Back-links are the backbone of the blogosphere, allowing bidirectional linking that creates a web of conversation. Several standards exist, so how do they stack up?

TrackBack

TrackBack is most commonly used by manually copying the TrackBack URL of an existing page and pasting it on a new page. Pages that support TrackBack include a TrackBack URL that can handle pings (e.g. http://somewhere.blog/trackback/demo-post). Supporting blog software includes a field in which to enter this URL, and will send an HTTP POST to the TrackBack URL when used. The post includes the source url as well as the following option parameters: title, excerpt, blog_name. The server responds with simple XML indicating success or failure. The spec also includes provisions for automatic discovery via RDF embedded in the target document. Typical exchange occurs as follows:

1. Alice writes a blog entry that interests Bob.
2. Bob writes a related blog entry, and manually copies and pastes the TrackBack URL in his blog software.
3. Without interaction, Alice’s blog software receives the TrackBack, and adds a link and excerpt from Bob’s blog.

TrackBacks remain the most widely supported linking mechanism. Compared to Pingback, “the other open standard”, the TrackBack standard includes article titles and summaries, which are important for judging whether a link is worth reading. On the negative side, usage is effectively manual because the autodiscovery mechanism is not commonly supported. The standard doesn’t include provisions for well-known error codes, but this complaint is somewhat academic.

Pingback

Pingback also uses HTTP POST’s, but requests are submitted via XML-RPC. Autodiscovery is done via the X-Pingback HTTP header or an HTML link element (<link rel=”pingback” xhref=”" mce_href=”" >). The server responds with an arbitrary string indicating success, or any of several defined fault codes – nonexistent URI’s, already registered, access denied, etc. In short:

1. Alice writes a blog entry that interests Bob.
2. Bob writes a related blog entry that includes a link to Alice’s blog.
3. Without interaction, Bob’s blog software uses the autodiscovery mechanism to send a Pingback to every URL in his entry, including Alice’s, that supports Pingback
4. Without interaction, Alice’s blog receives the Pingback, and adds a link to Bob’s blog

Although less widely supported than TrackBack, Pingback has supported autodiscovery since it’s inception, so cut-and-paste is not necessary. The standards-based XML-RPC and error reporting mechanism are also arguably cleaner than TrackBack. Pingbacks are thus more transparent, and this transparency facilitates backlinking for incidental non-blog URL’s (e.g. news.com). Unfortunately, the spec doesn’t include provisions to relay a title or summary, which are important for judging relevance. Several implementations support these features, but it would be nice if the spec defined them.

Backlinks

Backlinks are the proprietary mechanism used by Google’s Blogger.com. Back-links and exceprts are generated dynamically using google’s blog search engine blogsearch.google.com (alternately search.blogger.com). Blogs that notify an update service, such as rpc.pingomatic.com, should soon be indexed by blogsearch.google.com. The scenario occurs like this:

1. Alice writes a blog entry that interests Bob.
2. Bob writes a related blog entry that includes a link to Alice’s blog.
3. Without interaction, Bob’s blog software updates several blog search engines using pingomatic.com.
4. blogsearch.google.com indexes Bob’s entry.
5. Users who visit Alice’s entry see Bob’s entry by virtue of the fact that it includes a link to Alice’s entry.

Backlinks are implicitly supported by any blog indexed by blogserch.google.com. This is convenient because any indexed blog that refers to the Blogger.com page is automatically in the Backlink list. On the negative side, blogger.com generates the back-link list dynamically, and Alice receives no notification nor history.

From telegram to telnet, from horse messenger to instant messenger — our communications arsenal continues to evolve. Web logs have existed for almost 10 years. So why the hype du jour surrounding the now-called Blogosphere?

Buzz

Simply put, the blogosphere is an interlinked network of web logs. Let me begin by stating that I, too, am resistant to seemingly meaningless buzzwords. To this day, I feel that the CD burning moniker is excessively sexy (violent?) for such a simple process. Another segment of the population resists buzz because it represents the infiltration of an in-crowd. And, just like when their favorite band “sold out”, the public at large now sings the praises of the blogosphere.

At first I found the term laughable — is this simply a case of the general public finally catching up? Back in 1993 I was reading ‘zines via gopher, right around the time Mosaic introduced graphical browsing. And before that I had some BBS experience, so I am no stranger to electronic expression. But this experience causes me not to dismiss blogs, but to appreciate the ensuing evolutionary changes to the web. And I am not alone – the blogosphere is growing exponentially (The Honorable David Sifry).

Backlinks

Older collaboration platforms such as Usenet, mailing lists, and forums are great tools, but they lend themselves to creating disjoint islands of data. For example, let’s assume that I find an interesting discussion on slashdot. I have no way to connect this discussion to the alt.politics post that inspired it. In turn, the alt.politics post might have used ideas from a post in a political forum. Thus the Slashdot discussion, alt.politics post, and forum post are stripped of their relationships.

Blog software addresses this continuity problem with TrackBack’s, Pingback’s, and backlinks. These are all two-way link mechanisms that allow a site to include links to all sites which refer back to it. This two-way linking allows ideas to flow more freely and the network gains value beyond the sum of it’s parts.

Bios and Blogrolls

Not only does greater connectedness add value to the data, but a new level of metadata becomes available by viewing a blogger’s related posts and fellow bloggers. Attribution in traditional collaboration platforms is largely limited to posts in any particular media, and anything the poster can summarize in a 3-line signature. The blogosphere again replaces this disconnected data with a stream of ideas, opinions, and other linked bloggers to better identify and engage people of similar interests.

“Beginners”

Pre-blog web outlets are decidedly geek-centric, often requiring some degree of computer expertise to navigate and add content. This too is addressed blogosphere and it’s associated technologies, which make it easy for people with little computer expertise to write about any topic and still reap the benefits of connecting ideas and people. Case in point – political blogs contributed heavily to making blogging mainstream.

Bottom Line

The blogosphere by any other name would connect as well. Facilities introduced by the blogosphere are fundamental to a well-connected read-write web, and they’re here to stay. Backlash inevitably follows any major trend, but the end result is undeniable. The Web, the little buzz phrase that could, should not favor the technically inclined. Ideas thrive in a free market, and the blogosphere offers that market. Disclaimer: throughought this writing, I have resisted the urge to discuss memetic evolution. Hard to defend one buzzword while using another .

When looking for a WordPress theme, it wasn’t long before I happened upon Lush, a theme that Jeriko ported from Typo. This beautiful theme features AJAX searches and an attractive color scheme, and quickly became my hands-down choice. Many thanks to the original Lush Typo theme from w3-labs.com and Jeriko’s faithful port. In the spirit of Creative Commons By-NC, I made a couple of changes…

I experienced duplicate HTML headers displayed by single.php when viewing individual posts, and I also got PHP “headers already sent” errors due to extraneous carriage returns at the end of functions.php. I also disabled the ladybugs, added orange RSS icons to the sidebar, and added a style-corrected calendar.

Here is my hack-job of a patch; sorry I haven’t better documented, or verified XHTML compliance, but I’m currently of the mentality post first, and ask questions later: lush-nathan.patch. Good night, and good luck x-)