These are the old pages from the weblog as they were published at Cornell. Visit for up-to-date entries.

February 12, 2004

RESTless Atoms

I know it is totally un-cool these days to have reservations about whether REST should be the one and only architectural style to build internet based applications and systems with, but I share Russell's frustration about the overloading of the PUT and DELETE verbs in the Atom API. I generally support the Atom format work being done, but I use the API as a showcase to students to demonstrate how a constraint interface can force you to make very awkward choices. Fortunately the we can expect an Atom server to also support alternative interfaces (read SOAP), so we may actually see developments in other styles.

Posted by Werner Vogels at February 12, 2004 12:27 PM


I would be very interested in your opinion where, exactly, the akward choices are - while I'm convinced that a REST-based design *will* lead to strange situations if applied to certain problems, I think the Atom API is actually a case where it works very nicely indeed. Could you elaborate? Thanks.

Posted by: Stefan Tilkov on February 15, 2004 02:08 PM

Right on! Agree totally.

Posted by: Randy Charles Morin on February 24, 2004 02:35 AM

Posted by: Randy Charles Morin on April 5, 2004 01:13 AM