Re: the meaning of stateless

Mike Schinkel wrote:
> 
> Well, it's unproductive if you already understand. So IOW you are
> saying it' a stupid question? Haven't we been socialized to tell
> learners there are no stupid questions so as not to retard
> learning? :)  
> 

Part of learning is understanding which are the right questions to
ask.  ;-)  After reading the entire thread, my best advice to you is to
get the book, "Software Architecture:  Foundations, Theory, and
Practice" by Taylor/Medvidovic/Dashofy.

This is the only REST book I currently recommend, because it's about
architectural styles, not REST per se.  The common thread throughout
the book is a "lunar lander" application -- the discussions and
exercises illustrate the different implementation details of a well-
known app as the architectural style is varied.

"I rehabilitate Web APIs, but I train developers..." the REST whisperer
method isn't about examples of specific solutions to specific problems,
so much as learning how to approach a problem in terms of applying an
architectural style as a means to an end.  The "end" would be the goals
of the REST style being realized in the resulting system.

I think I would have grasped application vs. resource state vs.
stateless messaging in far fewer years, if I'd known to start with the
concept of information hiding.  And REST itself, had that book been
published a decade earlier, since REST was my introduction to the
concept of software architecture and the books of the time didn't take
Roy's work into account.

So I think many of the questions you've been asking here, would be
clarified not by REST examples, but by the cross-style lunar lander
example in the textbook.  It'll help you know where one style starts
and another leaves off in typical REST examples, regardless of source.

-Eric