@DaveWiner, SUP and pubsubhubub

Dave Winer has been suggesting alternatives to Google’s pubsubhubub and Friendfeed’s SUP.

Both of these methods are aimed at making the polling of RSS feeds more efficient.

Neither of these protocols use RSS. Which, come to think of it, is really strange. The RSS protocol has been around for a good number of years. Its well understood. and built into most Blogging software and CMS software. Not to mention RSS readers everywhere.

Dave Winer made a good case for prior design as a, well, design method. The idea being that things should be designed from the get go to be backwards compatible.

Those of you watching my Twitter feed will see that I fired off a few tweets to Dave saying that RSS can do the job in and of itself.

Consider. All we are really saying is that we should have a single feed or resource detailing what blogs  or feeds have changed in the last x number of minutes/hours. The idea being that you then go and poll only those feeds. Thus drastically cutting the number of requests you have to make you check ALL the blogs/feeds in the directory, or that are in your RSS reader.

So RSS basically details a number of entries, including title, link, author, date, etc. These entries typically refer to blog posts, and the like.

Capture

So assuming your software knows that directory y is a hub for x number of feeds, by checking y’s RSS feed, you can determine what blogs have updated because they are in Y’s RSS feed. Any blogs that are not in the feed, have not updated. There is a date element in the feed as well, which could be used as well.

This requires simple changes in the logic of the software, rather than the support of an entirely new protocol. I think everybody wins here.

One thought on “@DaveWiner, SUP and pubsubhubub

Comments are closed.