[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: netcdf / g++ Transition

On 07/13/2015 02:19 PM, Emilio Pozuelo Monfort wrote:
> On 13/07/15 14:15, Alastair McKinstry wrote:
>> On 13/07/2015 12:50, sebastic wrote:
>>> On 2015-07-13 13:40, Alastair McKinstry wrote:
>>>> So, whats the current status of the netCDF transition?
>>> We still need to verify all reverse dependencies before we can request
>> a transition slot.
>> Ok, I'll start with my packages then. Experimental builds.
>>>> What about collisions with the gcc-5 transition?
>>> The GCC-5 related transitions are going to be painful, that's why I
>> hope to get the gdal & spatialite transitions done before GCC-5, but it
>> seems unlikely.
>> Yes. This gets confusing.
>> I have a package libdap, upon which gdal depends. Its just undergone a
>> transition 3.12 -> 3.14.0 that included SONAME changes. It has two
>> rdepends, gdal and grads.  I had set an RC bug against libdap to block
> No. There was no RC bug at all.
>> it from transitioning until g++-5 transition (as it has c++11 symbol
>> changes) but its been removed and libdap 3.14.0 has transitioned.
>> gdal has been binNMU'd and will transition.
> Right.
>> So what happens? Does testing get rebuilt with g++-5 ? (unlikely, but a
>> good idea, I think),
> Not going to happen.
>> or is another transition necessary for
>> libdapclient6 -> libdapclient6v5 with the g++ transition?
> You need another transition if the ABI changes with the new libstdc++.

yes, there are a few packages which need both the libstdc++ and the gfortran
module transition (for now I see slepc and metview).  So I would prefer if you
could wait until after that one.  Btw, I'm still looking for somebody who wants
to handle the gfortran module transition ...


Reply to: