From: Stephen D. Williams (firstname.lastname@example.org)
Date: Thu Mar 16 2000 - 09:30:49 PST
Kragen Sitaker wrote:
> JoeBar (should I stop calling him that now that he's left Microsoft?) writes:
> > I've NEVER seen an environment where using threads
> > willy-nilly was a good idea.
> Well, from reading about VolanoMark, I seem to remember that Java
> doesn't have non-blocking I/O operations, which means you need one
> or two threads per network connection. If that's true, then in Java,
> in some cases, using threads willy-nilly isn't just a good idea ---
> it's the law!
This was a really stupid decision in the design of Java and its I/O semantics.
There are also ways to get around it, as I'm implementing.
The other performance evils for both C++ and Java are marshalling/unmarshalling
and object creation/copying and the associated memory allocation and collection.
> See also the Tera MTA. There's an explanation at
> http://www.tera.com/tech/teramta.html; unfortunately, it sucks
> royally. More detailed papers might be available at
> <email@example.com> Kragen Sitaker <http://www.pobox.com/~kragen/>
> The Internet stock bubble didn't burst on 1999-11-08. Hurrah!
> The power didn't go out on 2000-01-01 either. :)
-- Insta.com - Revolutionary E-Business Communication firstname.lastname@example.org Stephen D. Williams Senior Consultant/Architect http://sdw.st 43392 Wayside Cir,Ashburn,VA 20147-4622 703-724-0118W 703-995-0407Fax Jan2000
This archive was generated by hypermail 2b29 : Thu Mar 16 2000 - 09:35:12 PST