While I was hanging out at the hospital waiting for the new addition, I got to wondering if RSS has to be “pull” by definition. Could you “push” updates from a site to a client?
Say you have a notification client — a little app that sits on your desktop. You log into it, and it notifies the site that it’s waiting for notifications on port X. This inserts the IP and port number of the client in a table of connected clients on the Web server. When the site is updated, it runs through this table and pushes XML to the IPs and ports in the table.
This came up because the nurses would often sit back at the station and monitor patients from there. Now, I realized you wouldn’t do patient monitoring via blog, but say there was some event that happened on a site that was important enough to warrant more than the standard “poll-every-15-minutes” notifications. If you pushed the XML like I explain above, end users would get it in semi real-time.
If you used this in conjunction with NewsGator’s ability to format and display arbitrary data, you could really move any information around quite well.
In the end, it’s a cross between the “push” of email and the “pull” of RSS, I guess. Quite possibly pointless.