Microsoft's long demise

I belatedly discovered, albeit with great interest, the writings of Paul Graham.

For those of you who don't know the guy, I highly recommend you take 2 hours one day and go and read some of his essays. He apparently wrote the seminal book on LISP, founded and sold an early internet company called Viaweb and more recently is the dynamo behind "ycombinator" a modern version of the incubators of yesteryear.

I mostly recommend reading the guy because he has a pulse on the "geek zeitgeist". I believe he correctly represents a large portion of what geeks think. Some of it is right, some of it wrong. In 2003, he wrote, a wonderful entry called "why nerds are unpopular" which is a must read for every geek in America. It completely captures "geek angst" I witness in teens in America. He is full of hope and optimism.

Paul Graham has recently caused a bit of a stir by writing a hugely popular entry called "why Microsoft is dead". His thesis is not so much that MSFT is dead, in spite of the attention catching title, rather he argues that MSFT is not feared as it used to be.

He argues that the internet, AJAX/RIA, apple, and google are the main reasons for MSFT dismissal. I thought about this for awhile. While I agree with the real conclusion, namely that MSFT is not as feared as it used to be, I mostly disagree with the reasons he states.

Missed Java and Linux.
At the top of my gripe list, he misses both Java and Linux, and the combination of the two as one of the major reasons for MSFT's failed attempt to conquer the world of the web.

See, portability was always what Java promised, and delivered. From the beginning we java puppies were insulated from Operating Systems, we just didn't care. Whether that platform was Windows or Solaris or Linux or AIX or Mainframe or Mac was truly a secondary concern for the java guys.

In a sense, I believe that Linux and java have robbed MSFT of growth on the server side. In the mid-90's we were all shaking in our boots that MSFT was going to eat everyone's lunch on the server side and that NT was a juggernaut no one could stop. How things come to pass! Linux came along and put a definitive end to that.

Server side linux is a huge success story if anything because it stopped MSFT in its "tracks of inevitability". Java was a coup de grace and it freed legions of corporate developers. Java on Linux means a lot of corporate apps work as is on Linux. The rest was history. We never looked back and certainly didn't see windows as an evil not even really a factor, just as a vehicle of distribution for java technology.

New markets out of bullying reach.
In thinking about the reasons Paul Graham invokes, I find myself agreeing with the internet one, of course, but not for the same reasons. I believe many of the new markets that emerged because of the internet (e-commerce, search, portal, software distribution, social networks, peer to peer) were mostly immune to MSFT attacks.

The main reasons is that unlike their PC equivalent the web and internet applications:

1- do not care what OS you use
2- are not bundled with a PC

In other words, MSFT couldn't leverage their client OS monopoly to bully distribution. It bullied distribution with IE and that was the end of that. From there on it was freedom.

Google, sure, but.
Google is certainly a reason why MSFT cannot come to dominate the early web products (search, SaaS office apps), but Google robbed MSFT of web-products related growth more than it is threatening its core franchise. I mean in the same sense that Linux robbed MSFT of Windows growth on the server more than it came to threaten windows on the client.

The Windows franchise is certainly NOT under threat, not by the likes of Google despite all of the excited noise on WallStreet. Google doesn't deal in OS. Until they do, it is not an argument. Even if it did I would argue that it would be a long hardous road for them. Maybe if they bought Ubuntu... mmmm, maybe.

The Office apps? sure... but not yet. I use Google apps, I love it, but seriously.

Mac? don't make me laugh
The argument Paul makes that Apple's MAC-OSX contributed to the supposed demise of MSFT is just laughable. MacOSX still shows up with a paltry 4%. Oh and please don't come with the Gandhi quote, it is not because people ignore you or laugh at you that you inevitably win. Doesn't work that way most of the time.

I love the "I am a Mac, I am a PC" commercials as much as the next geek, they have a funny bone about them. But lately I am starting to suspect that these commercials maybe generating a backlash. Think about it, who IS the funny guy, who do you feel bad for? In other words you do you emotionally respond to? the smug mac kid or the PC guy? I like the PC guy, and he is a regular on "the Daily Show".

Stateful conversations as RIA enabler.
I hate to bash Paul on this one but I believe he is just trying to sound technically deep, and misses the target. Paul invokes the "HTTPSession" objects as an enabler of rich applications and one of the reasons MSFT is being humbled. He must have heard that in a VC meeting from a kid developping a web2.0 app, remembered he didn't have such an object in LISP and thought "wow, cool". Ah, what a curse it is to lose one's technical edge :)

Java had "HTTPSessionContext" objects since the late 90's. Second these developments reflect the progression of modern OO language to catch up and provide stateful conversations for applications over HTTP. You can really view this as a "half empty glass". Further proof is that these technologies are still under development (SEAM with webbeans is but an example).

In other words, while RIA/Web 2.0 rich applications are a potential future threat to desktop applications, the technical story is still unfolding, largely undecided and therefore cannot account for MSFT's humbling in the past. BTW, the topic of whether web apps will replace desktop apps, at all, is still hotly debated by geeks.

In MSFT's defense
Despite widespread belief to the contrary, there is a lot of innovation to come out of MSFT. A lot of basic research too. MSFT bashing was, still is popular in certain development circles. I get annoyed at the persistent myth, widely believed in OSS circles, that MSFT doesn't innovate.

You have to admire a company that has basically told Wall Street to "fcuk off" with its short term greed. I remember vividly a Wall Street investor commenting privately about how pissed off they were that MSFT just ignored demands by Wall Street in general to reduce their R/D, that they not invest and show margin expansion. There is an agenda of innovation a culture of R&D and corporate will and spine to keep the Wall Street greedier elements at bay, go them!

One thing that causes me more alarm, is that I see a lot of startups using MSFT technology these days. See, startups give you early warnings on trends of what people use. Something is going here. Startups are the early warning signs of on-coming floods.

Finally, I love my XBox 360 with billg's signature on it.

As the French used to say "The kind is dead, long live the king!"

marcf