On 2021-05-31 14:45 +0200, Marc Nieper-Wißkirchen wrote:
> PS A functional updater should never allocate a new structure unless
> necessary. The semantics of SRFI 113, SRFI 146, ... regarding the question
> of what "newly allocated" means, are not very good and should eventually be
> corrected.

This is a good point that I'd like to address in SRFI 224 (still a
draft).  How can we get this right?  I'm inclined to discard the
"newly allocated" terminology, which I agree is unclear.

The point about the difficulty of providing both efficient
mutating and functional implementions seems critical; thanks to
Shiro for raising it.

-- 
Wolfgang Corcoran-Mathe  <[email protected]>

"Earth is the cradle of the mind, but one cannot live in a
cradle forever." --Konstantin Tsiolkovsky

Reply via email to