On Tuesday 20 March 2001 17:10, Nick Bensema wrote:
> In article <3AB7921F.B26BB1E / hurstlinks.com>,
>
> Guy N. Hurst <gnhurst / hurstlinks.com> wrote:
> >I don't see a problem here, because I only meant
> >for the '.' to be higher precedence than a space.
> >If there is no space, than the lack of it is higher
> >than the '.'
> >
> >Foo.new(1).bar("a")  #-> (Foo.new(1)).bar("a")
> >Foo.new (1).bar("a") #-> Foo.new((1).bar("a"))
> >
> >Can you think of any problems still?
>
> I think this makes sense.  After all, since we're giving the option
> of using whitespace in lieu of parens, we might as well make that
> whitespace mean the same thing as parens.  Maybe.  This is just at
> first glance, of course.  It sends the visual message that the
> low-precedent code on the right is separate from the high-precedence
> code on the left, in the invisible parentheses.  But...
>
> What about people who like to do things like this:
>
>   Foo.new 1, (1).bar("a"), 2


To construe any meaning whatsoever from whitespace (or lack thereof) is 
probably a dangerous thing.

Regards,

Kent Starr
elderburn / mindspring.com

(Notoriously inconsistent user (abuser) of whitespace!)