As I May Think...

Beginning RSS and Atom Programming

My copy of Beginning RSS and Atom Programming by Danny Ayers and Andrew Watt arrived in the mail today and I must say that I'm terribly impressed. There have been a number of books written about RSS, however, this 700+ page tome is unquestionably the biggest and most complete. It is also the most up-to-date. It provides in-depth coverage of everything from MCF and CDF (ancestors of today's RSS and Atom) to the FeedMesh.

Weblogs.com now supports Extended Ping!

I was very pleased today to see Dave Winer's post saying that Weblogs.com has been updated to support the Extended Ping format which enhances the old Basic Ping format by providing a pointer to an rssUrl. While most sites that ping weblogs.com are still only sending basic pings, my hope is that the transition to Extended Ping will be rapid. Ideally, all of the entries in weblogs.com's changes files will soon have rssUrl attributes!

Basic Ping is becoming historical. Extended Ping grows in use.

Matt Mullenweg, creator of WordPress, wrote on the FeedMesh mailing list:

WordPress, a significant source of pings to Ping-O-Matic, will now be sending out extended pings. [link]

This is excellent news! While once it might have seemed sensible to ignore RSS and Atom files when announcing blog updates (as the old "basic ping" ignores them), that simply isn't reasonable anymore. Blogging has moved beyond the idea of "personal webpages" and begun to pioneer new forms of content syndication that rely on machine processable Web 2.0 data structures like RSS and Atom files.

My sincere hope is that WordPress will soon completely drop or deprecate support for basic pings and that all other blogging software will do the same. Only extended pings, which include pointers to a site's RSS or Atom file, make sense today. Some techies might argue that tools should continue to support both forms of ping as a way of ensuring that users have "choice" -- but that is just silly. Forcing the non-technical users, who are the majority of bloggers, to make the choice between the old basic ping and the newer, more powerful extended ping is simply adding useless complexity to the problem of managing a blog. We need to make blogging simpler, not more complex.

"Mention them and they will come!"

It's seems like years now that I've been saying: "Someone once said 'Mention them and they will come!'" as a way of explaining how to attract someone's attention in the Blogosphere. On his blog today, Stephen Spencer talked about this technique and how it relates to "a new set of rules" for the PR industry:

"If, for example, you want to catch [Robert] Scoble’s eye, then you would say something interesting that somehow relates to Scoble and work in a mention of his name. Scoble, like many other bloggers, follows what’s being said about him in the blogosphere by subscribing to a PubSub search results feed for the word “scoble.”

I can't remember who coined the phrase "Mention them and they will come!" and when was it first used. Do you know? If so, please drop me a line or a trackback.

bob wyman

Zipf still rules... (InDegree, OutDegree in PubSub Data)

Discussion of the Long Tail seems to have supplanted what was once a vibrant discussion of power laws, Zipf distributions, etc. and their presence in the blogosphere. Perhaps most folk have simply accepted that the Zipf distributions are there and have moved on to discussing implications... But, for those that haven't yet been convinced, I offer below yet more evidence of power law relationships in blogging. The data below is based on links in blog posts discovered by PubSub on 14-Mar-2005. This data conclusively demonstrates Zipf distributions in the degree of in-links and out-links in the blogosphere.

FeedMesh works! (blo.gs data)

As many have noticed, since December just about every major blog monitoring or search service has shown a significant jump in the number of known blogs. While there are many explanations for this, one reason is that more and more services are benefiting, either directly or indirectly, from the FeedMesh.

Recently, Jim Winstead of blo.gs posted statistics showing the impact of the FeedMesh on the number of blogs known to blo.gs. His numbers show that during its first month in the FeedMesh, blo.gs discovered more than 1 million new blogs than it did during the previous month and more than during any month in blo.gs's history. Even after having "caught up," blo.gs is still discovering over 100,000 more new blogs monthly than it did without the FeedMesh. Other services that have been connecting to the FeedMesh are seeing similar dramatic increases in their ability to discover new and updated blogs. Clearly, FeedMesh is working...

What if PubSub had come first? (i.e. JEP-0060)

Peter St. Andre asks: "What if every user's Jabber ID were a PubSub node?" (Note: Jabber = XMPP instant messaging).

Well, as I've often said, if "Jabber PubSub" had been JEP-0001 instead of being defined in the much later JEP-0060, many of the other Jabber protocols would never have been created or would be very, very different in the current form. The reality is that many protocols, such as presence, chat, etc. are simply special cases of the more general case defined by Publish/Subscribe. Many other protocols can be easily viewed as simply extensions to a core publish/subscribe system.