XForms and Internet Applications: Web 2.0, Copernicus and Spartacus: Moving the centre of the web

Although it was an amazing mental feat for Copernicus to finally nail the idea that the earth was at the centre of the universe, he did replace one central location with another--the sun. In our enlightened times though, we know that there is no centre...or perhaps post-Einstein/post-Foucault, we might say that anywhere can be the centre.

I've had this image knocking around in my mind for a few years now, because it's the way that I think about the internet--as far as I am concerned, I should be at the centre of the web, and at the same time, so should you. When each of us uses the internet we want information to come to us, rather like a spider waiting for a tasty insect meal, and so, as with Einstein, where exactly the centre is depends on your 'frame of reference'.

I'm Spartacus

That might seem obvious to some, but it's not the internet that we have at the moment. We have many, many servers vying for our attention, offering to store all of our data, process all of our feeds, talk to people on our behalf, and each of them presents itself as the centre of the web, with each of us as satellites or appendages. This is the portal model, and of course it works for a lot of users, but the very fact that things are changing and evolving in the direction of Web 2.0 shows that it certainly doesn't work for everyone.

Since many of those central points don't quite do what many of us need them to, alternatives spring up that offer themselves as a new centre that can bring together exactly the things that we are looking for. The problem with this is that each time we add a new possible 'centre' we create more fragmentation, and, it has to be said, a new nervousness. Should I have my diary with a portal like Yahoo! or on HipCal.com? Since calendars are much of a muchness, the answer might be whichever one gives me the best set of additional features, like contact management, or diary sharing; these features might be all I need...for now.

But come next week I may really wish I'd chosen the calendaring system that gave me maps to my destinations, and sent a message to my mobile phone telling me what time I needed to leave to get to the meeting. What do I do now? Do I move my data? (Oh no, not again, you say, having just spent hours typing in your kids' school holidays and your parents' wedding anniversary.) Or do I stick with the service I chose, and hope that it improves pretty soon?

But how can it hope to improve in all of the directions that I want it to go in, when each day I'm coming up with new things that I want to be able to do? And of course, just as I have my new requirements, so do you, because the exciting thing about what everyone is calling Web 2.0, is that each time we see that something is possible we immediately think of something else that we'd like to do.

The growth in mash-up services is a reflection of the desire we all have to solve this problem. Services are set up that try to aggregate other services, but on the negative side, bringing a number of other services together on yet another server just creates yet more centres shouting "look at me" ("I'm Spartacus").

But the positive side of mash-ups is really positive, since unlike portals that simply try to get you in with any bit of information they can think of (horoscopes and TV listings), many mash-ups consist of some clever juxtaposition of pieces of information--in other words the really exciting bit is the very fact that the data has been combined, rather than the way that it was achieved.

Building Web 2.0

Ultimately I don't think Web 2.0 can be built on server mash-ups; if we have to keep adding new servers to get the kind of interaction we want then something is very wrong, and this is what we will have to do until we put each and every user at the centre of the web.

This means that Web 2.0 will continue to evolve and develop, showcasing ideas for the explosion to come, but for the great leap forward that will finally warrant the step up in version number, I think there need to be two developments.

"Show me the data!"

The first is that many sites need to slim down rather than rack up, or perhaps new sites coming online will just be more focused. Either way, my point is that if you run a calendar site, why not just stick to that? Why add contacts and maps? Why not just create a simple RESTful API so that we can all upload our data, and then let us query it and get the results as RSS 2.0? Add value by having loads of other calendars that we can subscribe to, like public holidays, 'this day in history', scheduled roadworks, and so on, and being the best source of data for calendars.

Of course lots of sites already have great APIs, so this is not exactly a prediction to rival Nostradamus'. But I think the point is that it will increasingly be the power of the APIs that determine which service you will trust your data to, not the web interface, lovely as they usually tend to be.

Rich Clients

The second development is actually the continued evolution of the rich client, by which I mean putting more functionality into the software on user's machines. This is the crucial development for Web 2.0, since to put me (and you!) at the centre of the web, I need to be able to build my own mash-ups, on my machine, accessing data sources that I want, and combining them with each other in ways that I choose.

For example, to put geographic information onto my images in Flickr I can either put them into the tags on the images themselves, or I can create a reference in del.icio.us and tag that. The second technique is way, way better than the first because it means that I can also add geographic information to someone else's images, and I can also tag images that I have stored in other online photo stores.

But as of today I cannot realistically do anything with this data unless I write a server-side application that combines the data for me. Of course that isn't difficult (even without Rails) but by doing that I have simply created another Spartacus. For the decentralised-web-with-each-of-us-at-the-centre vision to really happen we need a new kind of client application that allows me to drag one lot of data and drop it onto another, and for something entirely obvious and meaningful (albeit very clever) to happen. Users need to be in full control of their own mash-ups, without the need for script or mark-up.

Web Application Viewers

But there's no way we can build such a client for the end-user if we stick with the view of the browser that we have today. A web application viewer is as likely to talk over Jabber as today's browsers talk over HTTP (and may do both at the same time). It needs to be able to learn what web sites you might be interested in (and the people you trust) from the email you receive and the online chats you have, just as easily as it could from your browsing patterns. And it should be possible to extend such a web application viewer with new viewers and processors, for new data types, as easily as you can download a web page today.

That's the distributed web I've been working towards for a number of years, and it's the reason why I got involved in the XForms and XHTML 2 initiatives at the W3C...because I'm convinced that we need new languages for this new client software, and we need this new client software if we are going to build a new web.

But despite the scale of what we need, I am feeling more excited about this than I have been, and that's thanks to the level of experimentation and discussion in the space that is being called Web 2.0. It's not that so much because the answer is out there now--in fact there are lots of old ideas knocking around that have been dusted off, and there is certainly a lot of work to do--no, my excitement is caused by the buzz, that indicates that people are ready to solve the problem. After all, the reason for the interest in Web 2.0, Ajax, mash-ups, social tagging, and so on, is that people really want 'version 2', and seeing more of what it might look like only makes us want it more.

Which in passing, is why I don't see any point in complaining or saying that Web 2.0 is already dead--if we naysay the name, the ideas and enthusiasm that it represents won't go away, so it seems to me as good a name as any to capture this exciting period. Everything I read today leads me to think that whatever we choose to call it, most of us are looking to achieve the same thing, as--in some small way in the spirit of Copernicus--established practices are being overturned again.