Earlier this week, at SIIA S4 conference, I participated in a short panel, nicely moderated by Newsgator's Greg Reinacker on Desktop vs. Webtop. The key premise was: since a lot of activities seem to be moving to the Web, do we still need a Desktop? This, of course, is a debate that has been discussed in the industry for years; it reminded me of the 1984 (yes 84 no typos) phrase coined by Sun's John Gage "the network is the computer".
In the panel (and outside of it as well) my position was the following: Yes, a lot of activities, since 1984 apparently :) have moved to the web, or more accurately, have benefited from being connected to the network; but No, I do not see it as a replacement of the desktop but as an indication of what "current/next generation" desktop applications must become.
It is quite clear to me that "everything else being equal" no one would prefer having a web interface over a local, rich and interactive experience. For example, there is no way I would prefer doing my emails using OWA instead of Outlook. Clearly, based on the success of the browser and application such as GMail, "everything else must not be equal". The core question is therefore, what are the features that drive people to the Web and accept inferior user experiences? I think they can be reduced to 3 things:
(Just to preempt a series of comments/emails; I am explicitly not including the "not having to run the operation myself" benefit and concentrate on the user experience part. Not having to run the "back end" and get the service from a SaaS vendor is true regardless of a web interface or smart client one.)
So going back to the Desktop vs. Webtop, the goal is not having a cloud-only service with a browser-only interface; cloud-only + web interface is a mean to satisfy the 3 attributes above. If we could satisfy the 3 features above while maintaining a rich experience the result would be even better. In other words, the debate is not Webtop vs. Desktop, but more: let's make sure that the design principles of the emerging applications include rich experience + anywhere access + collaboration + trusted installation.
If you think about it, this model is clearly what is happening in the industry. "Browser-only is not hip" anymore. Salesforce.com is shipping a offline edition, Adobe is shipping Apollo and of course Microsoft with Silverlight, WPF, Office, MOSS etc. has always been embracing the smart client connected to the cloud service model.
Don't get me wrong, I am not suggesting that having a browser interface is wrong, it has its merit and certainly its place in today and tomorrow's applications; it is just that (a) it is not the only option that vendor should consider/provide and (b) not something that users should necessarily settle for.
As a small side note, It was ironic (funny?!) for a Microsoft guy like myself to see Google's Matthew Glotzbach delivers his Webtop session (just before my panel) at SIIA, describing that everything was moving to the Web, the browser was the only thing you needed etc. using a very "local" PowerPoint presentation :)