From: Nathaniel Talbott <nathaniel@...>
Date: 2003-08-13T11:20:02+09:00
Subject: Re: Nested class/module namespace

Christoph R. [mailto:chr_news@gmx.net] wrote:

> If this does not benefit you just don't use the new feature.

Well, I'd like the language to improve if possible... and it seems that the
new feature's usefulness is not quite what it could be. If it can be
improved, I think that would be great, don't you?

  
> Remember the protracted trouble with the scoping rules of 
> Ruby's class variables (it took Matz a fairly long time to 
> get things right).  Ruby's scoping are already complicated 
> enough (especially in combination with eval scope).

Really? I've never had any trouble with it. What causes you difficulty?


> The feature you are requesting run IMHO squarely against the 
> current scoping rules and might push them over the edge (of being too
> complicated) ...

How can that be so, when there is already a construct that behaves exactly
like I want the new construct to behave? All I'm wondering is why the new
construct doesn't behave like the old one... I still haven't seen any
concrete examples of why the current behavior is good.

If the reason is implementation difficulty, I can understand that... but I
worry that the differences between

  module M
    class C
    end
  end

and

  class M::C
  end

will confuse many a newbie (and some oldies, too).

Can you help me understand why the current behavior is better?


Nathaniel

<:((><