Khaotic Musings

I get to hear about the biggest threat to Microsoft's hegemony and
market dominance quite a bit. It often goes something like this

"Linux isn't Microsoft's biggest threat, Google is."

Alternatively, as raised recently by ZDNet blogger Adrian Kingsley-Hughes:

"The biggest threat to Microsoft’s bottom line isn’t Mac or Linux -
it’s netbooks!"

Adrian continues:

"The problem isn’t that netbooks are cannibalizing Windows sales
(they aren’t, especially when you take into account that Windows-powered netbooks out-sell Linux models by a significant margin), the problem is down to the fact that Moore’s Law has finally caught up with Microsoft and the OS is rapidly becoming one of the most expensive components of a new PC. And as hardware prices continue to fall (which they will), this is only going to get worse for Microsoft."

It seems that Microsoft's margins are being squeezed because netbooks
are on the scene, and netbook prices are plummeting, putting further margin pressure on Microsoft.

Now, here are some questions about these supposedly major-domo competitors to Microsoft, and I'm sure, that you can easily guess the answers:

1) What was it that made Google such a serious net platform player, with enough power and flexibility to serve the world's Internet needs?

2) What was it that made netbooks possible in the first place?

3) What gives netbook manufacturers enough leverage to screw Microsoft down on OEM licence costs?

4) What is it that makes products like Amazon's Kindle and hundreds of other e-book, portable media and mobile Internet gadgets possible, all by eschewing Microsoft's embedded OSes?

The answer of course, is Linux.

Without Linux, there would be no Google. No startup could afford to
build a platform on hundreds of thousands of servers which required
either proprietary hardware (ie, Sun servers, circa 1998) or proprietary
system software (ie, Microsoft's Windows Server.) Further, how
comfortable would Google be in competing with Microsoft in the Internet space, if Microsoft 'owned' its server platform, the basis of Google's ongoing business?

Without Linux as a viable desktop play, Asus could never have produced the Eee PC 701. No Linux. No netbook market.

Furthermore, without the presence and ongoing viability of desktop Linux for small form-factor devices (eg, Eeebuntu, Android), there's no way that Microsoft would have done an about-face on scuttling XP in 2008, nor could they have been beaten down on OEM licence costs. In fact, the sheer presence and viability of Linux puts an upper limit on how much Microsoft dares charge for Windows 7 OEM licences for these netbooks.

And once ARM-based netbooks start selling for under US$200, removing any "breathing space" for OEM licences fees at all, watch Microsoft get squeezed out of this market in a big way. As an aside, Microsoft's key OSes and the 'Windows software ecosystem" don't run on non-x86 CPUs like the ARM; the Linux ecosystem does, and has for many years. (I remember being involved in a project to build a Linux-based web-tablet, on an ARM platform, in 2000.)

So, everytime I hear pundits say that something-or-other-besides-Linux
and open source is the "real threat" to Microsoft, I reach for my
clue-bat.

And now you know why.

Matthew McKenzie over at InformationWeek raised some points recently which got me thinking about the level of uptake of desktop Linux.

Among his comments were:

"Conversely, Windows 7 delivers a clear message to every Linux distributor aiming for a piece of the business IT market: Aim high, execute flawlessly, and deliver amazing service -- or save yourselves the trouble and don't even bother trying."

In general, I like Matthew's pieces, but I think he misses the mark with this perspective, and here's why. Microsoft itself needs do none of these things in order to sell hundreds of millions of Vista licnces. It needs deliver neither a flawlessly executed product, nor particularly good, let alone amazing(!), service.

In trying to get more businesses to adopt Linux and open source software over the past 18 years or so, I've discovered that no matter how good Linux becomes, one hurdle stubbornly remains - in fact, it's the biggest hurdle that Linux and free/open source will ever face: vendor lock-in.

Here's what makes vendor lock-in such a difficult customer for any market-newcomer like desktop Linux and free/open source software:

It doesn't matter if Linux has become better than Windows if most users are too afraid to move to an alternative desktop.

It doesn't matter that Linux uses an open platform API, if most line-of-business apps are written to the obfuscated and proprietary Windows API, and Microsoft does its absolute best to ensure that few of these apps 'fly the coup' to Linux's open platform.

It doesn't matter that OpenOffice.org is free and uses a truly open document standard, if most users are tied to Microsoft's office suite because a proprietary binary file format is almost impossible to inter-operate with and thus completely convert documents over to OpenOffice.org.

It doesn't matter if Mozilla Firefox is truly standards-based if so many 'core-business' sites, banking, e-government etc, mandate the use of Microsoft's Internet Explorer 6, with its corresponding support for broken HTML.

Yes, in the long run, even vendor lock-in roadblock will be removed. However, in the meantime, industry observers must bear in mind that much of the retardation of market penetration that Linux and free/open source software see isn't inherent to that software, but something largely out of its control.

I've been working on a project to outfit schools with 'spray and wipe' versions of live educational USB-key Linux distros for netbooks, for quite a while; I'll write about this work on here sometime.

I was therefore interested to hear about another project, with similar aims, being run by one of my colleagues here at work. While his team's efforts aren't education-industry specific, they are very applicable within that context.

Briefly, here's what they've built: Imagine a live Linux USB key, which, upon bootup on a PC or laptop, pops up a menu offering a range of distributions to install. One of these is Edubuntu, the education-oriented version of the popular Linux distro. The user selects the required distro, and then walks off, perhaps with another USB key, to the next PC or laptop, and so on, around the whole school computer lab.

The installer will proceed to do a complete install, including all those apps that an average desktop Linux user wants, but which aren't installed by default, codecs, plugins, mplayer, educational software, etc. It fetches the latest package updates, and can even email a hashed version of the newly-created system password, in a secure way, to a designated address, say the school IT manager's. It also sets the OpenOffice.org default Save As documents to be DOC, XLS, and PPT, to make integration into a school's Windows network 'community reality', easier.

Using this USB installer system, a non-techie teacher can build a complete education-oriented desktop, on a PC, laptop or netbook, with about 2 minutes of actual effort, and only one decision (ie, what distro to install).

With a handful of such keys, a lab of 15 PCs can be built in about 2 hours. All this, with no requirement for any technical expertise.

As 2GB USB keys, the storage space needed to support this installer, are only a few dollars apiece now, this solution scales without much expense.

Yes, none of this technology is rocket-science, but that's the point. Linux and open source software make it feasible for even small, minimally resourced groups to build solutions which benefit users and communities, particularly communities which don't have substantial funding, like schools.