From: Timothy Byrd <byrd.timothy@...>
Date: 2005-11-19T06:47:20+09:00
Subject: Re: embedded ruby limitations

> What are you expecting (your users?) to do in Ruby
> that could require multiple interpreters?  Are you
> writing a web app using a not-ruby-aware framework,
> so you need to have separate "contexts" for different
> users of the app?

(pardon my Windows-centric terminology)

Imagine a command shell that can run individual Ruby commands
maintaining a persistent Ruby environment.  For example (given a syntax
of "%@ruby[" <ruby_statement> "]"):

    %@ruby[b = 42]
    %@ruby[p b]

would print 42 on the console, and then

    %@ruby[`set q=#{b**3}`]

would set the environment variable q to 74088 - *back in the calling
environment*.

Meanwhile, the shell would be able to directly run Ruby scripts, but
might well want to run them in a fresh, clean environment, as is done
by invoking ruby.exe on a script. And after the script has finished,
the b would still be 42 in the command shell's persistent Ruby
environment.

-- Timothy