From: matz@... (Yukihiro Matsumoto)
Date: 2004-04-19T17:46:40+09:00
Subject: Re: proposal Hash eql? , ==, ===

Hi,

In message "proposal Hash eql? , ==, ==="
    on 04/04/16, walter@mwsewall.com <walter@mwsewall.com> writes:

|Hash#== should be for would be content equals and not
|content and default value/proc equal. I think that most common case 
|for == would be content equality.
|
|Hash#=== as Nobu pointed out === would make more sense to be a 
|membership test, i.e., alias for Hash#key?.
|
|Hash.eql? should be content equal and default value/block (i.e what 
|the current == is now).  There is some precedence for this in Numeric 
|where 1 == 1.0 is true but 1.eql?(1.0) is false.  I see this as being 
|very similar circumstance.

I agree with both Hash#== and Hash#eql?  But I'm not very positive
about making Hash#=== a membership test.  Some usage of "===" are
membership, for example, Array and Class, but they are not general
consensus.  I'm not sure how much useful that making === membership.

							matz.