There’s a lot of current debate on the growing complexity of Web services (big W) which in general conversation most take to mean the SOAP/WS-*/etc stack. I’d outline it, but Martin LaMonica’s done a great job of that already here.
I’ll perhaps explore that in more depth later, but what I’m more interested in at this point is the potential emergence of simple but highly useful web services (small w) with consumer facing web applications.
Flickr was one of the first to expose its API along with del.icio.us. Now just a few days ago, Bloglines followed suit.
The product of all this? Composite applications that blur the line between web and local clients simply (much like Longhorn promises to do, actually), and actually offer very perceivable benefits to users. It’s interesting that many of these clients began as primarily web based platforms, but I’d argue that that’s just a natural consequence given that the real power is not, as I’ve argued (ad nauseam), in the client software but the network.
Looking down the road, the implications of these web services are as broad as they are simple:
1. Metadata driven local bookmarking might actually give us a system that actually works – take something like this that sync with a del.icio.us account, for example, in the Bloglines/FeedDemon fashion and now you’ve got something
2. Family/friend group photos from Flickr integrated into an OS and serving as wallpaper? Why not?
3. Bloglines/local reader combinations that not only provide seamless transitions in read/unread items, but connect to local information crawlers like X1 or Beagle? Sounds good to me.
The lesson learned here, for me anyway, is that simple protocols != simple functions. Of course, we all knew that (TCP/IP ring a bell?), but nonetheless I think it’s possible that these basic – or primitive, as some would call them – web services may have a more immediate and profound impact than their enterprise counterparts. There’s room for both, in my view, but I’m more excited by the simple at the moment.
