More Fedora mock breakage

[Tom Lane]

I'm getting a pile of these:

ERROR with rpm_check_debug vs depsolve:
rpmlib(PayloadIsXz) is needed by exim-4.69-12.fc12.x86_64
rpmlib(PayloadIsXz) is needed by crontabs-1.10-31.fc12.noarch
rpmlib(PayloadIsXz) is needed by logrotate-3.7.8-3.fc12.x86_64
rpmlib(PayloadIsXz) is needed by tar-2:1.22-6.fc12.x86_64
rpmlib(PayloadIsXz) is needed by openldap-2.4.16-2.fc12.x86_64
rpmlib(PayloadIsXz) is needed by ncurses-libs-5.7-3.20090207.fc12.x86_64
rpmlib(PayloadIsXz) is needed by grep-2.5.3-5.fc12.x86_64
rpmlib(PayloadIsXz) is needed by ncurses-base-5.7-3.20090207.fc12.x86_64
rpmlib(PayloadIsXz) is needed by gdbm-1.8.0-33.fc12.x86_64
rpmlib(PayloadIsXz) is needed by xz-4.999.8-0.8.beta.fc12.x86_64
rpmlib(PayloadIsXz) is needed by ustr-1.0.4-9.fc12.x86_64
...

while trying to build rawhide packages under mock.  Host is a fully
up2date Fedora 10 x86_64 system, selected mock config is
fedora-rawhide-x86_64.

What's especially curious is that it seems to work if I rm -rf
the contents of /var/lib/mock and /var/cache/mock first...
but I'm not eager to re-download all that stuff every time.

			regards, tom lane

[Mathieu Bridon]

>  ERROR with rpm_check_debug vs depsolve:
>  rpmlib(PayloadIsXz) is needed by exim-4.69-12.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by crontabs-1.10-31.fc12.noarch
>  rpmlib(PayloadIsXz) is needed by logrotate-3.7.8-3.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by tar-2:1.22-6.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by openldap-2.4.16-2.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by ncurses-libs-5.7-3.20090207.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by grep-2.5.3-5.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by ncurses-base-5.7-3.20090207.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by gdbm-1.8.0-33.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by xz-4.999.8-0.8.beta.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by ustr-1.0.4-9.fc12.x86_64
>  ...
>
>  while trying to build rawhide packages under mock.  Host is a fully
>  up2date Fedora 10 x86_64 system, selected mock config is
>  fedora-rawhide-x86_64. 
I had the same on F11, building in a Rawhide mock.

I was advised to update rpm to 4.7.1 that was in updates-testing,
which fixed the problem.


----------

Mathieu Bridon (bochecha)

[John5342]

2009/7/31 Tom Lane <tgl(a)redhat.com&gt;:
>  I'm getting a pile of these:
>
>  ERROR with rpm_check_debug vs depsolve:
>  rpmlib(PayloadIsXz) is needed by exim-4.69-12.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by crontabs-1.10-31.fc12.noarch
>  rpmlib(PayloadIsXz) is needed by logrotate-3.7.8-3.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by tar-2:1.22-6.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by openldap-2.4.16-2.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by ncurses-libs-5.7-3.20090207.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by grep-2.5.3-5.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by ncurses-base-5.7-3.20090207.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by gdbm-1.8.0-33.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by xz-4.999.8-0.8.beta.fc12.x86_64
>  rpmlib(PayloadIsXz) is needed by ustr-1.0.4-9.fc12.x86_64
>  ...
>
>  while trying to build rawhide packages under mock.  Host is a fully
>  up2date Fedora 10 x86_64 system, selected mock config is
>  fedora-rawhide-x86_64.
>
>  What's especially curious is that it seems to work if I rm -rf
>  the contents of /var/lib/mock and /var/cache/mock first...
>  but I'm not eager to re-download all that stuff every time. 
I think this is down to the new xz/lzma compression for rpms in f12 as
mentioned here:

https://fedoraproject.org/wiki/Features/XZRpmPayloads

Probably need a more recent version of rpm with xz support.

