Reinhard Poetz pisze: > Ralph Goers wrote: >> Carsten Ziegeler wrote: >> I suppose REST is fine for some trivial applications on the web tier, > > I disagree with this statement. REST is an architectural style and I'm > sure, following those principles will help you to build applications > from trivial to very complex. > > I also think that the mentioned "without sessions your apps don't scale" > argument is bogus. If that was true, HTTP (which is the most prominent > implementation of REST) wouldn't scale. > >> but many modern web frameworks (i.e. JSF, Spring Webflow, Wicket) >> require that state be maintained on the server simply because they >> have a requirement that pages be accessed in specific orders. > > Using sessions for that purpose is just one way to implement it. > > If you implement a RESTful web tier, your clients (e.g. web browsers) > have to become more powerful. In times of mature Ajax frameworks, Flex > etc. this isn't as much of a problem like a few years ago. > > It took me a bit longer than Stefano, but in the meantime I also think > that web framworks as we know today have already passed their zenith. > Next generation web frameworks will make it simple to implement your > webapps following REST and I think that Cocoon with its XML pipelines > can shine again. > >> I believe even flowscript and Javaflow require this to preserve the >> continuations. REST between the web tier and the business tier is a >> different matter. The business tier should always be stateless and >> REST (as a concept) works very, very well for that. However, when you >> start talking REST vs SOA vs RPC the lines can get very blurry. I >> would contend that it is possible to actually implement something in >> the business tier that is actually all three. > > yes, definitly possible but IMHO really not disired. > > > Although I wouldn't recommend the usage of sessions, I agree with Ralph > and Carsten that we shouldn't forbid the usage of sessions. From our > past experiences with our own contracts (e.g. the Cocoon environment > abstraction, the FOM) we know that all those attempts haven't been very > successfull over the time.
Nothing more can I add to Reinhard's statements. I agree wholeheartedly with all that has been said above. Since I see valid use-cases for session sharing I'm not going to obstruct changes heading in that direction. -- Grzegorz Kossakowski Committer and PMC Member of Apache Cocoon http://reflectingonthevicissitudes.wordpress.com/
