[gclist] C++ proposal

Jim Larson larson@kesey.jpl.nasa.gov
Wed, 29 May 1996 10:28:58 -0700


In message <199605290314.NAA10129@goanna.cs.rmit.EDU.AU>
Richard O'Keefe writes:
>Drat.  If this proposal is accepted *and implemented* it will remove
>my biggest objection to C++.

Feh.  Surely the fact that the language isn't well-defined ranks a
close second?  (Of course, C isn't well-defined either, but at least
you can hold the entire language in your head and reasonably intuit the
*spirit* of the standard where the *letter* falls short.)

>On the other hand, it has been explicit for some time that Ada *may*
>be garbage collected, and where are the garbage collecting Ada 
>implementations?  Perhaps it will be the same with C++.

I'd guess that each C++ vendor would like to add the "GC" bullet item
to their feature list.  Of course there's no reason to think that the
different implementations will be compatible.  Look at what happened to
"templates".

Hans Boehm and others have done amazing work to retrofit GC into
uncooperative environments.  But don't we get so much more out of GC
through the indirect benefits it has in allowing powerful languages,
reusable libraries, and modular code?

It frustrates me to see so much effort spent to graft yet another
antler or tentacle onto C++ while languages like Scheme and ML languish
in academia waiting for a production-quality implementation.

Jim Larson                      Jet Propulsion Laboratory
james.s.larson@jpl.nasa.gov     #include <disclaimer.h>