Day: January 5, 2003

Recovered from the Wayback Machine.

I feel it it my bones: 2003 will be the Year of Linking Dangerously. It will be the year that we reject page ranks and popularity-based ‘s/he with the most links bubbles to the top of the heap’ skimmers.

It’s in the air. It’s viral. It’s contagious. Hold your breath or you’ll catch it.

Case in point: Joshua Allan writes elegantly and compellingly in defense of the Semantic Web in response to somewhat disparaging comments made by Mark Pilgrim and Dare Obasanjo. In his post, Joshua makes the point that the Semantic Web is about people, and about making people’s voices audible and indelible.

On audibility, Joshua writes:

 

Today I get most of my lies from whichever barbarians have clawed their way to the top of the local and national media outlets. But sometimes when I see an advertisement for an interesting new product, I want to be able to pick up my remote control and click on “connect me to five people who hate the product and ask them why”

 

On indelibility, Joshua continues:

 

The story of Babel is a metaphor for what later happened at Alexandria; a reminder that we all suffer when we lose our ability to pass lessons to future generations…

Again, the Internet has had a profound impact on our ability to preserve our collective memory, but we are still very fragile.

 

Joshua sees the Semantic Web as a way of storing our collective knowledge in a manner that is easily accessible, based on methods more effective than today’s crude raw scans of hypermarked text. He also sees that the so-called problem child of the Semantic Web, RDF/XML, is more of a “red-herring” to the discussion rather than an actual impediment. RDF/XML is really nothing more than the selected method used to record knowledge, chosen from among all possible methods. It’s a means to an end, not the end itself. He answers the argument about RDF/XML being too complicated by saying:

 

The primary serialization for RDF is XML. This really starts to hurt your brain when you realize that RDF and XML are almost the same thing. Too much meta and your mind can’t bootstrap.

 

Joshua also states that he agrees with the consensus that, overall, RDF is to complex for most renmin (“people”), something I, of course, <shamelessplug>hope to disprove with my book</shamelessplug>.

Joshua Allan’s post is extremely well written and well argued, and you should take time to read it in its entirety — whether you’re a techie or not. I not only agree completely with Joshua, I’ll take it farther by saying that the Semantic Web will never be built with today’s system of dumb links — links with no inherent meaning attached to them other than their numeric value. All links do is push a resource’s rank up, piling link after link like pieces of wood for a bon fire; except instead of using dead trees we’re using dead links.

Torch the piles! Reject the concept of “all we need is links, sweet links”! Burn down the house that Google built before we become as dependent on Google as we almost became on Microsoft Windows.

So where do you start?

First, enable Trackback. It’s the first semi-intelligent threading implementation that’s actually starting to get fairly widespread use.

The Trotts from Movable Type have provided standalone trackback servers for those who don’t use Movable Type. The technology has already been integrated into Bloxson. If you’re using a weblogging tool that hosts the pages, ask the tool vendor to incorporate trackback functionality if it hasn’t already.

Once your tool supports Trackback, use it. It’s there to allow people to visibly show their connection to your writing. It let’s your readers know that others have something to say about your post, good or bad. It continues the conversation. It breaks the hell out of this idea that weblogs are just some form of electronic journal, written in isolation that just happens to be published for posterity’s sake.

More than that, though, Trackbacks provide deliberation and some intelligence with the link. Not a whole lot other than the words of the associated post, but it’s a start. There’s a tiny bit of RDF/XML associated with the Trackback link — a placeholder for future information, future bits of knowledge. A base on which to build.

At a minimum, Trackback doesn’t break into your space like comments does. You don’t even have to post the excerpts associated with the trackback — just the link. By allowing trackbacks, you’re providing a way for others to participate in the thread you started. When you don’t implement trackback, you’re breaking the thread.

Of course, the use of Trackback, as well as enabling comments is a personal choice. If you wish to separate yourself from others and exist in splendid isolation, far be it from me to get in your face about it.

Next week, part two of Year of Linking Dangerously — getting rid of your blogrolls.

(Thanks to Kevin Marks for sending me the link to Joshua Allan’s wonderful post.)

I’m firing on all (one) cylinder today.

Sam Ruby references a citation at Simon Willison, who quotes Tantek:

 

“…we now have Trackback and Pingback to help automate generating comment hyperlinks to blog-on-blog commentary. While I certainly applaud these efforts at automating the plumbing, I must ask – why is there any distinction in the presentation? I ask because many blogs present separate and different interfaces for their comments, trackbacks, and/or pingbacks.

 

Good points. After all, these technologies are nothing more than threads to a communication.

For Movable Type, it’s fairly simple to make a modification to your individual archive page to list trackback entries along with your comments. I’ve made this modification to my individual archive pages and thought I would pass on the how-tos of my mod.

Warning: To implement trackback within the archive page following my prefered approach, I did need to make a minor modification to one of the Movable Type’s Perl modules, Trackback.pm. It’s a minor change: it forces a re-build of the archive page when a trackback occurs so that the new trackback entry displays in a manner similar to how new comments are added, automatically, to the page. You can download the modified file here and replace the in your MT directory (put it into /lib/MT/App/). However, you do so at your own risk. You can find the edits I made because I surrounded the edit with comments containing my name, ‘Shelley’.

Repeat: You do so at your own risk. This modification is not vetted by Movable Type’s creators, Ben and Mena Trott.

For those taking the leap of faith, to add the trackback entries to your individual archives, add the following to your individual archive template:

 

<MTEntryIfAllowPings>
<MTPings>
<div class=”comments-body”>
<a name=”<$MTPingID$>”></a>
<a target=”new” href=”<$MTPingURL$>”><$MTPingTitle$></a><br /><br />

Excerpt: <$MTPingExcerpt$>
Weblog: <$MTPingBlogName$><br />
Tracked: <$MTPingDate$><br />
</MTPings>
</MTEntryIfAllowPings>

 

Note that the re-build of the page does slow the trackback ping, and if the remote site is having performance problems, the rebuild may not occur. However, the exact same process is used with comments, so whatever performance problems we’ll have with comments, we’ll have with trackbacks. Additionally, malicious people (known as spammers) could exploit the ping to add trackback entries pointing to junk — but they can do this anyway with the existing system. Web services are vulnerable that way.

Other trackback embedding approaches are discussed at the Movable Type forum on a thread related to this issue. I didn’t care for the approaches mentioned, excellent as they are, primarily because I would rather put the processing burden on the instance when the trackback occurs, rather than each time the individual page is accessed (by accessing MySql or forcing the page to be PHP or using SSI). I’m putting the burden on the ‘write’ because trackbacks follow the ‘write once, read many times’ pattern.

Still, don’t you like it when you’re given ten different ways to do something?

A discussion broke out at Sam Ruby’s talks about weblog popularity. In particular, Jim Winstead wrote:

 

The navel-gazing among some webloggers is pretty monumental, and systems like technorati, blogdex, and popdex reflect that.

 

I agree with Jim. As I mentioned in my own post, Third Generation of Weblogging, following an analogy of human growth to weblogging generations, this last blog generation is similar to our teen years with our fixation on popularity. S/he with the most links, wins.

The hypertext link is still our most direct tool for connectivity, but we need to make links ‘mean’ something other than bland acceptance and hence popularity. We need to start tying communication together and the link is the tool of choice, but we need to start using links effectively.

Of course, this leads me back to my own personal bugaboo: blogrolls. My blogroll replacement functionality is complete except for one thing — the blogroll links themselves. If I pull these links from the main page, delink them all, this action results in one less ‘class president vote’ so to speak for all my blogroll members. This impacts on each weblog’s ranking in technorati and blo.gs most watched and the blogging ecosystem, not to mention Google rank. My decision to not go with a blogroll impacts on other people.

Is it fair of me to implement a new way of doing things in my weblog that influences the participation of each of my blogroll listees? The old “my weblog, my choice” of course takes precedence, except that now we’re seeing twistie little threads of community, sticky strands of webbing that tie us together a tad more than is implied by ‘journal’.

After all, if we didn’t care about interaction within the weblogging community, we wouldn’t implement comments, or trackback, or link to each other, or correspond with each other on the phone or within emails. We’d just write and be indifferent to that which doesn’t impact directly on us — the pure journalist who never sees his or her audience.

(I just realized something — going back to writing in a void is almost repugnant to me now. What have you all done to me?)

Yet at the same time, there is a quality of “recommendation” to blogrolls that I don’t care for, above and beyond the political use of blogroll links; an assumption that blogrolls are made up of people whose weblogs, and indirectly writing and opinions, we recommend. You may not mean this by your blogroll, but that’s how people can, and do, perceive them. Do you want to recommend the weblogs you read regularly? Do you read weblogs regularly that you don’t recommend so therefore you don’t list in your blogroll?

At the Blogstreet site, they define Blog Neighborhoods as:

 

It is a set of Related blogs based on the BlogRoll of the analyzed blog.
For Authors, their Neighborhood is a pool of blogs to track.
For Readers, it helps them to find more blogs similar to a blog they have liked (emphasis added).

 

But, but, but — the weblogs in my blogroll aren’t similar to mine. In fact, many are quite different. Blogrolls are just like HTML elements: there’s an interpretation attached to them that can vary, considerably, from the original intent.

blogroll == em
blogroll == img
blogroll == a
blogroll == blink!

Anyway, the approach I’m considering at this point for my blogroll replacement is a PHP page that lists links to all of the weblogs that are among my significant quotes, with a caveat that no similarity or agreement with the writing is implied by the link — follow at your own risk. If the buzz sheets are smart enough, they’ll follow a link from my main page to this page, and people will get their votes. If they care. This is in addition to the PHP code that will randomly pick a weblog quote for the weblog main page.

What think?