Khaotic Musings

By now, you've probably caught up with the news that (apparently) hell hath frozen over, and Microsoft has contributed GPL code to the Linux kernel.

One of my colleagues at work also raised an eyebrow of surprise that Microsoft is also a Platinum Sponsor of the Apache Software Foundation, considering that Apache httpd is the biggest competitor to Microsoft's own IIS web server.

Neither of these moves should really be a surprise.

Microsoft's strategy should be (and is) to do all that it can to run as
much of its technology stack, on as many computers, as is possible; this translates into ensuring that Apache, which has far more market share than Microsoft's own IIS server, runs as well as it can atop
Windows, and therefore makes Windows a viable host for what would
otherwise be an ocean of *nix-only LAMP apps.

Obviously, Microsoft would prefer that everyone used IIS to deploy web systems, but that's not going to happen. It therefore has to accept the second-best option of trying to ensure that Windows isn't also cut out from the web market, by supporting Apache (and MySQL and PHP) as well as it can. Thus the support noted above.

In addition, this isn't the first instance of Microsoft redistributing or contributing GPL code; they've done so for years, for their Services for Unix (SFU) product, which reused a non-trivial volume of GPL/GNU code.

Microsoft is taking this latest move (in the virtualisation market) because they need to compete more completely against market leaders like VMWare. If VMWare didn't exist, or if Microsoft was a dominant VM technology player, then this code release would never have happened.

It should be fairly clear that the more IT industry mindshare that Microsoft loses (or fails to gain) in various markets, the more ecumenical (in a platform sense) they will need to become, just to remain in the game; this is because most of their serious competitors are multi-platform, and inclusive of technologies like Linux etc.

Each such move therefore forms part of an accelerating vicious circle
for Microsoft, further shifting the centre of gravity away from their
technology stack. In the end, there's not much Microsoft can do to stop this.

All they can do is maximally slow the trend down, which is what they're aiming to do by supporting Linux "just enough" to remain competitive against multi-platform players like VMWare, VirtualBox, Xen etc.

You've probably caught up with the news that Microsoft is going to offer a version of Windows to the European market, without Internet Explorer pre-installed. This seems to be a ploy that isn't making anyone happy.

My interest here isn't in what Microsoft and the EU agree to as an appropriate remedy for Microsoft's market dominance in web browsers or past legal transgressions. My interest is in ensuring an increase in the global competition in operating system platforms. In short, unbundle IE in Europe? Why stop there? Why not unbundle the whole of Windows from all OEM PCs shipped in Europe?

This isn't a new argument; I, and others, have made this suggestion before. In short, the best thing that the EU regulators can do to increase competition in the computer platform market is to ensure that consumers are offered a choice of operating system for every computer they buy.

At the moment, pretty much every computer that a consumer buys from a brand-name vendor comes mandatorily pre-installed with one of Microsoft's operating systems. Furthermore, there is no viable way for that consumer to avoid paying a licence fee for that Microsoft software, even if they know they aren't going to use it.

It's now time for the EU regulators to step forward to stop this grossly unfair market-dominating mechanism which benefits Microsoft to the detriment of its platform competitors; players like Canonical (Ubuntu Linux), Novell (Suse Linux), Red Hat (Fedora Linux) and Sun (OpenSolaris).

There are a number of models for how this can be achieved, without creating needless impact on those consumers who do want to purchase and use Microsoft's operating system.

One approach is to ensure that consumers are offered, at the point of retail, the option to have their new PC 'imaged' and shipped with either a Windows OS or one of the competion platform, or with no operating system at all.

Another, possibly less impactful option, is for the EU to require that Microsoft ship zero-cost copies of OEM Windows for use on new PCs, which stop working after a certain number of reboots. Microsoft has already been embedding this kind of technology in Windows for years, as part of its product activation anti-piracy technology, so we know that the vendor is perfectly happy to include 'lock-out' and 'sunset' machinery in Windows.

At a desginated point in time, this free version of Windows will stop working as a functional desktop, and instead offer users an option of pulling out a credit-card to purchase a real Windows licence, to allow ongoing use, or offer pointers to download alternative operating systems, like the aforementioned Ubuntu, Federa etc.

Yes, I know there are arguments raised as to why such changes in industry behaviour are difficult or impossible, but these arguments are spurious.

The EU regulators have to know that forcing Microsoft agree to change its practices in order to increase competition in web browsers or media players will mean nothing in the long run, if Microsoft enjoys a near 100% monopoly on pre-installed operating systems on OEM PC hardware. The only thing which brings long-term relief to this ongoing market strangulation by any single vendor, is to make sure that that vendor's "PC bundling air-supply" is switched off.

So, EU regulators, are you up for this, the real battle to stop anti-competitive practices and open up the platform market to consumer-benefiting real competition?

I had a lengthy discussion with a new acquaintance the other day, and as it so often does in conversations I'm involved in, topics eventually shifted to the philosophy of knowledge and ideas, and their impact on the improvements within human societies over time.

My conversation partner knew something of these topics, but didn't know much about the computer industry. She was quite surprised when I introduced her to that concept of open source software, and how it was merely an extension of, if not the Scientific Method per-se, then at least the underpinning drive and methodology of Science.

This in tern got me to digging up an essay from many years back, which I thought may be of interest to this forum's members.

Enjoy. Feedback most welcome.

###

One of the overlooked advantages that Open Source development affords, is that it imitates perhaps the most fruitful and beneficial of all human endeavours: Science. How has the scientific-method evolved, and what can it teach us about the future possibilities of software construction?

Science, in its clearly understood modern guise, is unique. This essentially Western tradition of open inquiry is believed to have developed only one instantiation throughout the whole period of human history. While almost all human societies have developed language, art, and music, open inquiry into the natural and philosophical world sprung only from the eastern rim of the Mediterranean sea, in a number of ancient Greek states, approximately 27 centuries ago.

Helped along by the advantages provided by the recently formulated Greek alphabet, the people of this region bought forth the makings of the primary conceptual and philosophical machinery that was necessary to develop an understanding of Nature that surrounded them. In short order, they had conceived ideas which led them to believe that the Universe was understandable, that it was measurable and that it could undergo rational analysis.

Through the advantages bestowed upon the Greeks by virtue of their written language and undoubtedly their open, democratic political environment, these concepts spread. Where once these people would have been manipulators purely of the physical world (pottery, sculpture) they now also became masters if symbolic manipulation. Mathematics, logic, geometry, geography, mechanics, hydraulics, medicine, architecture, astronomy and cosmology, optics and dozens of other disciplines flowered. There has never been a similar period in human history, with the possible exception of the 18th century 'age of reason', the Enlightenment, interceded by the Dark Ages following the fall of Rome and the burning of the great library in Alexandria.

Luckily for all of us, these same ideas were storehoused and enhanced by the Islamic world. Eventually, after the fall of Muslim-held Toledo in the 11th century, these same ideas resurfaced slowly but steadily into Western and Southern Europe, to precipitate the Renaissance; the re-birth (of civilisation.) The arrival of these Greek texts coincided with the development of the university as a legal entity with political and intellectual autonomy. Once again, the openness to new forms of thought, the cheap, efficient and accurate transmission of ideas through the wonderful machinery of technology (Gutenberg's printing press,) brought forth an explosion of creativity and propelled Western civilisation forward. Open Source software is a direct descendant of this culture of open thought, as it prizes the same properties and philosophy which form the basis of the driving force of Science.

Openness is a key area of the scientific process which holds our interest here for comparison with Open Source development. Science is the process of verifying or culling hypothesis, and is in essence an open and self correcting system. Because of this, progress occurs at a much faster rate and in a more dependable/trusted fashion. This doesn't mean that the self correction happens in minute, continuously flowing 'chunks'. In reality, corrections arise as mini-revolutions, characterised by philosopher of science, Thomas Kuhn, as paradigm shifts. Nonetheless, over longer periods of time, progress does occur. In many ways, this progress is accidental, as there is often no 'vision' or nomenclature to describe where Science is heading, until after it has arrived.

The speed of progress is greatly enhanced by virtue of the fact the practitioners of Science publish not only results, but methodology, and techniques. In programmatic terms, this is equivalent to both the binary and the source code. This not only helps 'bootstrap' others into the field, to learn from the examples set, but makes it possible for others to verify or refute the results (or techniques) under investigation. In an almost guided-Darwinian evolutionary fashion, this makes the scientific process a powerful tool for the highlighting, analysis and possible culling of ideas and concepts; less useful ideas and hypothesis die, and likely contenders come sharply into focus. Newton made his famous comment about 'standing on the shoulders of giants', in part, to indicate that his contributions to human knowledge could not have been achieved solely. He needed the 'firmament' beneath him hypothesised, tested and confirmed by generations of scientists, philosophers and thinkers before him, over centuries. With Science, in the medium to long run, all other issues ('marketing' and previous acceptance) fall by the wayside, and merit alone is the main attribute of the victorious memes.

By analogy, making the source code available for peer review and extension, is perhaps Open Source's most powerful advantage as a development methodology. Besides the verification provided by peer review, refutation for issues such as security is possible, learning of techniques by new practitioners is a great advantage, as is modification and redistribution of the code under similar conditions. It makes it possible for anyone who has a background in, or can acclimatise to the technology and skills required, to continue development, extending the code into whatever direction that they need. This, in turn can be plowed back into the original system, causing eddies of strengthening feedback. This results in an ever growing base (or firmament) of quality code, upon which more and more programmers can benefit from, and contribute back to. The open communication of source and ideas, while progressing slowly at first, builds momentum as more and more practitioners learn from, extend or revive from obscurity, more and more code.

Another facet of similarity between Open Source and Science, is the respective cultures. Both are strongly technical, perhaps verging on the geeky, and both are meritocracies. A large part of the impetus of scientists is the applause of others. Peer recognition is also perhaps the single most important reason attributed by Open Source advocates and developers as to the reason why they pursue Open Source methods of software construction. Finally, much like the contemporaneous invention of the printing press helping propel Science, the Internet has provided both the human inter-communication and resultant software distribution vehicle, to help catalyse the Open Source process.

This chain of attributes and contingencies which made Science perhaps the most effective and pervasive of successful human endeavours, is finding resonant similarities in the burgeoning success of the Open Source process and may help explain its rise and rise.

A few people have recently sent me links to articles, such as this one, here on ZDNet UK, which report that Asus and Microsoft are joining forces to reduce the demand for Linux in the netbook market. These correspondents claim surprise that Asus, which was just recently a booster for Linux on netbooks, should have now switched sides, and be seen to be backing Microsoft.

I can't say I share this surprise. At no point did Asus switch sides.

To Asus, Linux was nothing more than a means to an end; namely, find some mechanism to make a lower-cost, small-form-factor portable PC possible. Microsoft wouldn't come to the party, either on morphing Windows to fit such a small-form-factor device (Microsoft was instead pushing Vista up that Sisyphean hill at the time), nor on price ("What, drop the price on Windows to make a cheap laptop possible? Are you nuts?").

So, Asus did something clever: they reached for Linux, which they could morph into a viable desktop environment, suitable for something like the original Eee PC, and which would also reduce their product bill-of-materials by a substantial amount, by avoiding the OEM Windows licence fee.

The Asus Eee PC was a rocketing success, perhaps the most successful new PC launched this decade. So much for desktop Linux not being viable for the masses.

If Asus had had their way, however, they would have shipped with a customised and super-cheap version of Windows. They calculated that if they could make the Eee product a success, Microsoft would come kowtowing to them, rather than the traditional power relationship of mid-tier vendors like Asus kowtowing to Microsoft.

So yes, in short, Asus used Linux as leverage cluebat, to ensure that Microsoft came to the pricing party with Windows.

And, as you can tell, it worked.

Microsoft dropped the OEM price for XP by some (to them) alarming percentage. Even more damagingly, the success of the Eee, necessitating the resuscitation of Windows XP to fill a hole that Vista could not, seriously wrong-footed Microsoft in the process - so much so that it was probably the biggest single reason for Vista losing market momentum and becoming the perceived dud that it is now.

So, here we are today.

Asus has guaranteed access to cheap Windows. It's essentially making 'netbooks' at inflated (ie, notebook) prices - none of its netbooks since the 701 model sold at a lower price. So, what's Asus market-position difference between now and 18 months ago, when it, like Acer and any number of mid-tier PC makers were Microsoft's biggest boosters, shipping nothing but PCs running Windows? Not much. So, you see, Asus never switched sides.

In reality, firms like Asus (and Acer, et al) were never, ever, pro-Linux. I'm surprised that people thought they were.

In general, the less brand-name power a PC vendor has. the more pro-Microsoft it is, as it can bank on Microsoft's brand. This is why it's much easier for IBM and HP to be pro-Linux than Lenovo and Asus.

The Asus gameplan has worked out quite well. The only thing that can derail it now are real, low cost netbooks. You know, like the dozens coming on the market later this year, all running Linux or Android on ARM CPUs...

Many years back, when I had the time to make the rounds at various government education departments, I asked the brass why on earth they didn't charge software vendors (like Microsoft, Adobe et al) for software the government's schools put in front of students?

What I meant by my question was this.

Rather than our public-sector education organisations paying for software, they should in fact not only get the software for free, they should actually be billing the software vendors for the right to place their wares and branding in front of millions of school students. Much like television and radio stations offer up a bidding system to advertisers who want to capture the eyes and ears of the 'audience', schools could operate a bidding process where vendors bid for accessing student's attention for their software.

I couldn't get one single cogent answer from my audience as to why this wasn't a viable idea. The idea would not only reduce the software costs to our schools by hundreds of millions, but perhaps, with the right kind of negotiation, wipe out the costs to schools for much of their IT needs, as the vendors' payments might just cover the costs involved for hardware acquisitions, networking and support.

Why aren't education departments worldwide pursuing this?

Why indeed.

You might ask: "Why would Microsoft et al pay to have their software placed in front of school-kids?"

And chances are, you'll already know the answer.

Imagine what would happen if a generation of school kids was raised, from an early age, to be Linux, OpenOffice.org and Gimp users?

What's that I hear you say? Kids can't use that software? Rubbish. I know of many experiments which had school classes split into two; one group learning something like the Gimp for image editing, the other Photoshop. At the end of training, the groups were swapped. And while yes, the Photoshop-trained kids complained that the Gimp wasn't intuitive, the Gimp-trained students said the same thing about Photoshop!

Now, imagine once more, if a whole generation of school kids was raised on Linux and open source software. Imagine what that would do to the current proprietary vendors who enjoy a strangle-hold in certain platform and application market segments. Imagine what that would do to their markets. Imagine again how much said vendors would pay to ensure that that never happens. And now you can begin to understand the incredible leverage that national school boards and departments of education have.

And yet, they do nothing with that power.

Why is it so?