Re: Stateful Web Services...

On Nov 4, 2004, at 10:51 AM, Mullins, Chalon wrote:
> But -- we a lot of this happens because there is no standard way of
> maintaining a reference to the state between client and server.

If you are waiting for The One True Way, then I recommend that you "get 
used to disappointment" (cf. The Princess Bride). If you look at the 
collection of arguments that I gave for the distributed file access 
case, it should be clear that different use cases and workloads are 
going to lead to different solutions. In fact (irony alert) the latest 
versions of NFS introduce de facto stateful sessions. And that's OK.

Of course we want to identify common patterns and refactor to make them 
reusable. (I'd prefer to do this on the basis of actual, real-world 
experience rather than trying to dictate standards ahead of time, but I 
guess that's a lost cause.) But the bottom line is that there will be 
multiple models, multiple standards. Silver bullets and Procrustean 
beds lead to bad engineering.....

Received on Thursday, 4 November 2004 16:20:18 UTC