[gclist] Re: gclist-digest V1 #43

[Robert A Duff]

Guillermo (Bill) J. Rozas <gjr@hplgr2.hpl.hp.com> says:

> However, the informal definition should be adequate.

Yes, I suppose so.  I apologize for being overly pedantic in my previous
messages.  The tail-recursion example convinced me.

> Didn't Pascal try this?  I seem to recall that there was new but no
> way to get rid of such structures.  Implementations provided different
> mechanisms (dispose, mark/release, etc.), but to my knowledge none of
> them implemented a true gc.

I believe that "dispose" is a standard feature of Pascal, all the way
back to Jensen and Wirth Pascal.  (I mean, to the extent that Pascal is
or was a standard.)  So, in this regard, it's just like C or C++ or Ada
-- there's a manual method, so implementers don't *need* to provide gc,
and they will only if their customers demand it, or they're inclined to
think gc is a Good Thing, or whatever.  Pascal isn't like Lisp, in that
you really can't get away with a non-gc implementation of Lisp.  (At
least, not unless you have GOBS and GOBS of memory -- more than all of
us put together have on all our machines.)

- Bob