Re: RFC 135 (v2) Require explicit m on matches, even with ?? and // as delimiters.

[Tom Christiansen]

>Is it accurate to say that the list of things that would have to be 
>addressed is table 3-3 on p. 97 of Camel III (unlabelled table on p. 84 of 
>Camel II)?

Well, mostly.  (Note that <FH> for readline(FH) also is <*.c) for
glob('*.c')).  I suppose we could forbid pick-your-own-quotes.

But remember how stultified tihngs like emacs have historically
gotten by s'foo'bar' and such.

There are unsolvable problems here, though.  Consider

    print rand +3

versus

    print time +3

The unary is

    print rand(+3)

but the zilchary (nilary?  nihilary?) is 

    print time()+3

Now consider

    use Frobnitz;
    print frob +3;
    print nitz +3;

Pure lexical analysis of those three lines cannot reveal the
parse, as we don't know the arialities of those operators/functions.

A simple solution is to require all function calls (and things that
are like functions, such as built-in CORE opcodes) to have parentheses
around their arguments, *always*.  Such a requirement would certainly
make us friends in many camps.  And enemies,  no doubt, in others.

--tom