On 09/07/2009, Michael Gogins <
michael.gogins@...> wrote:
> This is a good idea, and I have thought about it myself. It would take
> a fair amount of work, but it might be well worth while. There are
> several issues
>
> (1) The opcodes might have to be "wrapped" in some abstraction layer,
> like a C++ virtual class for opcodes that exposes all the opcode
> hooks. But the existing opcode table might actually be enough, as it
> already contains all the function pointers and in/out type codes. So,
> such an abstraction layer could be quite thin, a skinny C++ class that
> plugs the opcode structure and function pointers into a more
> accessible form, perhaps with template metaprogramming to
> automagically create getters/setters for all the in/out values.
>
> (2) Many opcodes call back into Csound, and often directly access
> objects in Csound. So, that part of the Csound "host:" would have to
> be re-implemented in a wrapper by the new host. Actually, I don't
> think doing this is necessarily so hard, especially since a lot of
> opcodes do this via the Csound API now, and the rest could be changed
> to use the API. Then all that is required is to re-implement those
> Csound API calls in the new host. A lot of this is easy - rates,
> messaging, etc. Function table access, not so hard. Memory management,
> a bit harder. I don't know what else would be required.
Binary compatibility with the real csound if you don't plan on
rebuilding the opcodes for each new host.
Saludos,
Felipe Sateler
------------------------------------------------------------------------------
Enter the BlackBerry Developer Challenge
This is your chance to win up to $100,000 in prizes! For a limited time,
vendors submitting new applications to BlackBerry App World(TM) will have
the opportunity to enter the BlackBerry Developer Challenge. See full prize
details at:
http://p.sf.net/sfu/Challenge_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel