Garth N. Wells wrote:
> Anders Logg wrote:
>> Yes, we really need it (for the SWIG interface).
> 
> OK.
>>
>> But since it is not visible in the interface we can name it anything 
>> we like (doesn't need to look nice). How about real_array?
>>
> 
> Is it work making it a template? It could be useful when communicating 
> real and integer arrays between processes with MPI. Or would this make 
> it unsuitable for SWIG?
> 
> Garth

I think a template should be ok. We should only need to change the 
typemap (directorin) in typemaps.i.

/Anders


>> /Anders
>>
>>
>> Garth N. Wells wrote:
>>>
>>> Garth N. Wells wrote:
>>>> Åsmund Ødegård wrote:
>>>>> 515>hg update
>>>>> abort: case-folding collision between src/kernel/common/dolfin/Array.h
>>>>> and src/kernel/common/dolfin/array.h
>>>>>
>>>> I don't see a file array.h, only Array.h
>>>>
>>>> Garth
>>>>
>>>
>>> I see that it was introduced in the last changeset.
>>>
>>> Do we really need array.h? It looks to me like one could easily make 
>>> a mess of it. The only functionality it provides over a plain array 
>>> is that it knows its length.
>>>
>>> Garth
>>>
>>>
>>>>> There are still computers out there that does not behave as they
>>>>> should when files have names differing only in case. To be more
>>>>> precise, to support OSX, either array.h or Array.h must be removed!
>>>>>
>>>>> mvh,
>>>> _______________________________________________
>>>> DOLFIN-dev mailing list
>>>> [email protected]
>>>> http://www.fenics.org/mailman/listinfo/dolfin-dev
>>>>
>>> _______________________________________________
>>> DOLFIN-dev mailing list
>>> [email protected]
>>> http://www.fenics.org/mailman/listinfo/dolfin-dev
>>

_______________________________________________
DOLFIN-dev mailing list
[email protected]
http://www.fenics.org/mailman/listinfo/dolfin-dev

Reply via email to