Lists Home |
Date Index |
> But consider that for the last several years, programmers have spent
> a massive amount of effort figuring out how to make a "web
> interface" to their software. Web application frameworks, J2EE, ASP,
> PHP, etc. etc. Part of this is because people have spent the last
> few years remapping their problem domains into the REST model.
I think the "web interface" here is important as a test of REST and
HTTP applicability. Not everyone needs, or wants a web interface...
though most people want to leverage ubiquitous networking easily.
At the end of the day, choice of architecture at this level is
dictated by larger application architectures... and people should
always remember that. Maybe HTTP/REST fits, maybe it doesn't.
The best part of Roy's thesis is the definition and methodology of
classifying and (subjectively) evaluating architectures. It provides a
nice basis for experimenting with, discussing, and evaluating new
network architectures and implementations thereof. It'd be interesting
to see someone design a protocol explicitly optimised for the REST
architecture... I'm picking it'd look something like BEEP.