Microsoft is Dead

Great points. I think that part of what microsoft doesn't understand is that it's no longer completely about the bottom line. Customer satisfaction is the driving force of the web 2.0 movement, and while I think that microsoft is hardly dead, it needs to evolve a great deal (basically reiterating what Paul said). I disagree that the desktop is dead though, I think that many people overestimate the capabilities of the of the web as it is today (give it a few more years). This might just be me overreaching and being completely wrong, but I believe that desktops will have a greater role in the far future because people will be able to host and manage their own slices of the internet. In a few years the desktop will be the equivalent of a glorified facebook profile and weblog hosted by the individual, and the user experience will be something of an integrated desktop/internet environment with web 2.0 companies reduced to widgets downloaded onto the desktop. Maybe. Thanks for the good essay Paul.

Speaking to a microsoftie this evening about all this and another lightbulb went on -- one of the things hobbling Microsoft is its historical focus on the partner as the channel for connecting with customers. Think about it -- how often do you buy a Microsoft product directly from Microsoft? Never? You always buy it through someone else, through the "channel" -- which might be a reseller, a var, or an oem. This disconnection from the customer is killing microsoft in the NEW world...

Might be true if you live in "Silicon Valley lead shielding" talking only with startup guys. Completely wrong outside. How many people know Y-combinator? Here you are...

Awesome.

Old news! I've been using Linux since 1993, been involved in startups since 1999, and about the last time that I paid any attention to Microsoft was when they launched .NET (around 2003). In fact, the .NET implementation (a belated catch-up in many ways to Java) was about the only semi-innovative development effort I have seen come out of MS in the last 10 years. Even so, it was little more than a tarted-up virtual machine, a cleaned-up set of Java-equivalent programming APIs, and a couple of cool features like crypto-signed DLLs, a global assembly cache (hmm - where do you put all your JAR files, then?!) and the ability to mix and match programming languages in a single application (though only one target language per DLL, or more strictly, per .NET "assembly").

Then, Miguel and the gang came along and wrote Mono, and pretty much re-implemented everything of value from .NET, without the licencing overheads of running in a windoze only environment. That basically took the wind out of the sails for Microsoft, and they've been floating around, almost impassive, pretty much just watching the emergence of ASP/SaaS computing, web 2.0, AJAX, Ruby on Rails, blogging, wikipedia, bittorrent, skype, asterisk and everything else from the sidelines ever since.

I think the real problem comes back to people. It's a bit like Paul's articles on Great Hackers and what motivates them. I know plenty of great hackers (and I would like think that a couple of startups later I am at least a good one! - though I find myself more and more on the business side in recent times, growing our company and developing our people - good young hackers themselves). I even used to know a few guys at Microsoft (Australia and Singapore). The last "good" person that I knew who went to Microsoft (from a PhD programme at Melbourne Uni) joined in 2002. Since then, nobody I know even went near the place. If you can't attract great people, then you won't build great products, and everything else is a forgone conclusion. Microsoft has built great cash-cow businesses, and great fortunes for its founders, but I don't believe it will ever be a player again.