How far can stack [LIFO] solve do automatic garbage collection and prevent memory leak ?

[Paul Rubin]

Elizabeth D Rather <erather at forth.com> writes:
>> Processors seldom could multitask, so it wasn't recognized that the
>> stack could be a performance bottleneck
> Lol.  Forth supported multitasking on every processor it was
> implemented on in the 70's, with blazing speed compared to competitive
> techniques. I have never seen stack operations to be a bottleneck.

I think "multitasking" in that post refers to superscalar execution,
which wasn't done in the 1970's except on supercomputers.  That the
stack is a bottleneck is the precise reason that optimizing Forth
compilers do complicated flow analysis to translate stack operations
into register operations.