mathlink unsigned types and bit sizes

[Art]

I have some related mathlink questions that I haven't found answers
for in the documentation. Answers to any would be great, thanks. Sorry
for posting several questions at once:

1. What is the best way to Put unsigned types. In particular, I have a
large 64-bit unsigned multidimensional array? I can't use
MLPutInteger64Array, right?

2. If I want to get Real or Integer packed arrays using, for example,
MLGetReal64Array, how do I know which bit size to use, 32, 64, 128?

3. Why can't we use the *data portion of the arrays that are sent to
mathlink without copying first and Disowning if we don't change those
arrays?

Art.

[Art]

[Art]

[Jens-Peer Kuska]

Hi,

1) Mathematica does not know an unsigned type and you have to send int.

2) MLGetReal64Array() are 64 bit reals, i.e. double ??

3) it is up to you when you disown the data. A function that
is called via the template mechanism never disown the array
itself.

Regards
Jens