TrackBack vs. Hyper-Cross-Referencing

A few days ago, Ed Brill brought up the question of the closed community of Domino bloggers. Methinks Ed worrieth too much, perhaps, but methinks he has a point nonetheless.

TrackBack is one of the supposedly cool blogging technologies that will connect us more closely to other bloggers, but I'm not willing to jump onto that bandwagon yet. The way I see the TrackBack specs, it is essentially an open invitation for spammers to autodiscover a way to disrupt all our blogs. Ed, I notice, does not have TrackBack in his blog, and even if this missing functionality is helping to keep the Domino blog community "closed", I'm glad! I'm not going to give any of my effort to contribute momentum to something so poorly thought out as that, and I hope that you readers won't either! We have this terrible problem with spam in our email, Usenet, and various other forums today because 20 years ago the designers of protocols didn't anticipate it. We know about the problem of commercial hijacking of "social software" tools now, so let's not make the same mistake! The designers of TrackBack have done so, and nothing good will happen if we go along with their mistake.

BTW: an aside.... I predict that the comment feature of our blogs will soon be hijacked by spammers, even without TrackBack. Finding RSS feeds is pretty easy, and can probably be automated at least partially. A tool that extracts link URLs from RSS feeds and matches them against a database that returns the pattern for forming POST URLs and form elements... that's a pretty trivial bit of programming (no! I'm not doing it! ;-)), and the effort required to collect the information for the database will be worth it to someone. Fortunately, we in the Domino blogging community will be able to defend ourselves against this, because unlike users of canned tools that generate lots of static HTML, we do everything dynamically. We'll have a pretty easy time changing our form names, field names, etc. as soon as the spam starts flowing. But if we implement TrackBack auto-discovery, we'll just be handing the information right back to spammers, so our countermeasures will be useless.

The alternative to TrackBack that I think we in the Domino blogging community should consider spearheading is an idea that I'm calling (for lack of a better name) hyper cross-refencing. This is a feature that I'm working on building into my aggregator application. In a nutshell, it does three things:

1. It retrieves the HTML for all articles referenced in RSS feeds, parsees it, pulls out all the hrefs and checks them against the hrefs of articles from other sites in its list of feeds, and builds records of any cross-references it finds.
2. Publishes a modified RSS feed containing items that give information about the cross-links.
3. Pulls in the modified RSS feed from other sites running their own versions of the aggregator, and imports records of the cross-references that the other sites know about.

This will expand the community for users of my aggregator (or of other aggregators that implement the protocol) by allowing me to learn about cross-references from people I trust, who in turn discover cross-references in the blogs they aggregate or learn about them from people they trust.

Maybe "hyper cross-referencing" is the wrong term for this. Perhaps it should be something like "XSRR" for "XML Syndication Reference Replication".

BTW: in addtion to Ed, credit for getting me thinking about this idea should also go to Ben Langhinrichs, for this reply to one of my earlier posts.