, I'm not Alex, but I can see what I can do, it looks like an easy fix to 
me. Though I don't have a Rawhide machine to test the result on.

        Kevin Kofler

[Jesse Keating]

On Thu, 2008-10-23 at 01:23 +0000, Kevin Kofler wrote:
>  Well, I'm not Alex, but I can see what I can do, it looks like an
> easy fix to 
>  me. Though I don't have a Rawhide machine to test the result on. 
Should be easy enough to test inside a mock chroot.

-- 
Jesse Keating
Fedora -- Freedom² is a feature!
identi.ca: http://identi.ca/jkeating

[Jef Spaleta]

On Wed, Oct 22, 2008 at 5:23 PM, Kevin Kofler <kevin.kofler(a)chello.at&gt; wrote:
>  Well, I'm not Alex, but I can see what I can do, it looks like an
> easy fix to
>  me. Though I don't have a Rawhide machine to test the result on. 
I got it. I was able to pull the source here. I'll have a fix
committed tonite after testing it.

-jef

[Kevin Kofler]

Jeff Spaleta <jspaleta <at> gmail.com> writes:
>  I got it. I was able to pull the source here. I'll have a fix
>  committed tonite after testing it. 
Actually, I already committed this:
http://cvs.fedoraproject.org/viewvc/rpms/python-basemap/devel/python-base...
and the build is now running. Does this look sane? (As for testing, it can be 
done once it's built, it's not like it can get much worse than not working at 
all. ;-) )

        Kevin Kofler

[Jef Spaleta]

On Wed, Oct 22, 2008 at 5:57 PM, Kevin Kofler <kevin.kofler(a)chello.at&gt; wrote:
>  done once it's built, it's not like it can get much worse
> than not working at
>  all. ;-) ) 

it can...if you are using it for science!  Not working at all is far
far better than working and giving you the wrong calculation.

I tested some of the exposed geos functions involving great circle
calculations and things seem okay.

But I did find that the examples scripts in the package haven't been
ported yet...la la la. Silly upstream need to file some patches with
them.

-jef

Make announcing soname bumps mandatory (was Re: rawhide report: 20081018 changes)

[Alex Lancaster]

> >>>> "JS" == Jeff Spaleta  writes: 
JS> On Wed, Oct 22, 2008 at 5:04 PM, Jeff Spaleta <jspaleta(a)gmail.com&gt; wrote:

> > -jef"Geos moving to 3.0 at this late date is a problem for
> me. Not
> > cool. This should have come in MUCH earlier in devel cycle"spaleta 
Yes, indeed bumping to 3.0 so late is not a good idea, nevertheless it
shouldn't be forbidden.  What should be forbidden is not announcing it
beforehand.  This is why I specifically reminded Balint to, at the
very least, announce it here on fedora-devel-list as per policy:

https://fedoraproject.org/wiki/PackageMaintainers/Packaging_Tricks#When_b...

I think that the above policy should be made mandatory rather than
"nice to" (unless you happen to be building all affected package
yourself in which case an e-mail is superfluous).  At the very least
it would get maintainers in the habit of running repoquery more
regularly and so they would get a better sense of what other packages
depended on their package (many maintainers seem unaware of this).
Ideally this would be automated and provided in regular reports, or
somehow incorporated into PackageDB.

Unfortunately this bumping the sonames without notifying the list
and/or without co-ordinating with other package maintainers keeps
cropping up and the evolution maintainers are particularly guilty of
this.  I don't think that the rebuild of evolution/e-d-s was announced
here either and it has caused more breakage in rawhide 20081022:

Broken deps for i386
----------------------------------------------------------
	evolution-brutus-1.2.27-1.fc10.i386 requires libcamel-1.2.so.13
	evolution-zimbra-0.1.1-2.fc10.i386 requires libcamel-provider-1.2.so.13
	evolution-zimbra-0.1.1-2.fc10.i386 requires libcamel-1.2.so.13
	planner-eds-0.14.3-5.fc10.i386 requires libcamel-provider-1.2.so.13
	planner-eds-0.14.3-5.fc10.i386 requires libcamel-1.2.so.13


JS> So yeah... not the packagers fault.... clearly upstream "fixed"
JS> their code to report version 3.0 instead of 2.3..well after the
JS> 3.0 release.

I just wanted to fix the broken deps with the most minimal packaging
fix, because without it building it wouldn't even be installable.

