From: Joel VanderWerf <vjoel@...>
Date: 2008-03-28T08:10:51+09:00
Subject: Re: instance_variable_[gs]et

Rados�aw Bu�at wrote:
> On Thu, Mar 27, 2008 at 9:07 PM, Joel VanderWerf
> <vjoel@path.berkeley.edu> wrote:
>>  I've got a couple of methods in library code I use frequently that
>>  initialize or populate an object from a hash using the loop above. I've
>>  always thought it was kind of a pain (and inefficient) to have to
>>  construct a string for each key.
> 
> Does it mean that everyone use/need it? I don't think so.

Don't see what that has to do with anything. There's lots of core 
methods that are rarely used. Have you used 
Module:protected_method_defined? recently?

>>  Why couldn't #instance_variable_set be agnostic about whether you refer
>>  to ivars as "foo" or "@foo" ? I don't see any ambiguity...
> 
> I don't know if it's main reason but AFAIR Ruby internally use
> instance variables without "@" character to store additional
> information about object (but from Ruby you can't get it).

That's a good point, but my suggestion was for #instance_variable_set to 
map both "foo" and "@foo" to the same entry in the table, namely the 
entry for "@foo". Hidden entries would remain hidden.

> I think that most important is consistent. We use @ in code and the
> same goes when we want get value by instance_variable_get method.

Maybe.

-- 
        vjoel : Joel VanderWerf : path berkeley edu : 510 665 3407