-- 
There are 10 kinds of people in the world: Those who understand binary
and those who don't...

[Tom Lane]

"Mathieu Bridon (bochecha)" <bochecha(a)fedoraproject.org&gt; writes:
> > ERROR with rpm_check_debug vs depsolve:
> > rpmlib(PayloadIsXz) is needed by exim-4.69-12.fc12.x86_64 
>  I had the same on F11, building in a Rawhide mock. 
>  I was advised to update rpm to 4.7.1 that was in updates-testing,
>  which fixed the problem. 
[ consults CVS... ]  So XZ support in F-11's rpm is less than a week
old, there is *no* support in F-10, and we're already requiring
the capability in order to do useful development work?

All I can say is WTF.

			regards, tom lane

[Josh Boyer]

On Fri, Jul 31, 2009 at 06:48:54PM -0400, Tom Lane wrote:
> "Mathieu Bridon (bochecha)"
> <bochecha(a)fedoraproject.org&gt; writes:
> >> ERROR with rpm_check_debug vs depsolve:
> >> rpmlib(PayloadIsXz) is needed by exim-4.69-12.fc12.x86_64
>
> > I had the same on F11, building in a Rawhide mock.
>
> > I was advised to update rpm to 4.7.1 that was in updates-testing,
> > which fixed the problem.
>
> [ consults CVS... ]  So XZ support in F-11's rpm is less than a week
> old, there is *no* support in F-10, and we're already requiring
> the capability in order to do useful development work?
>
> All I can say is WTF. 
Did you have a better plan for migrating to an XZ payload for F12 in time
for Alpha?

josh

[Bastien Nocera]

On Fri, 2009-07-31 at 18:48 -0400, Tom Lane wrote:
>  "Mathieu Bridon (bochecha)"
> <bochecha(a)fedoraproject.org&gt; writes:
>  >> ERROR with rpm_check_debug vs depsolve:
>  >> rpmlib(PayloadIsXz) is needed by exim-4.69-12.fc12.x86_64
>  
>  > I had the same on F11, building in a Rawhide mock.
>  
>  > I was advised to update rpm to 4.7.1 that was in updates-testing,
>  > which fixed the problem.
>  
>  [ consults CVS... ]  So XZ support in F-11's rpm is less than a week
>  old, there is *no* support in F-10, and we're already requiring
>  the capability in order to do useful development work?
>  
>  All I can say is WTF. 
An rpm with Xz support is in F11 updates-testing. I'm pretty sure it's
also in updates-testing for F10.

[Tom Lane]

Bastien Nocera <bnocera(a)redhat.com&gt; writes:
>  On Fri, 2009-07-31 at 18:48 -0400, Tom Lane wrote:
> > [ consults CVS... ]  So XZ support in F-11's rpm is less than a week
> > old, there is *no* support in F-10, and we're already requiring
> > the capability in order to do useful development work? 
> > All I can say is WTF. 
>  An rpm with Xz support is in F11 updates-testing. I'm pretty sure
> it's
>  also in updates-testing for F10. 
It is not in F-10's package CVS, let alone in updates-testing.  Yes,
I looked before complaining.  This was not well-managed.

			regards, tom lane

[Orcan Ogetbil]

On Sat, Aug 1, 2009 at 12:09 AM, Tom Lane wrote:
>  Bastien Nocera writes:
> > On Fri, 2009-07-31 at 18:48 -0400, Tom Lane wrote:
> >> [ consults CVS... ]  So XZ support in F-11's rpm is less than a week
> >> old, there is *no* support in F-10, and we're already requiring
> >> the capability in order to do useful development work?
>
> >> All I can say is WTF.
>
> > An rpm with Xz support is in F11 updates-testing. I'm pretty sure it's
> > also in updates-testing for F10.
>
>  It is not in F-10's package CVS, let alone in updates-testing.  Yes,
>  I looked before complaining.  This was not well-managed.
>  
Is this such a big issue to whine about? The same thing happened to
many of us. Just grab the SRPM and build your RPM.
Here, I did it for you:
http://koji.fedoraproject.org/koji/taskinfo?taskID=1571414

Orcan

[Dan Horák]

Tom Lane píše v So 01. 08. 2009 v 00:09 -0400:
>  Bastien Nocera <bnocera(a)redhat.com&gt; writes:
>  > On Fri, 2009-07-31 at 18:48 -0400, Tom Lane wrote:
>  >> [ consults CVS... ]  So XZ support in F-11's rpm is less than a week
>  >> old, there is *no* support in F-10, and we're already requiring
>  >> the capability in order to do useful development work?
>  
>  >> All I can say is WTF.
>  
>  > An rpm with Xz support is in F11 updates-testing. I'm pretty sure it's
>  > also in updates-testing for F10.
>  
>  It is not in F-10's package CVS, let alone in updates-testing.  Yes,
>  I looked before complaining.  This was not well-managed. 
rpm with xz support for F-10 is available at http://fedora.danny.cz/rpm/


		Dan

[Daniel P. Berrange]

On Fri, Jul 31, 2009 at 08:41:17PM -0400, Josh Boyer wrote:
>  On Fri, Jul 31, 2009 at 06:48:54PM -0400, Tom Lane wrote:
>  >"Mathieu Bridon (bochecha)" <bochecha(a)fedoraproject.org&gt; writes:
>  >>> ERROR with rpm_check_debug vs depsolve:
>  >>> rpmlib(PayloadIsXz) is needed by exim-4.69-12.fc12.x86_64
>  >
>  >> I had the same on F11, building in a Rawhide mock.
>  >
>  >> I was advised to update rpm to 4.7.1 that was in updates-testing,
>  >> which fixed the problem.
>  >
>  >[ consults CVS... ]  So XZ support in F-11's rpm is less than a week
>  >old, there is *no* support in F-10, and we're already requiring
>  >the capability in order to do useful development work?
>  >
>  >All I can say is WTF.
>  
>  Did you have a better plan for migrating to an XZ payload for F12 in time
>  for Alpha? 
IMHO there should be a much larger window between providing an new 
feature in RPM, and requiring it for development. eg, new features
should go into RPM in rawhide & F11 at least 2-3 months before we 
require them. Yes this would have required XZ support to be merged 
much sooner in the F12 schedule, or alternatively merge XZ support
in F12, but don't use it till F13. 

Daniel
-- 
|: Red Hat, Engineering, London   -o-   http://people.redhat.com/berrange/ :|
|: http://libvirt.org  -o-  http://virt-manager.org  -o-  http://ovirt.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: GnuPG: 7D3B9505  -o-  F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

[Tom Lane]

"Daniel P. Berrange" <berrange(a)redhat.com&gt; writes:
>  On Fri, Jul 31, 2009 at 08:41:17PM -0400, Josh Boyer wrote:
> > On Fri, Jul 31, 2009 at 06:48:54PM -0400, Tom Lane wrote:
> >> [ consults CVS... ]  So XZ support in F-11's rpm is less than a week
> >> old, there is *no* support in F-10, and we're already requiring
> >> the capability in order to do useful development work? 
> >> All I can say is WTF. 
> > Did you have a better plan for migrating to an XZ payload for F12
> in time
> > for Alpha? 
>  IMHO there should be a much larger window between providing an new 
>  feature in RPM, and requiring it for development. eg, new features
>  should go into RPM in rawhide & F11 at least 2-3 months before we 
>  require them. Yes this would have required XZ support to be merged 
>  much sooner in the F12 schedule, or alternatively merge XZ support
>  in F12, but don't use it till F13.  
I would be satisfied if there was any window at all.  The upthread
suggestions that it's okay to expect every individual Fedora packager
to cope with this for themselves were insane, not to say insulting.
I have real work to do, and dealing with that sort of make-work is
not it.

			regards, tom lane