From: "Brett H. Williams" <brett_williams@...>
Date: 2003-10-09T06:22:44+09:00
Subject: Re: Ruby Patriotism: Python+XML v. Ruby+YAML

On Oct  9, why the lucky stiff wrote:
> We've got a good old-fashioned derby going on in blogoland.  Perhaps
> some of you fully-vested shareholders can give me a hand here.
> 
> Some disparaging Ruby+XML benchmarks from yesterday:
>   http://www.intertwingly.net/blog/
> 
> And now I'm making a case for the speed of Ruby+YAML:
>   http://www.whytheluckystiff.net/
> 
> So, if this gets heated, I'll end it pronto.  This isn't name-calling or
> brute campaign (I hope).  A friendly face-off.  Like a pick-up game of
> basketball.
> 
> Here's what you can do: 
> 
>   - I've opened a Wiki page. [http://www.intertwingly.net/wiki/pie/ThePickupGame]
>     Add your suggestions for improving our chances in the running.
> 
>   - Check out the scripts [http://www.whytheluckystiff.net/ruby/yaml-vs-atom/]
>     and take a stab at improving the converter.
> 
>   - Perhaps someone can better represent Ruby+XML?  Can we benchmark
>     Ruby+libxml2?  Any other good XML libs out there?

Ruby+libxml2 certainly works well, and borrows a lot of the nice API
concepts from REXML.  Certainly it should be benchmarked if you're trying
to examine Ruby and XML together.

It's only drawback (other than library dependency on libxml2) is that it is
not documented.  You just have to search the modules/classes and see what
methods are defined and figure out how to use them.

Certainly if you just want to throw out a number for parsing the file (like
for Python2.2 + libxml2) it's really easy to do.