I can't help but think of "640k should be enough for everybody" when I read this. Desktop computing isn't as sexy as mobile computing, but it isn't dead yet. There are plenty of innovations left to be made.
This is fair, but I also think about an essay I read (from Kevin Kelly? It might be in his new books, _What Technology Wants_), in which he points out that, basically, we've hit a plateau of about 60 MPH landspeed for most journeys. If you chart a graph of "how fast people get around on land" from 1715 to about 1950, you see a steady rise. After that, you just see... us being steady.
Ditto for air flight: from the dawn of aviation to about 1960 or 1970, you see steadily increasing speeds. Today, most planes still fly at about ~500 MPH (I think), just like they used to.
Although there are still lots of micro-innovations going on with cars, rail, and planes, they've basically hit a maximum airspeed.
I suspect something similar might be going on with the current paradigm of desktop operating systems. Since around ~2005 (when Windows XP's service packs finally caught up to virus / worm writers and OS X Tiger brought stability and full text search), nothing in operating systems has made a tangible difference in my life.
Compared to the jump from DOS to Windows 95, or from 95/98 to XP, or the late versions of Mac OS to the early versions of OS X... the current changes in operating systems just aren't that great. Which isn't to say "consumer OSes are dead"—but they've probably hit a spot where the major problems have been solved. The innovations are smaller than universal protected memory, fast, full-indexed search, and the like.
The key words above are "for this paradigm." Mobile represents the next obvious computing paradigm, and that might be a good thing. For better or worse, Microsoft has "won" the desktop and will probably never lose it. The company appears to mostly be competing with itself these days (I say, writing from an iMac).
It's interesting to see this stated now, right as Google is moving forward with automated cars. Once all highway traffic is automated we can push to much higher speeds safely (and drive much closer to each other).
Yeah, mobile is great and all (I say that as I type this on my iPad) but there is still much that can be done with larger form factors and more powerful devices. But then I'm the guy who still doesn't understand why no one even tried to come up with ways to use two mice. Imagine using one hand to grab and hold, and the other to manipulate. It's not like we haven't had USB for over a decade, and both Microsoft and Apple make and sell mice. You'd think that doubling e market would be attractive.
Two mice means two cursors, this would require window managers and the UI libraries on on of them to be largely rewritten. As they have been, for the mobile space.
Multitouch in desktop OSes has been tacked on as a layer over the traditional cursor-oriented APIs, which is why it feels that way when you use it.
You are probably not familiar with some niche markets in engineering and 3D design, where this is part of day-to-day operations. Check www.3dconnexion.com for more info.
This 60mph landspeed limit isn't true in non-North American countries. High speed trains are a reality in many other places, where they actually invest in infrastructure. And let's not forget the German autobahn (where I recently drove 125mph for more than an hour at a time… sweet!).
