Back in a Bit

The reverse-chronological arrangement of weblog posts has frustrated me for years. It’s a format that works for about the first week in the life of a weblog, and after that it becomes more and more of a headache.

One of the most basic principles of information architecture is that the good stuff should float to the top. Of course, what that means in real terms comes down to your definition of “good”. In the world of weblogs (and many, many other varieties of sites), this refers to the most recently added content, hence the reverse-chronological ordering of posts. But is the most recent thing you’ve written necessarily the good stuff?

For regular readers of a particular weblog, having the most recent content at the top makes a lot of sense—they’ve read what’s gone before and don’t want to have to wade through it to get to the latest addition. But for everyone else, this structure makes little sense. If you arrive at a weblog via Google, for example, you might just be lucky enough to find a snippet of information on the subject for which you were searching, but this information is almost entirely disassociated from any other information on the site. The author may have put out other information on the subject, and you could go back to Google and redefine your query in order to try and find them, but this is all very hit-and-miss. There’s little to help the user find related content, or even to suggest that related content exists.

As weblogs’ content management systems have evolved, solutions have been thrown up and deployed to try and get around this problem. A common solution is to assign each post to one or more broad categories. These categories are then listed so that the user is aware that related information exists and can traverse categories at will. However, with this approach comes yet more problems.

The first and most obvious problem to arise from the categorisation technique is what categories to have. Weblogs are usually a complex mix of the author’s interests, and some (if not all) of the content can be difficult to pidgeon-hole. Further to this, if the author writes a post on a subject they’ve not covered before and are unlikely to again, they face the dilemma of whether to create a new category for this post or simply to dump it in a catch-all category called something like “general” or “personal” or, at worst, “miscellaneous”. The catch-all category almost defeats the object of having categories in the first place.

The next problem is one of style. Considering that every post will require categorisation so that it will make sense within the archive, authors may (consciously or otherwise) limit the content or variety of their posts so that they fit more comfortably in a given category. They’ll find something and think “I’ll blog that”, but then hesitate because it doesn’t exactly fit into the grand scheme.

The next problem is a practical one. Many of the most popular weblogs have been running for quite a while, and have accumulated a considerable amount of content, yet when they started out these considerations didn’t occur. Go to any weblog author who’s been posting for more than two years, suggest they switch over to categorisation, and watch the colour drain from their expression: the prospect of going back through years of content post-by-post and assigning each and every one to a category is, at best, daunting.

One very interesting thing about all of this is that there have been advances in the ability to find and guide users to related information by people other than the author of any particular post. Comment systems, trackback etc. allow the scope of any particular post to be expanded by including the thoughts of others. What these systems are unable to do with any great competance is link this piece of information with other pieces of information by the same author—it’s difficult to guide users through a thought-process, specific chain of events or in any other way connected posts on the same weblog.

Let’s take this post as an example. At time of writing, I’m not suggesting solutions: I’m merely sketching out problems, mostly to aid my clarity of thought on the subject. In the future, it’s almost a given that I will publish more of my thoughts on the subject. But, if you arrive at this or any subsequent post via a Google search or a link on a weblog or any method other than religiously reading my every post as their posted, it’s difficult for me to guide you through my thought-process in a coherent way. Ideally, I should take you to the first post, and then guide you though successive posts so that you may understand not only the place but the direction.

What we’re looking at here is context. We have the technology to show where this collection of thoughts sits amongst the thoughts of others, but not so well within our own.