Tom Horsley asked about the status of these ``experimental'' aspects of the
language.  Ron Guilmette posted a somewhat cryptic reply, which I'd like to
expand a little.

At the July meeting of X3J16 in Seattle, the committee decided to adopt the
proposed template mechanism into the working draft for the standard.  At last
week's meeting in San Francisco, the committee decided to adopt the proposed
exception handling mechanism.  In each cases, ``the proposal'' was the text of
the appropriate chapter in ``The Annotated Reference Manual'' by Ellis and
Stroustrup (net jargon: either ARM or E&S).

To the extent that such action ``defines'' the C++ language, these decisions
make the features part of the language.  At the very least, they indicate that
the syntax and semantics as described in the book are likely to end up in the
final standard -- vendors of C++ implementations can begin work on these feature
with some confidence that the definitions will remain (relatively ;-) stable.

Of course, we'll need to provide feedback on these definitions, based on our
experience implementing and using the features.  The sooner these features show
up in C++ implementations, the more time the committee will have to incorporate
the feedback.

--
Mike Vilot,  ObjectWare Inc, Nashua NH
mjv@objects.mv.com  (UUCP:  ...!decvax!zinn!objects!mjv)