The Sad State of Smart Clients

Well the year is officialy over now, or will be within a matter of hours and like a beaten candidate on election night it's also time for me to concede victory in my bet from last year to Rob and all of the other AJAX smitten developers of the world. 

While losing this bet is of little consequence for me personally, I fear it reflects an overall sense of stagnation in the smart client world.  Windows Live services which at this time last year was launching several smart client applications has not really released anything significant since with the notable exception of the Live Writer blogging tool.  Click-Once deployment has become considerably easier and more technologically feasible as the installed base for the .NET Runtime has grown but we still see very little in the way of consumer rich client applications deployed via the Internet.  Office 2007 represents a significant improvement in what I would consider to be one of the most successful set of smart client applications, but one suite of applications can't change the world by itself.

It would seem that most software companies are willing to settle for "good enough" with the features of AJAX for fear of alienating the minority of folks either not running Windows or who have had their PCs locked in a cave such that the .NET Framework hasn't made it onto their systems.  I won't bother debating whether this fear is warranted, the end result is that the adoption of smart clients - particularly external to an enterprise, has simply not hapenned.  In the end, this willingness to settle for the limited functionality available with AJAX costs our users in their experience and ease of use of the software we are creating.  And we, the collective software industry, seem to be just fine with that.

The good news is that new technologies like WPF/e seem poised to further bridge the gap between the conceptual ease of centralized deployment and the rich content of smart client applications.  Perhaps with the adoption of Vista and thereby WPF, which I believe will be a very slow process, XAML will put AJAX scripting to rest once and for all and our users will not have to suffer with "good enough".

4 Comments

  • Stay tuned on WPF/E and the state of smart clients. I think you will be pleasantly surprised in 2007.

    Scott

  • Paul, I've been thinking these thoughts for many years - really since the days Web Services became somewhat standard I've thought that there would be some sort of fallback back into smart client applications. And year, after year I've been surprised to not have seen it happen - most development continues to be focused on Web applications and desktop development in general seems to have stagnated significantly in general for these years. In the work I do many opportunities for smart client applications was not even considered even when it would have made perfect sense.

    It'll be interesting to see how WPF/E turns out, but I'm not having high hopes given the anti-pathy against Microsoft only technology and the fact that other ubiquous technologies like Flash have minimal penetration for real applications... WPF/E has an uphill battle even if developers may drool over the idea of what it might become (if not what it is now in its limited form). It may simply be too late to unseat the deeply ingrained Web mindset...

  • Actually this is pretty much true. I rem how excited I was with the "Click-Once deployment" but at that time I never realized that Linux and MAC were so popular among the Tech Crowd. In this age if something isn't cross platform its going no where. One reason why Firefox will continue to be adopted at a larger scale.

    @ScottGu isnt WPF/E a Subset of .net 3.0 Framework. .net is not Cross platform so even WPF/E work on the client side i.e Cross platform it wont make a difference. Smart clients Are dead and are going no were even with the release of WPF/E nothing is going to change since MS is not going to release a Linux Version I saw the Video at channel9. Where WPF/E was show for the first time and I remember the guy asking "So will this work on linux" the Programmer clearly avoided the question and answered back saying we will provide the tools so that People can make it cross platform.

  • I rather have companies think about those minorities instead of ignoring them imho :)

Comments have been disabled for this content.