Bug#836447: transition: proj
On 05/09/16 14:37, Sebastiaan Couwenberg wrote:
> On 09/05/2016 10:54 AM, Sebastiaan Couwenberg wrote:
>> On 09/04/2016 05:32 PM, Sebastiaan Couwenberg wrote:
>>> On 09/04/2016 12:02 PM, Emilio Pozuelo Monfort wrote:
>>>> On 03/09/16 21:41, Sebastiaan Couwenberg wrote:
>>>>> proj (4.9.3-1) has been uploaded to unstable, and has been built on all
>>>>> release architectures. The ports need a little more time.
>>>>
>>>> binnmus scheduled.
>>>
>>> Many thanks for those.
>>>
>>> In general the rebuild look good so far. The binNMUs did reveal a few
>>> packages with issues on some of the architectures.
>>>
>>> postgis FTBFS due to test failures caused by the fakeroot fix for
>>> #830912 not being used yet. Those builds should succeed when they use
>>> fakeroot (1.21-2). Can dep-waits be set for those?
>>
>> postgis still needs to be rebuilt with fakeroot (1.21-2). The following
>> should do, perhaps the libproj-dev dependency should be specified too:
>>
>> dw postgis_2.2.2+dfsg-4 . ANY . -m 'fakeroot (>= 1.21-2)'
>>
>>> mapserver FTBFS on mipsel due to a fatal error in the Java Runtime
>>> Environment. I think retrying the build on a different buildd should let
>>> that succeed too, mipsel-manda-01 had issues with mapserver before.
>>
>> mapserver is now OK thanks to the rebuilt on eberlin.
>>
>>> qlandkartegt FTBFS on some architectures where char is probably unsigned
>>> by default. I've added a patch which should fix the issue, but still
>>> needs verification on the porterboxes. I'll upload a new revision if the
>>> builds on the porterboxes confirm the fix.
>>
>> qlandkartegt is now OK too with the changes in 1.8.1+ds-7.
>>
>>
>> The affected packages in experimental need to be rebuilt with proj 4.9.3
>> too:
>>
>> grass (7.0.5~rc1-1~exp1)
>> mapnik (3.0.12~rc6+ds-1~exp2)
>> spatialite (4.4.0~rc1-1~exp1)
>>
>> python-mapnik (1:0.0~20160810-3a60211-1~exp1)
>> spatialite-tools (4.4.0~rc1-1~exp1)
>>
>> I think the attached wanna-build commands should do for the above
>> packages in experimental.
>
> vtk6 on armel still hasn't starting building yet, can the priority be
> increased to hopefully prevent another entanglement issue like near the
> end of the gdal transition (#823335)?
There's no rush here. gdal is different because there all packages need to
migrate at the same time.
I have given back vtk6/mipsel, let's see if that was a transient issue.
Cheers,
Emilio
Reply to: