I really like this analysis:
…need to ponder and digest it some more.
I get the sense that we’re kind of swirling around this stuff. The HTML standard that evolved did a really good job of describing machine-to-human distributed interaction, and the Interwebs is teh result.
Now we’re struggling with how much of that we can map to the machine-to-machine space, flying in the face of all the machine-to-machine protocols we’ve sorted through in the last 50 (60?) years or so.
I agree with mca in another post that when the dust settles, and the hype dies down, REST will have been a pleasant idea and fun mental challenge that we chased for a while. The challenge is to capture the simplicity and benefits of ALL the distributed communication protocols & architectures that we’ve learned from in the last 50 (60?) years and come up with something super simple and powerful.