JS> Alex if you have the time, please hunt down and patch the runtime
JS> version query that is keeping this from importing.  I'm not gonna
JS> be able to get to this till the weekend it seems.

Looks like Kevin beat me to it...

Alex

Make announcing soname bumps mandatory (was Re: rawhide report: 20081018 changes)

[Jesse Keating]

On Thu, 2008-10-23 at 01:17 -0700, Alex Lancaster wrote:
>  I think that the above policy should be made mandatory rather than
>  "nice to" (unless you happen to be building all affected package
>  yourself in which case an e-mail is superfluous).  At the very least
>  it would get maintainers in the habit of running repoquery more
>  regularly and so they would get a better sense of what other packages
>  depended on their package (many maintainers seem unaware of this).
>  Ideally this would be automated and provided in regular reports, or
>  somehow incorporated into PackageDB. 

Something that I'm planning for our infrastructure in the future is the
ability to do a build in the buildsystem, have it go through some
post-build examination such as:
  Did a soname change?
  Did provides change?
  Did requires change?
  Did files get removed?
  Did files get added?

so on and so forth, all geared toward "would this cause problems if
introduced into the package set?".  The maintainer would get a report
and if some things triggered poorly (like so name changes) the build
wouldn't get tagged into the collection, and the maintainer would have
to interact with something to either force the tag (which could then
trigger the emails you're talking about, soname bumping etc...) or throw
the build away and do something different.

Of course this type of functionality is blocking on a few things, like
an automated testing environment.  But hey, we'll get there at some
point!

-- 
Jesse Keating
Fedora -- Freedom² is a feature!
identi.ca: http://identi.ca/jkeating

Make announcing soname bumps mandatory (was Re: rawhide report: 20081018 changes)

[Ignacio Vazquez-Abrams]

On Thu, 2008-10-23 at 10:05 -0700, Jesse Keating wrote:
>  Something that I'm planning for our infrastructure in the future
> is the
>  ability to do a build in the buildsystem, have it go through some
>  post-build examination such as:
>    Did a soname change?
>    Did provides change?
>    Did requires change?
>    Did files get removed?
>    Did files get added? 
There's my cue...

https://fedorahosted.org/bodhi/ticket/79

-- 
Ignacio Vazquez-Abrams <ivazqueznet(a)gmail.com&gt;

PLEASE don't CC me; I'm already subscribed

Make announcing soname bumps mandatory (was Re: rawhide report: 20081018 changes)

[Jesse Keating]

On Thu, 2008-10-23 at 13:17 -0400, Ignacio Vazquez-Abrams wrote:
>  There's my cue...
>  
>  https://fedorahosted.org/bodhi/ticket/79 
Right, checking deps is one of those automated tests.  The trick is
getting it done after the build, but before the tag, so that if
everything is successful it just goes right in, but if it's not
successful it can either be waived (think groups of packages all updated
at the same time for new sonames), or thrown out.

-- 
Jesse Keating
Fedora -- Freedom² is a feature!
identi.ca: http://identi.ca/jkeating

[Matt Domsch]

On Sat, Oct 18, 2008 at 10:09:59AM +0000, Rawhide Report wrote:
>  Compose started at Sat Oct 18 06:01:17 UTC 2008
>  
>  New package ipmitool
>          Utility for IPMI control
>  Updated Packages:
>  
>  OpenIPMI-2.0.14-3.fc10
>  ----------------------
>  * Thu Oct 16 18:00:00 2008 Jan Safranek <jsafrane(a)redhat.com&gt; - 2.0.14-3
>  - split ipmitool to separate package 
Yea! Thank you!  I've argued unsuccessfully for years that ipmitool
should have been its own package.
-Matt

-- 
Matt Domsch
Linux Technology Strategist, Dell Office of the CTO
linux.dell.com & www.dell.com/linux

[Peter Robinson]

> > New package ipmitool
> >         Utility for IPMI control
> > Updated Packages:
> >
> > OpenIPMI-2.0.14-3.fc10
> > ----------------------
> > * Thu Oct 16 18:00:00 2008 Jan Safranek <jsafrane(a)redhat.com&gt; - 2.0.14-3
> > - split ipmitool to separate package
>
>  Yea! Thank you!  I've argued unsuccessfully for years that ipmitool
>  should have been its own package.
>  -Matt 
What was it part of before? I just thought it had never been packaged.

Peter