Bug#918750: transition: simbody
Control: tags -1 confirmed
On 10/01/2019 12:16, Jose Luis Rivero wrote:
> On Wed, Jan 9, 2019 at 10:11 AM Emilio Pozuelo Monfort <pochu@debian.org>
> wrote:
>
>> On 09/01/2019 01:27, Jose Luis Rivero wrote:
>>> Package: release.debian.org
>>> Severity: normal
>>> User: release.debian.org@packages.debian.org
>>> Usertags: transition
>>>
>>> Dear release team:
>>>
>>> simbody 3.6.1+dfsg-1 is now in experimental, we can start the transition
>>> for the existing package in the archive currently using it.
>>> The following source package need to be rebuild:
>>>
>>> gazebo 9.6.0-1
>>>
>>> I think that in terms of 'ben' lingo, the transition has the following
>>> parameters:
>>>
>>> Affected: .depends ~
>> /\b(libsimbody3\.6|libsimbody3\.5v5|libsimbody3\.5v5\-dbg)\b/
>>> Good: .depends ~ /\b(libsimbody3\.6)\b/
>>> Bad: .depends ~ /\b(libsimbody3\.5v5|libsimbody3\.5v5\-dbg)\b/
>>>
>>> Sorry for sending this close to the freeze but it will kill the 2 RC
>> bugs pending on Simbody.
>>> Please schedule binNMUs for gazebo packages on all architectures.
>>
>> simbody failed to build on several architectures:
>>
>> https://buildd.debian.org/status/package.php?p=simbody&suite=experimental
>>
>> Please fix that before we consider starting the transition.
>
>
> I've upload simbody 3.6.1+dfsg-3 which:
> - fixed: all, mips, powerpc, powerpcspe, ppc64el, ppc64
> - waiting but probably fixed: mipsel, mips64el, kfreebad-amd64
> - still failing: i386, hurd-i386
>
> The build is failing on i368 (will require a bit more of work) but it is
> already failing on unstable so there is a big gain on architectures
> supported (+6 at least) and no regression as far as I can say.
My concern here is that the way to fix the build on all those architectures was
by ignoring the failing tests. If the test cases themselves are buggy then
that's fine (though it'd be good to forward that upstream and get the tests
fixed). However the tests may be failing due to bugs in the underlying library
code, in which case ignoring them is not really a fix.
In any case the situation in sid is bad too as you said and I imagine that the
version in testing (which seems quite similar to the one in sid) would be
affected by these build failure problems too, so I guess we should go ahead with
this version.
BTW I assumed that gazebo builds fine against this new simbody, is that right?
If not, that is obviously a blocker. If it builds fine, then go ahead and look
into the remaining build issues.
Cheers,
Emilio
Reply to: