Prototype or production?

[VHD]

>> >If you
>> >need some speed enhancements you can always write or use an external 
>> >command or a shell application to do the heavy processing.
>> 
>> Being able to do this is what makes RR even more interesting. 
>> I like the idea of delagating heavy processing stuff to 
>> another language. (But then again I have read that transcript 
>> is actually quite fast.)
>
>It's not fast enough for Neural Nets for example but it's fast enough
>to manage a lot! And you have to be careful how you call your shells 
>or RR will sit waiting until it receives the answer from the shell. 


Now that you mention Neural Nets, have you tried linking RR with octave, scilab or matlab?


Ben