EQ? non-essential?

[Bill Rozas <JINX%OZ.AI.MIT.EDU%xx.lcs.mit.edu@CSNET-RELAY.ARPA>]

    I also disagree strongly with JAR's latest suggestion that EQ? be made
    non-essential.  Doing so doesn't solve the problem, it just passes the
    buck!

I disagree.  The point of the agreement that JAR and I reached (I am
the unnamed flamer) was that implementors would have the choice of
having EQ? or a compiler which can split without thinking about the
consequences.  The essence of my (and GJS's) position is that if
splitting is allowed, the language has changed to the point we don't
feel comfortable with it anymore.  The difference in our positions is
that I'm willing to do away with EQ? as long as if it is implemented
it has the properties that I consider essential.  GJS disagrees since
he argues that there is no way to obtain that power if it is not
provided.

When I originally answered JAR's poll, I answered I to the relevant
questions.  After rhh sent his message, I revised my position after
thinking about it again.  One of the things which I've always liked
about Scheme is the ease with which one can do things like that, which
I believe are intuitive and natural.  Adding locations explicitely
into the code just to make EQ? work (as was suggested to me) seems
artificial and no better than emulating closures in languages which
lack them (like C).