From: Jeremy Kemper <jeremy@...>
Date: 2009-11-08T06:19:56+09:00
Subject: [ruby-core:26606] Re: HashWithIndifferentAccess to core

On Sat, Nov 7, 2009 at 12:25 PM, Rick DeNatale <rick.denatale@gmail.com> wrote:
> Are there benchmarks showing that implementations such as in
> ActiveRecord are bottlenecks?

There is no claim to bottlenecking nor is that a necessary condition
for speedup. Hash access is so extraordinarily common that getting it
as fast as possible is a worthy goal, particularly for library code
wishing to lower overhead for its users.

> Are there benchmarks showing that any of the proposed implementations
> really are faster?

See the README at http://github.com/methodmissing/hwia

Speedups:
key?   4.3x
[]     6.8x
[]=    5.1x
dup    5.1x
merge  6.3x

jeremy