From: Dave Thomas <Dave@...>
Date: 2000-12-23T02:36:37+09:00
Subject: [ruby-talk:7969] Re: defined? problem?

Kevin Smith <sent@qualitycode.com> writes:

> Hold it right there. Is it really valid to 
> "remember the fact that m is a variable"? It 
> seems like that decision should be held off until 
> runtime.

Well, what happens is not that it remembers that m is a variable, but
that it has seen m being treated as a variable.

Holding it off until runtime also leads to undesirable side-effects.

My feeling here is that we're making a compromise. For the convenience 
of allowing parameterless method calls without parentheses, we
occasionally force the interpreter to silently resolve ambiguities
during the parse.

I very much liked the idea of issuing a parse-time warning if the
interpretation of a symbol within the same scope changes. That would
probably overcome most problems (after all, we all _do_ run our
programs with -w...)


Dave