Publishing flamebait [Fwd: Pragmatic Bookshelf releases "From Java To Ruby"]

Bryan Sant bryan.sant at
Wed Jun 28 16:03:57 MDT 2006

On 6/28/06, Hans Fugal <hans at> wrote:
> I'm also not very good at enterprise-level performance testing/thinking,
> but I'll venture that insofar as rails is not hindered by the ruby
> threads implementation it's by working around the limitations. What are
> the limitations, exactly? In my mind the primary problem is that if you
> block a thread in certain ways, you block the whole process. Aside from
> that, a virtual thread isn't all that different from a real thread, and
> if your program is a primary contender for resources you will probably
> not notice much detriment due to scheduling.
> Maybe Stuart or someone else has more insight into this than I do. If
> not, I can ask the URUG people or maybe the ruby-talk list, I suspect
> someone there would have a better answer.

Your application will get zero benefit from multiple CPUs/cores if you
don't use real threads.  On a single CPU (and single core) system,
this doesn't matter much (as you've described).


More information about the PLUG mailing list