[PD-dev] FIXED problem defining t_outlet array

[Mathieu Bouchard]

On Thu, 17 Jul 2003, Brian Sheehan wrote:

> Thanks Mathieu, that worked. Is it so that we never
> have to worry about deallocating memory that
> outlet_new is the only way of creating t_outlet
> structs?

There may be several reasons. That may be one of those. If there were
outlet_init() that would not allocate but just initialize, then there
would be a need for an outlet_finalize() that deinitializes but without
deallocating, and maybe the memory allocation would become the caller's
responsibility, or else there would have to be a flag in t_outlet to
determine which way the object should be gotten rid of, finalization or
also deallocation. That would be somewhat more difficult for almost no
reward. That is, unless completely automated deallocation (refcount or
marksweep) is implemented in PD, but I don't think that will ever happen.

Else it allows the size of t_outlet to change in PD without recompiling
all externals.

________________________________________________________________
Mathieu Bouchard                       http://artengine.ca/matju