Update: Unfortunately, since writing the first version of this post, Paul Graham has released an update and clarification of his original post. Obviously, I should start blogging over the weekend and give up this whole “having a life thing”. 😉
In terms of disclosure, I attended the first Startup School in Cambridge in 2005. Other than the obligatory “thanks!”, I've never communicated with him and have no interest in Y Combinator, etc.
Paul Graham has a tendancy to ruffle feathers. I have a number of close friends who's reaction is anywhere from annoyance to outright disgust of everything he says. So when he puts forth his newest premise, it's sure to tick off a few people. He believes the overwhelming threatening specter that is Microsoft is dead. Not only dead, but he believes it's been completely replaced by none other than the new beast on the block: Google. He identifies the four major aspects of this decline as: 1) Ajax, 2) the rise of desktop-like web applications, 3) broadband Internet, and 4) Apple.
The first all tie closely together, so I'll address the last first: Apple. In the past few years, I've noted that most of the smartest hackers I know either own Macs or are lying in bed dreaming of them at night… like myself. Woz noted the sheer number of Macs at the first Startup School. I was finally convinced last fall looking over Mike Ho's (of QCodo fame) shoulder during DCPHP. I saw him developing and testing a .Net application on IIS via Parallels on his MacBook. Unless you're developing a desktop application, I don't need to have a Windows box at all. Even if you occasionally need one, virtualization can handle it. So on this point, I have to agree with Paul. Microsoft is effectively dead in this space.
Now, on to the bulk of his points… is Microsoft dead?
Dead? No, I don't think so. They still have massive revenues and profits, massive cash reserves, and a product suite so deep and wide that no other company is even close. So no, of course they're not dead… but as described above, they might be irrelevant. Let's look at the recent past.
Windows XP was leaps and bounds beyond the previous generations of Windows 95 and 98. The stability was just the most obvious aspect. Has Windows Vista been a similar leap? In the years leading up to the release, the coverage focused on delay after delay and which features were being dropped. In the months since the release, the bulk of the discussion I've heard is centered around a) complaints of “Vista-capable” machines being marginally functional at best, b) incompatible hardware drivers, and c) incompatible software. Each of these individually might not be deal-killers, but all of them combined together can make your computer an ugly paperweight.
As noted yesterday, I recently received the opportunity to upgrade to IE7 and get tabbed browsing. Unfortunately, Microsoft is the last one to this game… which further leads to Paul's Ajax point. In the distant past (think: Netscape-era), they could change internal libraries and behavior to make life difficult for third party developers. In the present, there are so many people using so many web applications, they don't have a choice but to behave. If they choose to make life difficult, there are alternatives and the willingness by users to use them. Combine this with the sheer number of vulnerabilities, etc that have come along for IE6 and most users just a bit of doubt that Microsoft has their users' interests in mind… Quite a bit different than the last Browsers Wars.
So basically, is Microsoft dead? Not a chance.
Are they irrelevant? To anyone focusing entirely on web development, yes.
Microsoft may have to change strategies eventually, but eventually they'll figure something out. Maybe they'll go the skunk works route and come up with something amazing. Maybe they'll comb through the Open Source Community and cherry-pick the most interesting and promising ideas to bring in house. Maybe they'll give up and get into live stock. I honestly don't know… or really care, but Bill Gates and crew didn't get rich by being dumb or lazy.