The Death of Microsoft is Greatly Exaggerated

by bob on April 7, 2007

I had a good chuckle today at this naive screed from Paul Graham, a Silicon Valley renaissance man by any standards, whose shoelaces I am no doubt unworthy to tie. Still, I think he’s full of … himself on this one.

Graham’s basic thesis is that Microsoft “died” sometime in 2005, it has become irrelevant, and it sucks.

This “Microsoft sucks” mantra that Graham is so fond of is mostly a combination of sour grapes and pandering to the youthful hubris of his constituency. I am not going to claim that Microsoft is “cool”, but I’m not going to fight them, either. I’ve done very well by myself and my clients by just accepting them for what they are: a useful 800-lb Gorilla.

I’ll be the last person to mindlessly cheer-lead for a company that has produced the train-wreck that is Vista, but neither will I count out any company with tens of billions of dollars burning a hole in its pocket. Especially one that is still capable of doing a lot of things right — most importantly to me as a developer, their managed code initiative and superb developer tools and libraries.

Microsoft would never admit it, but at some level I think even they realize that fat client operating systems are collapsing under their own weight; they are unworkable. They are simply milking the last few billions of dollars out of it while they gestate other things. And the transition away from a Windows-centric world will take years; it isn’t going to happen in the blink of an eye. Inertia is Microsoft’s friend, as long as they recognize that it’s not going to benefit them forever.

Maybe it’s just the circles I move in, but I’ve yet to have a client approach me with “Hey! Dude! I want to sponsor an open source project in Rails!” It’s still uniformly along the lines of “I need an asphalt paving estimation system enhanced.” And those kinds of customers already have Microsoft technology, and no pressing reason to re-invent the wheel to satisfy my sense of elegance. Besides, a lot of Microsoft technology is elegant — C# and .NET, for example, are widely respected even by Java partisans.

Graham is right that the desktop era is coming to an end, and the instrument of its destruction is the convergence of widely available broadband with increasing acceptance of the idea of applications and services in the cloud. He even admits that Microsoft knows this. What he fails to understand is that Microsoft is perfectly capable of reinventing itself (again) and is already doing so. A day late and a dollar short as usual, but “embrace and extend” has served them well before.

There is potential for Microsoft to stumble to the point that it slides into irrelevance. It may be that they will never again enjoy unopposed market dominance. The Mantle of Invincibility in this industry may well have passed to Google. But I’m not willing to say that “Microsoft died sometime in 2005”. It’s far from a done deal. Let’s all reconvene in about five years and have another look, shall we? Maybe things will be clearer by then. And my guess is, we’ll all still be grousing about “how much Microsoft sucks” and Microsoft will still be ignoring us, and smiling all the way to the bank.

Whether you think that is good, bad or indifferent, you have to admit, it’s much more likely than that the world will be “Microsoft-less”.

Update: Paul Graham posted this clarification about his “Microsoft is Dead” assertion, partially in response to this critique from Don Dodge of Microsoft’s Emerging Business team. Graham asserts that all he means by “dead” is that people at the leading edge no longer have to think about Microsoft. I wonder though if startups have ever had to worry about Microsoft. A business plan that appeals to a start-up is probably not even on Microsoft’s radar. Inherently, a company the size of Microsoft isn’t going to pursue something on a startup scale unless they see huge potential in it. Even then, they have a history of acquiring startups after their ideas are proven.

Graham also asserts that he just wanted to be the first to call it. Well … I guess he succeeded at that!

{ 1 trackback }

infurious » Blog Archive » Schrödinger’s Microsoft
April 8, 2007 at 12:32 am

{ 11 comments… read them below or add one }

Anonymous April 7, 2007 at 7:08 pm

Sooo.. While Microsoft is dying, they still have a lot of money?

cyber_rigger April 7, 2007 at 8:58 pm

Companies are starting to sell computers without any Microsoft products.

Bob Matsuoka April 7, 2007 at 9:39 pm

I think Paul’s essential message was that Microsoft has followed IBM’s path, and I think he’s dead on.

Robert April 7, 2007 at 11:08 pm

I was under the impression that Paul Graham had defined what “death/dying” meant. That meaning that it was no longer a threat. He properly stated his definition and carried out with his explanation.

I don’t know whether you understood that but it doesn’t seem that you’re refuting his definition but rather have your own undefined idea of what it means.

Bob responds: Okay, I’ll bite. A meaningful — as opposed to merely provocative — definition of “death” from my perspective is when the Microsoft dev stack is no longer viable because the marketplace as a whole has lost confidence in it. From the perspective of the total company, “death” means that enough areas of their business portfolio experience that same problem, such that the company ceases to be worthy even of Paul Graham declaring it dead. From where I sit, with work coming out my ears that is either specified to be Microsoft-centric or the customer doesn’t care and doesn’t object to Microsoft — I see neither even close to happening. I see it coming close only in terms of the Windows OS, and even that is far from “life-support” status.

Lorien Dunn April 7, 2007 at 11:24 pm

Managed code is a good thing you think? Explain please how an extra level of instability, bloat and lock in is “good”. And “superb development tools and libraries”(!!) They STILL haven’t managed to put code refactoring into Visual Studio and it’s 2007. Try the competitions products and libraries I suggest.

