Scoble and the Microsoft Longhorn team are right: it is time for rich clients to return. I, for one, am still looking for the return of Pointcast -- it was by far the best tool I have ever used for daily review of news headlines and a stock portfolio. Sure, the web is free and easy but it wasn't designed for high-performance interfaces. Thankfully, it works great for stuff like blogs -- its simplicity is incredible for micropublishing.
But not for complex applications. You've never heard of Lacerte, the largest maker of professional tax preparation software (and of late a division of Intuit). They sell to accounting firms which pay $3000 or so each year for a key and indispensible part of their business. Products like Turbotax (also Intuit) are nowhere close since they don't provide the level of coverage, high-speed data entry, or volume features that professionals need. Lacerte, beginning in 2000 or so, made a very ambitious effort to port their entire package to the web. At first, it was free for beta testers. I think the firm I used to work for paid an extra $100 for it last year. Their theory was it was an investment in disaster recovery. Because it just wasn't useful for a tax professional. It was great software -- I used it the past couple years since I didn't need to install an 80MB thick client on my machine (I still provide IT support for the firm remotely). This was as good as it gets for a browser-based ActiveX thick-client accounting solution. But while it was fine for me doing one personal tax return (fringe benefit of working there), the extra minute or two each hour due to noticeable pauses and sluggishness made it unacceptable for the accountants. A pure web-based solution would be even worse. Lacerte announced this spring that they would no longer offer a web version due to lack of interest.
People want snap. They want instant response. They want things happening in the background, predicting what they do next. The web gave us freedom: from captive data and inaccessibility to small businesses and individuals. The web will always be there as the lowest common denominator. But there is no longer any excitement in browsers. I used Netscape recently and what struck me was the snap. It seemed much quicker than IE. But not enough to face the compatibility issues (having both installed on a machine can be truly painful with tweaking). There is no hope for Netscape. Microsoft has given up interest in IE (no money). .Net and Java are getting the little guy back in the game. Loads of power are lying there untapped (unlike the 486/66 I began surfing the web with). It is time for the pendulum to swing back to rich client, if only to accelerate the following swing to the future of the Net (which we are making dismally insufficient progress towards). The gradually building shift among the blog cognoscenti to RSS aggregators is the first indicator that rich clients are returning to the desktop.
Hats off to Microsoft for thinking two steps ahead of the game. After all, Longhorn is it until 2009.