On Thu, Oct 16, 2008 at 9:36 AM, Karl Meerbergen <Karl.Meerbergen@cs.kuleuven.be> wrote:
Correction done.

Karl Meerbergen wrote:
> You are right of course. This should be corrected.
>
> Thanks for spotting this
>
> Karl


Thanks for the quick update.  After thinking about this for an hour or so I have another idea.  What about testing something like:

if( transpose_flag != 'N' ){
    transpose_flag = 'T';    // For single or double precision
    transpose_flag = 'C';    // For complex
}

This would allow a templated function to use the same flag for both complex and non-complex types, and have the binding sort it out.  Would this bring bad karma?

Thanks again,
Jeremy