Recently, one of the bloggers I read, Caffo over at Pattern Recognition, brought up one interesting issue in his blog, complaining about the monotony that pervades the blogosphere both in terms of layout and site structure. This is a really trick subject, with many ramifications, involving even the culture of the blogosphere itself.
In a comment on his blog, I said that the biggest problem was inertia: most of us are simply used to what already exists, which is also what almost everybody uses, and there’s a certain reluctance to change what, so it seems, works well. That comes hardly as a surprise since innovation, in all areas of human knowledge, usually proceeds from a considerably smaller subset of the group of people working on those areas. In the Web, a imitative medium where dissemination of customs and patterns of use is so easy, that becomes especially true, even if the opportunities to innovate are considerably larger on it as well.
To ascertain the situation, you just need to look at sites on the top of the curve of the power law distribution that describes the blogosphere. Even those sites, counting with a superior potential in terms of what is doable, tend to gravitate around the same tried-and-true structural patterns. I say that not as an offense since my own site, which doesn’t even figure on the power law’s curve, follows the same old and boring model of a blog, with content of the left and a sidebar on the right with dozens of links that conflict with each other losing most of their meaning and usefulness. The fact remains, however, that little work is being done on the subject. Part of the problems comes, as I mentioned before, from the blogosphere’s own culture. A simple example is the blogroll, which is almost mandatory on the front page of a blog. As an information item, there’s not specific need that requires a blogroll to appear on the main page of blog. If the intention is to show what sites the person behind the blog is interested in, there are many other points in the organizational structure of the blog in which the blogroll would fit better. Blogrolls appear in first page of blogs because, in most of them, they fulfill an implicit need to call the attention of the linked sites, bringing more visibility to the linking site. Then again, almost all of us are or have been guilty of this. Similar reasoning applies to many other common elements in blogs, including, for instance, comments.
All that makes me think about two related, yet distinct, points about the subject at hand. The first point concerns to the structure of sites and the existence of alternative forms to organize them in a way that allows a more efficient use of the resources of the Web, the available screen state in a page, and the content of the site itself with regards to its users. The second point, indicated by the title of this post, concerns to the use of a blog — and as a matter of fact, any site — as a form of information space. In this sense, a blog transcends its role as a simple aggregator of posts whose relationship among themselves is almost entirely fortuitous.
With regards to the first point, there are certainly better ways to make content available in a blog. Driven by his own dissatisfaction, Caffo modified the front page of his blog, putting a bigger emphasis on the content, prioritizing what was newer and more relevant, while also highlighting some other content of additional interest that would otherwise quickly disappear from the main page of an ordinary blog, usually organized in reverse chronological order of posts. Although that reorganization is just the first step for a more concerted change, it’s a excellent beginning.
Considering the existence of numerous patterns of information design (even for purely personal sites), we can only wonder why they are not used. As I also commented on Caffo’s blog, I attribute that to at least three big problems.
The first is knowledge, or the lack thereof. Changing things in the layout of a blog is a reasonably simple activity. And I’m not talking about creating a compelling and pleasant visual presentation. I’m talking about just changing simple graphical element s of the page, something that can be easily done in the site’s style sheet. Any person with a minimal knowledge of Web technologies can do that. Experimenting with the information structure of a site is another matter altogether. It requires much more knowledge, and it’s a much more involved activity. A simple example of this problem in the duplication of content in many blogs, with archives the carry the same posts multiple times in their entirety, without aggregating any added information to them.
The second problem is the difficulty of gathering metadata. As it has been discussed exhaustively before in the Web itself, gathering metadata is a expensive chore, close to impossible in many situations. Gathering reliable metadata is probably not even feasible in most contexts. If people interested in the subject have trouble doing it, it’s too much to expect that simple users will do it too. Unfortunately, metadata is required to enable the kind of content transformations that will increase the usefulness of a site. For instance, metadata is necessary to create reliable relationships between diverse content.
The third problem is the lack of tools — and integration among existing tools — to maintain the information present in a site. Even ignoring the metadata problem, the current applications used to power most sites lack the required tools to manipulate a site’s structure. A simple example is the trouble of maintaining the valid markup of a given piece of content — markup that would enable the use of that content in different contexts.
All those problems contribute to make the task of using alternative structures a difficult predication, creating a steep ladder of acquisition that prevents most users from adequately exploring the potential of their sites. And it’s equally hard to talk about solutions when the problems are so diverse and generic. It’s a situation in which small modifications are better than nothing, because they at least they form the basis for deeper changes.
Thinking about the second point now, which is the use of blogs as information spaces, it’s easy to see that it depends, almost on its entirety, on the first point. Transforming a blog — or, in fact, any given site — in a space that aggregates diversified information allowing it to be accessed in a more efficient manner requires much more than simply entering information: it requires extra classification and categorization work to enable it to reach a better state of usefulness.
Considering my own site, I can think in a lot of things I’d like to do to aggregate more content to individual posts. For instance, I’d like to link terms requiring explanation to the corresponding entry on Wikipedia. I would also like to gather information related to each entry, together with similar entries, and present that in a more useful way. And I would also like to allow entries to be freely annotated. I can think of others things that would add value to a given entry.
In the context of the site itself, although that is even harder to do, there are more things to aggregate. The recent rise of blogmarks, links worth remembering but not commenting on, is a clear indication of those possibilities. There’s a real need to preserve the memory of a browsing session. A blog as an information space would enable that since it would be a real web of archived information whose pieces become more and more related as time passes, mapping the acquired knowledge into a coherent form.
Take the elements above and apply them on a interface that allows each piece of information to be annotated, Wiki-style, and that helps to create references, and you get a space much more useful than a simple site, with keeps a qualitative state and accomplishes better the vision a Web in which hyper-documents are the norm, not the exception. Of course, an interface like that would require some form of transclusion. The key issue here is to create links between newer content and older content, preventing the latter from becoming obsolete just because of obscureness.
Of course, such sites would not remove the need and function of simpler sites—or more complex ones, for the matter. There’s always a need for different kinds of confluence between content and structure.
The next stage, which is not where I want to go, is the Semantic Web, which would (supposedly) work in a level orders of magnitude higher than such a simple information space, even in more restricted conditions. The idea, repeating what was aforementioned, is to create a more organized site, making better use of the information that accumulates on it with time. It’s not to create a intelligent and self-managing process, able to modify itself to conform to the needs of its users.
Going back to the beginning, we come again to the problem with inertia. Perhaps, in a near future, tools will be developed to handle some of the problems described here, giving raise, then, to new problems. For now, we can only forging ahead in small steps, fighting the entropy that threatens to engulf all the content we create.