J2EE equivalent in Python

Bill Tate tatebll at aol.com
Mon Oct 22 21:28:17 EDT 2001


Robin Smith <smithrc at zdbaora.nat.bt.com> wrote in message news:<x2het5jflz.fsf at zdbaora.nat.bt.com>...
> I like J2EE but I am also an open source fan. I don't like what I read
> about Sun and restricting Java -
> http://www.onjava.com/pub/a/onjava/2001/10/10/osjava.html .
> 
> Is there anything similar to J2EE for python?
> 
> Robin

Robin,
At best its J2EE light.  However, I would definitely keep an eye on
Skunkweb and STML.  Reasons: 1. Skunkweb forks processes rather than
using threads (app servers tend to be I/O bound not cpu bound as many
have noted previously) - see Sam Rushing's site on this topic - he's
got some great references there.  2. Since forked processes means the
Python interpreter can take advantage of multiple processors where a
threaded model means being able to only take advantage of a single
processor no matter what. 3. The STML implementation is by far sight a
superior alternative to DTML and it is extensible.  I know the Zope
folks are working on alternatives but it takes some real effort to
figure out what's going on.  STML is also logical and much more
straight foward to understand.  4 With Skunkweb you have access to
Python - Zope presently doesn't get high marks on that score - again
they are addressing that I'm to understand - and updating python code
without shutting down the server is is facilitated by its use of PAR
files (akin to JAR files).  5.  I like Webware - a lot - but it is
multithreaded which gives me pause to think scalability could be issue
and I'm not to keen on PSP.  Since one of the reasons J2EE is pushed
is the ability to handle huge transaction volumes, scalability is
clearly an unavoidable core issue wrt alternatives. That's before you
even get around to discussing entity/session/messaging beans.



More information about the Python-list mailing list