RMS's response to "Re: ARPA still undermining Ada"

[kenner@lab.ultra.nyu.edu (Richard Kenner)]

In article <Cy5nxM.pn@research.att.com> bs@research.att.com (Bjarne Stroustrup <9758-26353> 0112760) writes:
>That description does not square with my recollection.
>
>After a talk I gave a long time ago (8 or 10 years ago, I think)
>at some Californian university, a wild looking guy walked up to
>me wrote half a dozen suggested changes to C++ on the blackboard
>and demanded (not suggested) I adopt them.  ... The guy yelled
>``I had not expected such a hostile reaction!'' did some strange gestures
>with his arms and hands, and walked off.
>
>I was somewhat startled by this unusual behavior - that is why I
>remember the conversation. People who had been standing around
>listening told me that that the guy was Richard Stallman. 

I forwarded this to RMS and here's what he had to say:

[I asked kenner to post this for me because I don't do netnews
myself.]

Today I received a forwarded news posting by Stroustrup, and read that
I had "demanded (not suggested)" certain changes in C++.  I think I
should say what actually did happen.

When I met Stroustrup, I had been concerned for some time about
certain minor aspects of the grammar of C++ which cause major
practical problems (ambiguity and gratuitous incompatibility with C).
These particular grammar points don't contribute anything to the
salient features of C++ because they are too minor.

I considered these serious problems and I was glad to have a chance to
talk to Stroustrup about them.

I did not, however, make a demand.  I expected he would share my
concern about the problems.  I expected he would do something about
them, not because I said so, but because it would be good for the
users.

I believed firmly in the importance of solving these problems.
Stroustrup may have mistaken firmness of belief for a demand.

But Stroustrup did not regard gratuitous incompatibility and ambiguous
grammar as problems and had no interest in these changes.  I went away
and decided not to use C++, at least in its current form.

I considered implementing a cleaned up version of C++ syntax myself,
with a -bs option to select the Bjarne Stroustrup syntax.  I spoke
about this idea at Tri-Ada.  But I never implemented this--I was too
busy with other things.

It may be too late now to consider changing these things.  But if you
*want to do the work* of implementing a rationalized C++, please
contact me, and I'll try to remember the issues which I've had several
years to forget.  At least it would be fun to have an excuse to have
an option named -bs.

(Please don't ask if you aren't prepared to do the work--I don't want
to spend time discussing these grammar issues unless it does practical
good.)