From: Robert Feldt <feldt@...>
Date: 2002-08-29T18:55:57+09:00
Subject: Re: ANN: RJudy-0.1 - Judy Arrays for Ruby

On Thu, 29 Aug 2002, Tom Sawyer wrote:

> > Bad news everyone...
> > This, and insertion being faster with Judy*, is an illusion :-(
> > Read below.
> >
> > It's a pity it isn't quite there yet, but we still have a lot of place
> > for optimization :-)
>
> sad to hear, it seemed we were looking a significant speed increase for
> ruby in this area. but perhaps optimizations will still do it.
>
> recently someone pointed out to me that Arrays were an order of
> magnitude slower then Hashes. that suprised me. has any of this work
> sparked notions on how to bring Array more upto speed with Hash? That,
> at least, would be a nice boost. just some thoughts.
>
In this discussion you should probably also consider the "Ideal Hash
Trees" by Phil Bagwell. He has indicated to me in personal communication
that they should have even better properties than Judy-style arrays/hashes
for some applications. I don't have the IHT papers fresh in memory but
someone (interested in Ruby internals or doing a fast hash extension)
might wanna take a look:

http://lampwww.epfl.ch/publications/papers/idealhashtrees.pdf

Might be some long-term optimization opportunities in there...

Regards,

Robert Feldt