RE: JAX-RPC 2.0 enhancements and changes - feedback requested

From: John Furton <>
Date: Thu, 15 Jul 2004 09:34:25 -0500

I always found strange that the number of class files generated by the
wsdl compiler is pretty high. I basically get 4 classes for each method
exposed through wsdl. After experimenting with Axis and .NET, I'd like
to see a reduction of the number of classes. In Axis, I have only two
classes but what's really nice is .Net where the entire interface is (or
even multiple interfaces are) compiled into a single .cs file.

Another topic where I'd like to see some improvement is the
document/literal implementation of xsd:anyType. This is basically the
type used do describe associative containers like hash, vector, etc, but
the current implementation appears to generate SOAPElement and not


John Furton

-----Original Message-----
From: Doug Kohlert [mailto:Doug.Kohlert_at_Sun.COM]
Sent: Wednesday, July 14, 2004 5:07 PM
Subject: JAX-RPC 2.0 enhancements and changes - feedback requested

With the implementation of JAX-RPC 2.0 underway, now is the perfect time

for you to let the JAX-RPC SI team
know what enhancements you would like added to the 2.0 implementation.
We of course cannot promise anything
but we would like to incorporate your suggestions if possible given
time-constraints, etc.

For current users of wscompile/wsdeploy, we would like your feedback on
some potential changes that the SI
team might make in 2.0.

When JAX-RPC 1.0 was specified, it was believed that web services would
focus on rpc/encoded. However,
since that time, the industry has shifted and it appears that
document/literal is the preferred mode. Currently,
both wscompile and wsdeploy default to rpc/encoded. The question is:

Should wscompile and wsdeploy leave the default mode as rpc/encoded for
backwards compatibility reasons, or
should the default be changed to document/literal? If the default is
changed, we need to provide a way to get
the old behavior so I pose a second question. Should we add new
commands such as wscompile1 and wsdeploy1 that
have the old behavior, or should we simple add a -f:rpcencoded flag to
the existing commands? If anyone has other suggestions
we would entertain them as well.

Thank you

Doug Kohlert
Sun Microsystems, Inc.   
To unsubscribe, e-mail:
For additional commands, e-mail:
To unsubscribe, e-mail:
For additional commands, e-mail: