Bug#796345: Status report on perl 5.22 transition readiness (30th Sept)

[Dominic Hargreaves <dom@earth.li>]

On Wed, Dec 30, 2015 at 12:49:50PM +0100, Emilio Pozuelo Monfort wrote:
> We're almost there. The last britney run reported:
> 
>     * ppc64el: cyrus-dev, libccs-perl, libcyrus-imap-perl, libdcmtk4-dev,
> libperl5.20, libvtkgdcm-java, libvtkgdcm2.4, linux-perf-4.2, oar-restful-api
> 
> Of those:
> 
> libdcmtk4-dev, libperl5.20, libvtkgdcm-java, libvtkgdcm2.4, linux-perf-4.2 are
> cruft, so should be removed at the end of the run
> 
> oar-restful-api is not in testing as I removed oar. I guess britney is trying to
> add it as my removal hint is for the old version that was in testing. I'll
> update it for the new version that is trying to enter now, so this shouldn't be
> a problem (and in any case, we could remove it afterwards).
> 
> cyrus-dev, libcyrus-imap-perl should be fixed in the next run with my
> cyrus-imapd-2.4 urgent.
> 
> libccs-perl is the one that we would break, as has already been discussed.
> 
> So the only real issue is libccs-perl. I am pondering a `force-hint
> perl/5.22.1-3 mono/4.2.1.102+dfsg2-5' now to finish this. Otherwise someone else
> will have to look at this, or this will have to wait. Any thoughts?

This sounds sane to me. As already discussed the actual user impact on
breaking libccs-perl should be minimal, and it looks like the HA team
are on the way to resolving that. I guess avoiding blockages in
testing until that can be fixed properly is unnecessary.

Thanks for all your work on this!

Cheers,
Dominic.