Fog Creek Software
Discussion Board

Linux as a Disruptive technolgy

This might be a really brilliant idea or the dumbest idea ever...(its kinda late), either way here it is:

Linux as we all know is a disruptive technology
(Disruptive as in defines by Clay Christensen in his book The Innovators Dilemma).

What if Microsoft started its own version of Linux and starts selling it for a fraction of the price windows sells for. The biggest competitors to Microsoft’s products like word, windows are previous versions of Word & windows. 10 years down the line it is going to be difficult to sustain the kind of revenue generation they currently have.

As everyone on this board knows it takes 10 years to make good software, so if Microsoft started now they will still be doing fine in 2014.

The other likely scenario is that someone else's (Novell) does this anyway, and Microsoft loses.


Prakash S
Wednesday, January 21, 2004

"The other likely scenario is that someone else's (Novell) does this anyway, and Microsoft loses."

Novell bought SuSE (the most popular Linux distribution in Europe). They'll still fail, because it's not technology that wins these battles.

Brad Wilson (
Wednesday, January 21, 2004

Let's see, Clayton Christensen's solution (not that I've read the sequel which came out this year titled The Innovators Solution) seems largely to be to redefine what it is you're actually doing.

So railroad companies realize they're in the *transportation* business and get into airplanes and cars instead of just trains.

What is it that Microsoft is doing that Linux is really disrupting... It seems to me Linux does largely the same things as Microsoft, just for free.
Wednesday, January 21, 2004

The more interesting point, I think, is that software doesn't have to cost anything. What will that mean for the industry? It will be interesting to find out.

Consider it, really: If I write a book, it needs to be published on paper to gain recognition (the internet doesn't count, because it is not the major venue for disseminating novels). If I design a new mousetrap, it must be manufactured. All of these are significant costs, but with software, I can decide to give it away for free via the internet (costs of bandwidth are negligable for many projects, and donated bandwidth has sufficed for many major projects).

Now, it still costs me time to develop, but I don't need to charge for it.

I can volunteer my time. I am not arguing whether it is good or bad to give things away, just trying to point out that software, by its very nature, is able to be given away, for better or worse. I just think it's interesting that there is nothing which requires software to be commercial.

Mike Swieton
Wednesday, January 21, 2004

I thought we paid for service - knowing someone will fix our problems if we pay them.
Wednesday, January 21, 2004


I agree marketting plays a BIG part in too.


I haven't read the Innovators Solution either. Microsoft has a better software development process (IMO) and are way better are marketting.

Linux is making major inroads where various flavours of unix and solaris resided. Guess who's next!

Prakash S
Wednesday, January 21, 2004

Mike - agreed.


Imagine annual licenses for Linux being handed out by MS...

Prakash S
Wednesday, January 21, 2004

Can you explain how Microsoft makes money by creating (or adding to) a new product, selling it for a fraction of what they're getting now, all the while advertising said product as Linux, which is free advertising for the hundreds of other companies selling linux and linux services?

At the same time, they're doing all this while creating a platform that has little in common with what all their certified "engineers" and MVPs have studied for, and a platform that makes it difficult to enforce any sort of tie-in to their existing products.

I haven't read that book you're talking about, but why exactly is this a good idea?

As well, why does it take ten years to create great software, and why would it take ten years to make a great Linux product?

Thursday, January 22, 2004

I dunno, explain how IBM does it.
Thursday, January 22, 2004

Long post, who's gonna read this stuff...

In the 1950s, the Russians, of all nations, were leaders in computers. Their pioneering BASM machines (built with vacuum tubes and later with transistors) were the most powerful in the world.

For a brief time, Russian computers ruled. But progress is slower where there is no incentive for progress, and the Americans caught up, and in the 70s and 80s the Russians were reduced to cloning IBM mainframes, then cloning the 8080, then cloning the 8088, then (as the pinnacle of their industry) cloning the 286. That was it.

Now, *American* 3-GHz CPUs with hyperthreading and 128-bit data buses are the norm. The Russians can still only make 286 clones.

My point?

The Linux people are so ecstatic about their KISS philosophy, and how you recompile the kernel every time you replace the mouse, and how you can pipe commands, etc... Sure, *today* Linux is stable and powerful, but how far away is Windows from that ideal, really? I haven't rebooted my XP in weeks, and this is just XP, not the end of the road.

Okay, Linux is better for some tasks like web hosting, but what huge technological hurdle is there to Microsoft rolling the same quality *tomorrow*?

Exactly as my Russians above, Linux thinks it's "a competitor" to Windows. But the truth is...

They really are just catching up to Windows. Everything on Linux is a panting, out-of-air response to Microsoft's fire and motion. But it gets worse....

Actually here is my point

The amount of features that future Windows versions will pack will leave Linux dead in the dust. Biometrics, micropayments, multimedia.

And it's not that the open source people can't "volunteer time" to catch up again. It will just be increasingly difficult: for one, there will be a shitload of proprietary formats and protocols.

If cloning Word was hard, think about cloning -- or duplicating -- Passport.

To sum up...

Linux == Lotus Improv.
Thursday, January 22, 2004

I've never read IBM financials, but the idea I get is that IBM is orienting itself towards services.

As well, you can certainly sell software for linux platforms; not everything has to be GPLed.

Thursday, January 22, 2004


Most of the important development on Linux isn't done by volunteers. Your notion of Linux is unfortunately about 10 years out of date. Linus Torvolds, Alan Cox etc all have jobs that pay them to work on Linux. They're not just doing it for kicks anymore.

Sum Dum Gai
Thursday, January 22, 2004, you're also wrong with your Russians metaphor:,1282,18024,00.html

Thursday, January 22, 2004

Microsoft delivering Linux wouldn't make any difference, they already have all the licenced technology they need to deliver a *nix, they choose not to.  Well not to any great extent anyhow.

Unless and until MS develop applications for *nix again there is no benefit to them at all.  Word and Multiplan were both available for Xenix and both dropped when Xenix was dropped in favour of OS/2.

If the trend oif public institutions and  government bodies adopting a Linux desktop continues (or even if its a real trend), then they may reconsider their position.  But for them to build cross platform applications now on top of the current codebase would be a gargantuan task.  Not beyond their means but its the kind of move that can break corporations.

Word and Multiplan were both cross platform applications, the actual application code was, if I remember aright, the same and delivered in pcode.

Simon Lucy
Thursday, January 22, 2004

Great topic, way to bring out the clueless muppets.

The chances of getting something useful said on a thread like this are tiny. Just let stupid topics like this die and not start them again. Jesus.

Thursday, January 22, 2004

You mean you guys don't remember Xenix?

Thursday, January 22, 2004


Why don't you enlighten those "stupid muppets" you mentioned in your post with your wisdom?

Everybody is clueless about something including you.

One Programmer's Opinion
Thursday, January 22, 2004

> Linux == Lotus Improv.

Improv is still my favorite spreadsheet for moderately complicated multidimensional financial models.  It had problems, but the dev team had a good handle on solving them when Lotus pulled the plug.

It's amazing how companies will throw away awe-inspriring products like Improv, yet keep pushing utter crap like Notes for decades.

If Improv had been Free or Open Source, I'd be using it on my Linux workstation right now!

Cow Orker
Thursday, January 22, 2004

Recompile the kernel every time you change mice?

Beware the strawman...

Thursday, January 22, 2004

It's pretty clear to me that MS's strategy against free commodity Unixes is to continue to add features to differentiate Windows. Longhorn is full of stuff that can potentially give Windows a big lead over Unixes.

This is of course in addition to the massive fud that they spew.

Cluless muppet
Thursday, January 22, 2004

Linux is more an accidental disruptive technoloy.  UNIX was dying a slow death in the late 80s.  People wanted something better.  I was a UNIX user back then, and while I freely admit it was the most stable system available, it was also obviously very behind the times in equally many ways.  Then Linux came along and thrust UNIX back into the forefront, but at the same time didn't really improve things much.  Throughout the 1990s, PCs got dramatically faster, which made UNIX much more attractive than it once was, but at the same time there wasn't much advancement.  Sure, we've got better versions of gcc and Perl.  Sure we have some desktop environments that look more like Windows.  But real movement forward just wasn't there.  Now you have Microsoft making dramatic, sweeping changes to the next version of Windows, and Linux is just puttering along same as it always was, without any boldness.  And really, no one wants boldness, because Linux is just trying to be good old UNIX.  So the disruptive nature is a peculiar hanging on to yesterday's tech, much of that hanging on caused by a dislike for the vendors of other products.

