py26-numpy +gcc44 requires gcc45 variant of atlas

[Michelle Gill]

Dear Rainer,

Thank you for your quick reply.
> 
> Usually all scientifc ports requring gfortran should be kept in sync and
> use the same version of gcc. Recently in r71201 [1] atlas have been
> changed to +gcc45 by default because of build problems on Macs with a
> Intel Core i7 processor, see #24869 [2].
> 
> Marcus, James, and others maintaining scientific ports, please discuss
> how to resolve this issue. Either going back to gcc44 or switching *all*
> ports in category science to gcc45.

I missed the issue about atlas going from gcc44 to gcc45, though I saw the one about it going from gcc43 to gcc44. I will watch for a decision on standardization of preferred gcc version and adjust my installation appropriately.

> +universal means building for multiple architectures and has nothing to
> do with the compiler being used. Usually you don't need +universal
> except you have multiple machines to which you will transfer archives.
> Als a dependent might not build for your current architecture (e.g. wine
> does not build x86_64), but MacPorts will then (re-)install with
> +universal automatically.

I understand. I have been incorrect in my assumption of what universal builds are for--it seems they are for architecture independence (the more standard use of the term "universal") rather than independence of compiler. Thank you for the clarification.
> 
> sudo port install atlas +gcc44
> sudo port install py26-numpy
> 

I've confirmed this works. 

Thanks again,
Michelle Gill