From: Paul Brannan <pbrannan@...>
Date: 2007-02-23T05:15:04+09:00
Subject: Re: Stateful I/O interface

On Tue, Feb 20, 2007 at 11:24:52AM +0900, Tony Arcieri wrote:
> 
>    Has anyone ever suggested adding a stateful I/O multiplexing
>    interface which could be used for things like network servers where
>    something like Kernel#select is less than ideal?

For now, I think this is best done as a library.  The reactor pattern is
a well-accepted pattern for implementing such a mechanism.  There is a
reactor library for ruby, but it needs to have additional I/O
multiplexing mechanisms implemented.

I would like to see a way to pick a different I/O multiplexing mechanism
at runtime.  IMO the problem isn't what Kernel#select does (you don't
have to use it), but what rb_thread_select does (you aren't given a
choice here).  Integrating ruby's event loop with a library that uses a
different mechanism requires that the library be able to use ruby's I/O
multiplexing system, but may not be possible or desirable, especially if
ruby is embedded in an application.

It's been suggested that Ruby use WaitForMultipleObjects on win32 (see
[ruby-talk:47186]).  There are pros and cons to doing this, but I'm not
qualified to enumerate them.

Paul