Hello,
Am 17.09.2010 00:59, schrieb Paul C. Anagnostopoulos:
At 9/16/2010 05:32 PM, Andrew Whitworth wrote:
I'm not against creating ops like this, but I would suggest instead
creating them as dynops. In fact, I would be happy to see all the trig
and hyperbolic trig ops moved to dynops, if they aren't already.
The existing operations are dynops. We would simply be adding additional
ones.
Note that all these trig functions are provided by the Complex PMC.
Just as a data point from a language writer: Rakudo defines its own
Complex type, because it was easier to get the dispatch right that way.
Only a few otherwise costly functions are privately re-dispatched to
parrot's Complex PMC.
So we Rakudo developers don't see much need for more trig methods in the
Complex PMC (I don't know if there is a need for real trig
functions/methods, but so far we've successfully worked around the lack
of them). Of course they don't hurt either.
Cheers,
Moritz
_______________________________________________
http://lists.parrot.org/mailman/listinfo/parrot-dev
