Re: [Boost-mpi] [EXTERNAL] Re: Request

Subject: Re: [Boost-mpi] [EXTERNAL] Re: Request
From: Belcourt, Kenneth (kbelco_at_[hidden])
Date: 2016-10-31 19:13:58

Yes, but this one’s a serialization problem, I think.  These also need to be fixed.

Noel

> On Oct 31, 2016, at 5:10 PM, alainm <alain.miniussi_at_oca.eu> wrote:
> 
> Is it this one ?
> intel-linux.compile.c++ ../../../bin.v2/libs/graph_parallel/test/distributed_property_map_test-2.test/intel-linux/debug/distributed_property_map_test.o
> ../../../boost/serialization/access.hpp(116): error: class "std::vector<boost::parallel::detail::untracked_pair<remote_key_to_global::key_type, color_t>, std::allocator<boost::parallel::detail::untracked_p
> air<remote_key_to_global::key_type, color_t>>>" has no member "serialize"
>          t.serialize(ar, file_version);
> 
> 
> 
> On 31/10/2016 23:12, Belcourt, Kenneth wrote:
>> Hi Folks,
>> 
>> It seems changes have been made to MPI that have broken the graph_parallel library in develop.  Could I ask that if you’re going to commit code to MPI, that you always test it first with both graph_parallel and MPI, and that you not push broken code into develop?  I use graph_parallel from develop and I’m now unable to merge graph changes to master because I can’t test them with graph_parallel on develop.  If I’ve misunderstood something, please let me know.
>> 
>> Noel Belcourt
>> 
>> _______________________________________________
>> Boost-mpi mailing list
>> Boost-mpi_at_[hidden]
>> http://lists.boost.org/mailman/listinfo.cgi/boost-mpi

> 
> -- 
> Alain Miniussi
> Pole Génie Logiciel Scientifique
> Observatoire de la Côte d'Azur
> Blv de l'Observatoire, Nice
> 
> _______________________________________________
> Boost-mpi mailing list
> Boost-mpi_at_[hidden]
> http://lists.boost.org/mailman/listinfo.cgi/boost-mpi