Web 2.0 & My Data

Typepad, Bloglines, del.icio.us (poor del.icio.us). Even the mighty salesforce.com. They’ve all suffered outages in the past couple of months. Services that we’ve come to rely on, not just as luxuries but in many cases necessities, are susceptible to the whims of all these wires and cables and hadware strewn all over the continent. I’m not even talking about the frustrating slow-downs and lagging pages on other hosted services that don’t crash per se, but just come to a painful crawl on occasion.

Which I think leads me to ask the question: Why are my measly 85 bookmarks dependent on all this…stuff? Yeh I know, I’m loving the convenience of having my bookmarks centrally hosted so I can access them from anywhere, but should they disappear if del.icio.us goes down? I don’t think they should. And now people are extolling the virtues of using a word processor or a spreadsheet inside your browser. I can see the initial appeal of this. No installation. No worries. Etc. But there’s a problem here.

As we think about Web 2.0, we should slice up the pieces that are getting delivered to us through the Internet. There’s interface. There’s logic. There’s data:

Interface. This is what web applications are all about. You deliver the interface over the wire via markup (with some CSS mixed in) and you’ve got an application. For a long time, it was a very static, page-based interface, but an interface nonetheless. Now with AJAX, the interface is feeling a lot more dynamic and a lot more like a desktop app (when used properly). Good progress.

Logic. Historically, we’ve tried to keep the “brains” on the server as much as possible. Sure, Javascript could be used for validation, etc., but that was generally frowned upon. Again, used properly, AJAX can make this happen a lot more seamlessly. Server-side validation can elegantly inject itself into a user’s experience without Javascript pop-ups or full page refreshes. Again, good progress.

Data. Now it gets cute. Another appealing quality of web applications is the ability to store your data centrally. I think the Web 2.0 hype has sort of lumped in this aspect of web applications with others. I think this has led to a gross over-simplification of how data should be stored centrally.

Of course I would like my bookmarks centrally stored. And yeh, it would be nice if my documents and spreadsheets were centrally stored somewhere (maybe). But why should that result in a full reliance on some other system’s infrastructure for me to get to my information? If del.icio.us is down, then I can live without tagging for a day or two, but I can’t live without my bookmarks.

There is a point of failure and what lies to the left and right of it. Make sure my data ends up on my side if the chain breaks. When it eventually reconnects, sync it back up. Now I’m fully aware that the web browsers can’t sync locally stored data out of the box. That problem will get solved with rich applications that have the proper hooks to sync as needed (XAML, some Flash derivative, etc.).

We can still have collaborative software like Writely and others without forcing individuals to be left without their data when something goes wrong. The interfaces around word processing and spreadsheets don’t change all that much. Beyond the “neat” factor of seeing a rich word processor in a browser, the functional value is limited and the risk of being left without my documents far outweigh any benefits.

I think there’s a sweet spot here that marries the richness that we’re coming to expect from zero-install applications and the reliance (and comfort) we have with knowing that our information is in our hands, yet synced centrally as-needed. I don’t think we’re there yet.

2 Comments Web 2.0 & My Data

  1. rob

    Yes. But wait, there’s more.
    Privacy. Security. Confidentiality.
    Go try and sell a bank a backup solution …
    Some people can handle putting their data at risk. Usually it is flimsy, non-mission-critical data. What if it were their social security number and personal email address? Hmmm …

    Reply
  2. JD

    Spot on.
    I don’t use centrally hosted apps for critical data specifically for this reason. I have no *real* guarantee that they’ll even be in existence tomorrow morning.
    Of course that has led me to re-invent the wheel tens of times when a new service comes out, but that’s another story.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *