Re: Stateful Web Services...

On Tue, 2004-10-26 at 12:22, Cristóvão Bento wrote:
> Hi to all,
>  
> As far as i understood, web services are stateless by definition.

This may be true of some implementations of Web services, but there is
nothing in the Web services architecture[1] that prevents Web services
from being stateful.

>  Although i already read some works about JAX-RPC and the possibility
> of creating a stateful web service. To do this we need to extend the
> javax.xml.rpc.server.ServiceLifecyle interface. 
>  
> Although i would like to know what are the developments in stateful
> Web Services. Is it possible already to have a full server running and
> responding to requests like for example a Windows DCOM Service? or a
> CORBA Server. These kind of server have their own cycle of life. For
> Web Services the cycle of life is controled by the Web Server, they
> are dynamically allocated when a request arrives and deallocated when
> that request/session ends. Also these kind of servers gives us the
> hability to maintain some usefull information in memory which can make
> their performance increse in several situations.

Again, you seem to be talking about a particular kind of implementation
that has this limitation.  

>  
> When this will be possible with Web Services? to have mass processing
> server running while some application server is running...

Reference
1. http://www.w3.org/TR/ws-arch/

-- 

David Booth
W3C Fellow / Hewlett-Packard

Received on Tuesday, 26 October 2004 17:15:57 UTC