Sam wrote:
> It appears that IMMEDIATE_FFLOAT requires TYPECODES. why?
Because this requires extra #ifdef in the code. And since IMMEDIATE_FFLOAT
can only work on 64-bit machines and they all use TYPECODES by default,
it's useless to add IMMEDIATE_FFLOAT support for HEAPCODES.
> does this means that on a 64-bit machine with HEAPCODES IMMEDIATE_FFLOAT
> is not defined and every single-float has to be heap allocated?
Yes. But don't bother: it's an unusual configuration.
Bruno
|
