Re: what about PTS (Re: libunwind and ia64)

["Adam D. Barratt" <adam@adam-barratt.org.uk>]

On 2014-01-27 15:19, Andreas Beckmann wrote:
> On 2014-01-27 14:12, Osamu Aoki wrote:
> > It seems that this "auto-*" transition should not block upload if old
> > libunwind7 library is used.  Especially for packages indirectly
> > depending on one of the auto-* packages.  That is good.
>
> Looks like the transition tracker should get an optional flag "ignore"
> for the transitions - and the auto-libunwind one should definitively 
> get
> this flag as it obviously causes lots of confusion :-)
> And thereafter the PTS needs to be taught about "ignored" transitions.

That functionality already exists, and is why the manually created 
libunwind transition doesn't show up on the PTS; it's just that the 
software creating the auto-* transitions doesn't support it yet.

[...]
> And regarding the icu52 transition: The transition tracker is not smart
> enough to detect that a package has finished the transition (i.e.
> testing has a version satisfying the "good" constraint.). But this does
> not happen too often.

The reason that the icu transition hasn't been closed is that libicu48 
is still in testing, so the transition is /not/ finished.

A list of packages which still need updating in (or removing from) 
testing can be found in the britney log or by "dak rm -Rn -s testing -b 
libicu48" on mirror.ftp-master.d.o.

Regards,

Adam