From: "David A. Black" <dblack@...>
Date: 2005-09-29T02:24:07+09:00
Subject: Re: creating independent lambdas in loops

Hi --

On Thu, 29 Sep 2005, Eric Mahurin wrote:

> --- "David A. Black" <dblack@wobblini.net> wrote:
>> On Thu, 29 Sep 2005, Eric Mahurin wrote:
>>
>>> It still seems like there should be a way to control
>> variable
>>> locality in blocks better.  Making the variable local just
>>> based on whether the variable already exists in the outer
>> scope
>>> doesn't cut it.
>>
>> It's never bothered me to view block variable assignment as
>> assignment to existing variables, but I know I'm in a small
>> minority
>> which, among other things, does not include Matz :-)  You can
>> expect
>> this to change in Ruby 2.0.
>
> Any idea what the planned change is going to be?  New construct
> like:
>
> localize(:a,:b) {...}

Hopefully nothing quite that obviously "added as an afterthought to
the language" :-)

> Or maybe an incompatible change like making all block arguments
> local no matter what.

See http://www.rubyist.net/~matz/slides/rc2003/mgp00010.html and the
slides following that one.  I think the idea would be:

   a = 1
   method_call {|a| a = 2; b = 3 }
   puts a  # 1
   puts b  # 3

> It would be nice if all definite changes were put on rcrchive
> and marked as accepted.  And somehow put the potential changes
> in there too.  Or is there another place where this information
> is held?

That's really not what RCRchive is for; it's for the processing by
Matz of change requests from other people.  I think discussion of
possible changes coming from Matz is spread around -- Matz's blog,
ruby-core, here, the TODO file, etc.


David

-- 
David A. Black
dblack@wobblini.net