Issue: PATHNAME-SYMBOL (Version 2)

[David A. Moon <Moon@STONY-BROOK.SCRC.Symbolics.COM>]

    Date: Tue, 2 Jun 87 00:11 EDT
    From: Kent M Pitman <KMP@STONY-BROOK.SCRC.Symbolics.COM>

    The grinding of the PATHNAME-SYMBOL proposal needs to be cleaned up.

This happens to most mail that passes through Xerox.  I don't think they
see it in their own mail reader, but in the outside world it gets
reformatted as alternating long and short lines if the lines are longer
than a small maximum.

    By the way, in the Symbolics release I'm using (7.1) allows (LOAD 'symbol)
    so the comment saying we've long since flushed the feature is not completely
    accurate. 

Try ZL:LOAD if you want to see what it does in Zetalisp.  But CL:LOAD doesn't
allow symbols in 7.1 for me, so maybe you've "fixed" it in your init file?
I don't think I've "fixed" it in mine.

	      It may be that Moon didn't mean to include LOAD, in which case
    the discussion should be phrased to make it clear that LOAD is not intended.

CLtL already does not say that LOAD accepts symbols as arguments, unless we
are to think that when it says "filename" it means "pathname", and when it
discusses what arguments named "pathname" mean 13 pages earlier, it means to
include LOAD.

I certainly didn't mean to propose to change the language to make LOAD allow
symbols.

Does this mean that we need to withdraw PATHNAME-SYMBOL in favor of a more
comprehensive proposal to fix all the ambiguities and incomplete specifications
in chapter 23?  That would be nice, but it's a big job.  Maybe we could just
cover every argument that is called pathname, filename, file, or new-name
and leave the rest of the ambiguities for another proposal?