So, I'm thinking about how to integrate the Abdera server code into a web app I've been playing with, and so far I've been hit by the following thoughts.
It's unfortunate that we don't have a way (or, if we do have a way we don't have an example) of how to use the server code without routing stuff through AbderaServlet. Lots of web app frameworks have you route requests through a single top level servlet. It would be nice if we could integrate with them wihtout people having to add this whole separate servlet bolted on the side of their app to handle the atom publishing protocol stuff. It seems weird that we essentially require parsing of the request URI twice, once in the Resolver to figure out what type of target a given request is for, and then later on in the Provider methods when you have to figure out what any given request is actually doing. It'd be nice if those two could be consolidated somehow, maybe allow a regex based resolver to make a note of what one hit and what captures occurred, so you can use that info later on? Also, for people who have used this stuff to build existing APP applications, how are you doing it? Are you using AbderaServlet to handle all APP type traffic? Using a custom servlet to do the same job as part of your application? -garrett