Personally I like some aspects of Linux, but I think there needs to be a lot less blind praise and zealotry for it.

Thursday, January 22, 2004


MS would make money on windows as well as Linux - instead of some other company making money on linux.

It is a good idea, because if they don't do it - someone else does and MS loses amrket share & money.

THat ten years thing was in refernce to one of Joel's articles, look in the archive.

we make the same points.... if MS chooses to adopt Linux..they would make it easy enough for my mom to use.
When MS sells windows and linux - people can decide if they want to pay more for more features (windows) or use Linux which would be cheaper and get most things done.

Every platform needs applications to surive, why can't MS iuse openoffice ? Doesn't have to be cross platform, (3 rd part applications can be written to make word documents compatible with windows & unix.)

Prakash S
Thursday, January 22, 2004

> Linus Torvolds, Alan Cox etc all have jobs that
> pay them to work on Linux.

Out of curiosity, how many people are in that "etc"?

(In case it's not clear, that's not a rhetorical question -- I really have not the faintest idea how many people are paid to work on Linux full time.  I'd be interested to know how that number compares to the couple thousand devs, testers, PMs, user educators, media designers, etc, etc, etc who are presently pumping about 50MLOC into Longhorn.)

Eric Lippert
Friday, January 23, 2004

Then you need to compare that "etc" to Microsoft kernel deparment exclusively. Linux is just kernel. Most of a functioning Linux system is indeed cross-platform and doesn't depend on Linux.

For example, Bind (Linux's default DNS server), has been around for a while before Linux. You can even run Bind to drive a Windows Server domain (many do).

Egor Shipovalov
Friday, January 23, 2004

"When MS sells windows and linux - people can decide if they want to pay more for more features (windows) or use Linux which would be cheaper and get most things done."

Why would a system around a Linux kernel be cheaper? Is there some "magic" to that 5 letter incantation that makes stuff appear out of thin air? And why on earth would MS need a second kernel? The development of the current one seems to be very far ahead of the alternative you suggest.

The one thing the Linux world has over the Windows world at this point is its wide configuration options that play well to some scenarios. You can find or make lean semi-dedicated setups by picking and choosing for many tasks.
I do believe MS is very well aware of both the advantages and the disadvantages of tightly integration the platform.
I hope that the giant R&D machine of MS will come up with solutions that can give us the benefits of integration without incurring the disadvantages.

Just me (Sir to you)
Friday, January 23, 2004

I wish all these little Linux losers would piss off out the back and do whatever they do so we can get on with work.

Friday, January 23, 2004

Not really, Bill. If you were serious you would already be hard at work ignoring this drivel. But seriously, people here talk as if Microsoft has done some software innovation. When? Their strategy has almost always been to purchase a solution and absorb it. Their power lies in business acumen.

Likewise, IBM is a huge entity with business plans that dwarf others. Personally, I fear the gnomes of Armonk. . . . Gnomes, Dwarves, hmmmm, I wonder if Bill is in the market for a ring?

Anyway, disruptive is good when it disrupts a monoculture. The disruptor should be Java. But it looks like MS has dealt Sun some serious blows. The MS virtual machine has been derailed as well by the court system, so we are left with .Net

We shall see what happens if MS revenues decline over the next few years. It may be that people will decide to go legal when the licenses get less pricey. (;-) That's a tin hat to protect me. And revenues may go up as a result.

In any case, I'm glad I got my copy of DaVinci's notebooks before Bill got his hands on them.

PS  Apologies to Joel for that horrendous sentence with the little people. Just remember, "To err is human." Resume rage is an ugly thing and I hope you were able to get your cool back in place. I used to lock myself away and howl with laughter. On the other hand, one makes mistakes.

B Calder
Tuesday, January 27, 2004

*  Recent Topics

*  Fog Creek Home