We’ve had a bit of a frenetic week here – it’s been very exciting to finally add the tag functionality and see how that ties together. There’s more to come, but I hope you’ve enjoyed playing with it as much as I’ve enjoyed building it.

I’m going to be in Granada, Spain for a week or so, hoping for sunshine and resisting the urge to tag the Alhambra. Therefore there will be no structural changes during that time.

It doesn’t get said enough, so thank you so much to the people who have sent emails offering to help develop, as well as all of you who have come in to have a look around since we opened the alpha in November. My next task is a load test to see where the system lags when lots of people are using it at once, and tighten the code accordingly – although there will be some completely new features as well.

So here’s a thought.

By the end of today, the Elgg software will have keywords attached to files and folders. That means you’ll be able to (for example) click on the keyword “social learning” and you’ll find links to everybody interested in it, everybody who likes or dislikes it, every available weblog post about it and every available file. Later we’ll add hooks to allow you to search through other Elgg installations elsewhere on the Internet, when they exist. Perhaps we’ll also add hooks to talk to Technorati, del.icio.us and similar.

Isn’t that essentially an organic community? Sure, you don’t have to be let in, and nobody’s moderating content, but it contains all the other community elements people talk about. And if that’s the case, do we really require the hard-coded “community” functionality, which would essentially do the same thing?

I’m not proposing we stop plans for the community element, but I’m beginning to wonder if it needs a harder think. Classes and so on are handled using online classroom software such as WebCT, which Elgg is not designed to replace, which leaves topic-based collections of people.

I’m curious. Am I right or wrong? What would you like to see from a separate community function, that wouldn’t be provided by the keyword functionality?

I’ve just found two of the coolest web apps I’ve seen in a while. There’s a little-known facility in the HTTP standard that allows a client to connect to a server and stay connected (as opposed to most web browser transactions, which are of the form connect – request – download – disconnect). This has allowed several companies, most notably Google, to create map sites that start out at the country level and allow you to zoom right down to individual houses, downloading information from the server as required.

Another site doing this is Search.ch, the Swiss search engine, which allows me to show you Elgg’s namesake in Switzerland.

This same stay-connected functionality is what powers Google Select; when I was out in Stanford talking to the Knowledge Media Laboratory, we discussed using this to let users in learning landscape systems pick keywords more effectively, auto-suggesting common words as you type. I’m certainly looking into it.

People seem to have gotten into the keywording habit. It’ll be very interesting to see how this progresses now that we’ve enabled keywords on weblog posts and will shortly do so on files … Here’s the almost-obligatory “random tags” page. Pretty much useless, but fun.

Oh, and another undocumented feature which some of you might want to use to find new friends etc: the “all new weblog posts” screen, which aggregates every new weblog post in the system. (Although, obviously, you can only see weblog posts that people have made available to you.)

Weblog posts can now be tagged with category keywords. As is the standard, keywords are comma-separated; once you’ve added a couple, these can be searched. For example, if you click on “elgg features” at the bottom of this post, you can see every post I’ve ever made about Elgg features. If another user tags a weblog post with “elgg features”, the system will then show their user icon at the bottom of the page, and you can click through to see every post they’ve made about Elgg features – and so on.

Technorati is an interesting site for exploring what’s semi-ironically known as the Blogosphere; furthermore, it has tags for categorising weblogs, and an API so you can plug stuff into it.

Book Talk is an interesting application that picks out the books webloggers are talking about at the moment. Unsurprisingly, given his sad death recently, the list is dominated by Hunter S Thompson – when that dies down, I expect there’ll be a much wider range of titles. I’ll be interested to see if they follow the New York Times bestseller list, or if the Blogosphere is a more original collection of people. (Incidentally, they’ve also got News Talk, for news stories; to be honest, though, Google News is a far better current events aggregator.)

Cunningly enough, each of the books also has a link to Amazon which includes an affiliate code, so Technorati is making a percentage of everything you buy if you click through. One to file under “I wish I’d thought of that”, methinks.

To reiterate what’s been said in both Dave’s and the news blog, we’re looking for programmers to help us develop the system. If you’ve got some experience with PHP, HTML and SQL and fancy helping out, please contact either of us and we’ll take it from there.

In the meantime, apologies for the lack of new features since November – more is definitely to come …