Bob responds: I think you’re working off dated arguments. Maybe *you* should try VS 2005. It has refactoring tools built in. If those aren’t enough, there are plenty of plug-ins you can get. Instability? I have not experienced instability at all. I have always appreciated platforms in which virtually all the bugs are mine. I can do something about those! Bloat? In what sense is the CLR any more bloat than the JVM? Lock in? I ported one of my projects to Linux last summer — compiled it in VS 2003 and it Just Works running under Apache and Mono. If I want to, there are tools to run .NET code under any Java engine. Where is the lock-in? Also, I have sampled the competition’s products — slung a little PHP this past year even. Same book, different paragraph. After you’ve worked in a few dev platforms and languages, you realize that no one is inherently much better than the next. Different trade-offs, to be sure, but none of them represents Paradise.

Eli April 8, 2007 at 12:32 am

From unopposed dominance to irrelevance? To Ballmer, death would be a better option…!

Bob responds: I’ll have to agree with you on that one! Those testosterone-crazed, take-no-prisoner types of managers don’t serve Microsoft or any other company well in the long run. Sometimes, not even in the short run.

what? April 8, 2007 at 2:31 am

Sounds like you misunderstood the original article. The company doesn’t have to be wiped off the face of the earth to rightfully be declared dead.

MS has become irrelevant in developing areas, they are no longer growing, no longer changing. It’s just there, no longer living, a massive corpse that will take decades to fully decay.

Bob responds: I agree they don’t need to be wiped off the face of the earth to become “dead” in the sense of “irrelevant”. However, I don’t think they need to be either “feared” or “dominant” or, especially, “cool” to be quite a viable and important factor in the marketplace, either. Personally I think Microsoft is becoming irrelevant in stagnating areas, like its desktop OS, not in developing areas. IIS 7, CLR 3.5 / 4.0, and other pre-release products show terrific promise, IMO. I have a lot of company in thinking that SQL Server 2005 and VS 2005 kicks ass. They have a thriving developer community. Large parts of their ecosystem work just fine, as far as I can see. They make huge ongoing investments in R&D.

But I’m not really here to evangelize MSFT. Peruse my posts on this blog and you’ll see that I’m perfectly willing to dis them when they have it coming. Someday when it makes sense, I’ll embrace some other platform. As a solo practitioner, though, there’s not enough of me to go around to be deeply proficient with multiple platforms, so I’m not about to take the risk and investment of jumping over to (say) the Java world just because the Microsoft world is un-sexy. I mean, that’s basically Paul’s argument — no one thinks Microsoft is cool. What a specious argument for change. The objective is to delight customers, not to be popular with runny-nosed, freshly-minted, hubris-filled programmers.

For those who truly want to define themselves in terms of what they are against, rather than what they are in favor of; if you really lay awake nights dreaming of the Paradise that will magically ensue when the Evil of Microsoft is wiped from the face of the earth — get a clue about what it will take for that to happen. It isn’t going to happen because you’re cooler. Millions of people like myself will have to decide that the pain of change is less than the pain of not changing. If you think that tipping point is past, or even within reach, then you are for the most part living in a dream world.

Zak April 8, 2007 at 8:12 am

Besides, a lot of Microsoft technology is elegant — C# and .NET, for example, are widely respected even by Java partisans.

I think that says it all.

Robert April 8, 2007 at 10:04 am

I don’t know what to say then Bob. I don’t see how you can write a response article without fully addressing Paul’s point. If you both have two different definitions then it’s a tad silly then isn’t it?

If you were to say … write an article about how he’s attempting to be sensationalist. That’d be a different story.

Bob responds: So I can debate his tone but not his content? His argument’s form but not its substance? Sounds like over in the non-Microsoft world, you’re not supposed to question Certain Things! Sounds kind of restrictive; I lampoon Microsoft all the time in this space, so I don’t come at my opinion by way of being a True Believer in All Things Microsoft. On the other hand, I think Paul’s argument makes him look like a True Believer in All Things Not Microsoft. It doesn’t lend credibility to his arguments.

C’mon, Robert … I can’t possibly disagree with anyone without having “different definitions”. I do appreciate that you gave me the opportunity to clarify some details of why I disagree with Paul’s thesis. But I don’t think arriving at different conclusions means I can’t debate the merits of Paul’s conclusions.

Jamey April 9, 2007 at 2:28 pm

Graham’s definition of “dead” is an interesting insight into modernity: you are “dead” to someone if you can safely be ignored by that person. So death is a relative notion.

There now exist parts of the computer industry that can ignore Microsoft, hence to them Microsoft (or the almost hysterical terror of Microsoft they had in 1999) is dead. But in a way that’s always been true. Perhaps Graham has just migrated to one of those areas.

JeffDevVb October 12, 2009 at 12:35 pm

Its funny to hear someone call the market leader in almost all aspects of programing and development tools …Dead. Even their desktop OS’s have the lion’s share of the market, can anyone dispute those facts. I’m down with saying they are overly aggressive in maintaining their business model and could use taking down a notch or 3 but right now sending in any competitor to tackle them is like asking Tiny Tim to get that 800lb. Gorilla in a headlock…It isn’t happening in the foreseeable future.

Leave a Comment

Previous post:

Next post: