Microsoft and OSS: another battle brewing
Is the sun dawning into a new day of brotherhood, as Roberto thinks? Should we think that this time it is different, that no harsh words were spoken? That critics are wrong to suspect that something is brewing? I believe that the initiatives described by Roberto are just a new front of an ongoing market (and mindshare) battle, that Microsoft is playing to guarantee its position in the IT landscape of the future.
A Mad World, My Masters by Matt West
If there is one thing that should be visible to every analyst in the IT market, is that monopolies does not disappear in the night. As I already wrote in the past, the fact that every year is believed to be the “linux year” remains wishful thinking; and I still believe that even with the many new low-cost devices designed to run linux, the linux desktop market share in my simulations does not exceed 5% for the end of 2009 (of course, I hope to be wrong, and that in a bold sweep some new company is capable of selling 20M pcs in one year). On the other hand, open source is clearly capable of entering in both new markets or to be the underlying basis for more traditional products, like Apple OSX or the iPhone. I believe that the new activities from Microsoft are the first mature attacks against the OSS ecosystem, designed to de-emphasize both the ethical aspects behind OSS and the differences in licensing that provide the real differentiators from the technical point of view.
Let me share with you some initial musings:
Microsoft is a development tool company, and primarily sells to other developers. This may sound strange- after all, Microsoft sells operating systems and office suites that are not developer oriented. The reality is that Microsoft has created mostly platforms for other to build upon, and by providing nice and centrally-managed software libraries for every conceivable task it simplified the work for those building on Windows, Office, SQLserver and now SharePoint (among many other things). This simplification allowed ISVs to write software that run conceivably well, on a large number of machines, without having to juggle with updates from many different vendors of a separate DB, a separate web server, a separate presentation layer and so on. I believe that it is this ease of integration of components (because they were mostly from a single vendor, with rather similar and laissez-faire licensing conditions) and the fact that most of spending could be reused for different applications by buying licenses centrally from Microsoft once, and reusing them for additional value. In fact, I suspect that part of the lackluster performance of Vista was probably caused by the fact that, similarly to Windows ME, Vista had very little of value to offer to developers when confronted with the additional hardware requirements and the additional licensing cost.
For Microsoft (and its partners) everything is a PC. Remember when Microsoft designed its first game console? It was a PC, with just some changes in the bios and startup circuitry. Media centers? PCs. Servers? PCs. Mobile devices? PCs with a small screen, and a small “start” menu. The only “outsider” is the Zune, that is clearly designed as a clone of a product designed by others, and that as such is somehow neglected even by Microsoft itself.
And now, what happened? Many different things. First of all, the web (and virtualization) finally managed to deliver on the promises made years ago; even with some immaturities, a modern web engine can deliver end-user applications with security, speed and central management that provides significant cost reductions and much less hassles for both users and administrations. This combination allows for near-unlimited scaling (horizontally and vertically) and when used with open source software require no licensing steps that may increase the time to market, that is fast becoming the deciding element for IT deployments. Call it Prism, Air, Silverlight, JavaFX, there are enough choices that by leveraging existing and new platforms can give to software vendors new choices. And now there are enough options for developers to be free from the Microsoft endless supply of libraries, and they can now search for their own liking.
On the other hand, low-cost devices, handheld systems designed for the web and embedded systems on one side, and very large scale systems are so different from a PC that trying to shoehorn a PC model there simply fails, and in this way Microsoft has left opened several breaches that were ineffectively guarded (like stopping a flood with barbed wire). Now, mobile internet devices like the iPhone/iPod touch, nokia’s own N770/N800/N810 tablet (and the other WebKit-based N-series phones) and the up-and-coming intel MID are all examples of a new kind of platform that Microsoft is not prepared to fight for.
So, after trying to ignore OSS, badmouth it, or scaring companies into cross-platforms agreements, now Microsoft is taking a more mature approach, that uses its innate developer-oriented strength to swoon developers to develop and deploy on Windows and with windows-oriented tools, by dangling in front of software vendors the promise of a much larger market and the support of an extraordinary marketing force. By doing this, of course, it creates an incentive to leverage Microsoft technologies whenever possible, to “adapt” licenses (avoiding copyleft-based ones, that prevent deep linking with proprietary software) and thus facilitating a progressive embrace into additional Microsoft (or partner) technologies that can be centrally controlled. I suspect that there will also be a licensing change in future version of Enterprise/Grid versions of Windows, to counteract the economic and licensing advantage of OSS-based virtualization; this may however be difficult to manage well, as it may significantly lower extractable prices for large-scale installations. Pushing effort to reengineer their software offering in a modular way may help the company to move into smaller scale computing, as well as large scale system, and at the same time maintain the comfortable development and deployment environment that has made Microsoft such a large scale success.
What will happen? If Microsoft is consistent in its “good spirit”, they may be able to reduce significantly the platform threat and create strong bonds with at least half of the commercial OSS vendors within 2010. On the other hand, this can increase the penetration and perception of OSS in general, and if a suitable service provider appears on the market it can capitalize on that “visibility asset” and weaken Microsoft position from the inside.
If Microsoft (and at this point I mainly think about Steve “chairs” Ballmer) shows its “bad face” it may polarize the market further, creating a cadre of “white knights” that show no compromise and gain visibility and interest from the part of the OSS community that believe in ethical and openness values, thus reducing the value of accepting the Microsoft compromise.