Analysis It's been a good year so far for Microsoft UK - a couple of high profile open source insurrections in the public sector headed off, and an honorary knighthood for Bill, so who could ask for more? But the UK could be where Microsoft finally begins to taste defeat on the desktop, and where the Windows franchise starts to buckle.

On the surface, the omens are not particularly good either for open source or for discount-chasers. Microsoft held on to Newham Borough Council at the eleventh hour via some pretty serious price cuts, but Newham is less of a major precedent for Microsoft pricing levels than might initially appear, and despite regular menacing noises emanating from Whitehall, the government remains largely wedded to the notion that Windows is what's there to work with, that open source is for the future, and that efforts should therefore be concentrated on getting better prices from Microsoft.

Which is something Microsoft can live with. It probably has very little experience of customers demanding higher prices, but it has spent years dealing with ones who want lower ones, and has managed pretty well nevertheless. And the 'no alternative' card plays well against customers who are convinced that there really is no alternative; UK.gov might well be running open source trials, and might well have a policy that departments should consider open source alternatives alongside proprietary ones, but actually it's business with Microsoft as usual until such time as a few trials provide overwhelming evidence that it works, it's easy to implement, and it will save scads of money.

The difficulty for Microsoft is that a revolution (albeit not the revolution) might, probably will, start without central government's imprimatur, leaving the Office of Government Commerce's bean-counters to catch up and claim the credit. Currently, there are three directions it could come from - more ground-up open source projects of the kind engaged in by Newham contractor Netproject, or Sun or IBM.

Microsoft views all open source as a threat, and Netproject as a particular one in the UK. In the grand scheme of things, however, small, isolated open source projects have largely nuisance value for the company, and it's the ancestral enemies who're the greater danger. Small companies are easy to spread FUD about and isolate (Open source "software is usually not in a form that can readily be integrated with existing systems or compatible with off-the-shelf packages. Users have to therefore provide their own, often extensive technical support" - Controller and Auditor-General's report, as quoted in the recent Public Accounts Committee's software licensing report). Sun and IBM on the other hand are known suppliers with longer experience of dealing with government than Microsoft.

In the case of both companies it's at least arguable that open source is a tactic rather than a religion. IBM initially ran Linux as an OS for mainframe consolidation, and was sufficiently influential with the OGC for its advice against immediate deployment of open source on the desktop to be seized on as the good excuse for not doing anything. This is immortalised in the PAC's report as: "OGC had advice that open source software was not yet robust enough for use with the standard office desktop computer." IBM is now for OSS on the desktop, and may have retro-fitted desktop trials into the pilot studies it announced with the OGC last September. At the outset, however, desktop didn't look like a major part of the package, and IBM probably has a great deal of catching up to do in order to be a credible contender here. Nor, indeed, is it entirely clear the company wants to be a contender - parts of it might, but other parts of it sell Windows PCs, and Big Blue has a long record of contradictory and conflicting signals in this department.

Sun's compelling moonshine

Sun doesn't have this problem, and indeed would benefit massively from the complete destruction of the client PC market. In addition to being able to have everybody singing from the same hymn-sheet (aside from minor 'would sir like Linux or Solaris with that' problems), the hymn sheet itself boils down to an eminently marketable soundbite - £65 per station per year, for everything. According to Microsoft UK sources, Sun is actually pitching as low as £40, but the price isn't particularly material, because the point is that Sun is willing to go as low as it takes, as the desktop isn't where it makes its money. It wants to sell the idea of server-based computing, and it wants to stop Microsoft from leveraging into the server market via the desktop, but apart from that the current client market is just so much roadkill for Sun.

In a world where customers are no longer prepared to accept Total Cost of Ownership analyses from either side, the brutal, cynical simplicity of this 'per year, per station, for everything' soundbite is a killer. Sun UK director of public sector sales Charles Andrews is however reluctant to reduce it to this level; rightly, because the reality is much more complex, TCO is going to come into the picture just as soon as the customer starts looking at specific deployments, and of course Java Desktop System is simply the client end of what Sun really sells, and really makes its money from.

So although Sun is currently involved in desktop trials with the National Health Service, and these could potentially result in a massive defeat on the desktop for Microsoft, it's currently far more important for Sun to deliver on the NHS backbone contracts it has already won. This certainly has implications for the desktop, because the more IT strategy you buy from Sun, the more concerned you're likely to be with the delivery of apps to the desktop, and the less with deploying apps on the desktop. If Sun succeeds in selling the server end to government, then the PC-centric desktop slowly gets sucked out of the equation anyway. "It's actually about how applications get delivered to users," says Andrews.

Sun's Java Desktop System is therefore part delivery mechanism, part interim measure. It will run on existing PCs, meaning that IT managers accepting the Sun pitch can cut XP, MS Office and the hardware upgrade out of the equation, but it roadmaps logically into an appliance-based approach, because in this phase adopters will effectively be running their existing PCs as appliances, and the logic when upgrade time comes around is surely not to buy more PCs.

The MS counter-punch

If you look at it that way, Microsoft's objections that it's Sun and IBM who're the big boys and Microsoft that's the challenger acquire a certain credibility. Microsoft is now a pretty dominant force on the desktop, but it doesn't have a vast number of customers who've bought the whole show from client to server, whereas Sun and IBM still do pretty well with the server. And Sun's complaints to the European Commission that Microsoft is attempting to leverage server market share from the desktop are mirror-imaged by Sun's efforts to drive Microsoft and/or the profit margin out of the desktop. If Microsoft simply attempts to defend the desktop franchise then its profits will slowly be squeezed out, so it needs to win server business from the opposition and to sell its own equivalent of Sun's 'everything' strategy.

Microsoft has not been particularly good at this, so maybe it's losing already. But Newham is an example of its willingness to price an entire corporate system low enough to keep a customer and consolidate its position there. Microsoft does not view Newham as a precedent, but one of the reasons it's important is that it provides a snapshot of the situation a reasonable slice of the public sector currently finds itself in. The current IT regime in Newham is striving to rationalise and modernise a collection of systems that is outdated and not particularly rational. A number of Microsoft systems there themselves needed updating, and the cost of modernising the desktops was clearly a major factor. But you could argue that the desktops themselves were only important because they represented a big minus sign. Take away the Windows and Office licences and you'd make a massive saving, as Eddie Bleasdale of Netproject argued, but it's what you do at the server end that's actually important.

That's what Sun would have argued in Newham (it tried to intervene, but was too late), and it's what Microsoft has tacitly accepted in holding on to the council. It has kept the desktop clients there, certainly, but it has secured Newham as a Microsoft system focused customer.

Both Sun and Netproject argue that the deal will cost Newham in the longer term, with Bleasdale in particular making no bones about the perils of locking yourself in to Microsoft. But Microsoft was able to benefit from a certain amount of inertia here; it traded down to what the council viewed as an acceptable price, with the possible costs and likely terrors of the unknown factored in, and in exchange for this it got the lot. Which is not a bad deal, provided the consequent long term income is sufficient to offset the immediate discounts.

Inertia means that Microsoft does not have to cut right down to Sun's desktop level in order to hold the account, but the size of the cuts the company is prepared to offer clearly relates to what else it can sell and the length of the contract it can sign the customer up for. Sun (and IBM, as and when it shows up) is up against this inertia, but in some senses the desktop is roadkill for Microsoft too, and the real war will be fought out at the systems upgrade level.

Another striking similarity is that Sun, IBM and Microsoft are all prepared to spend money in order to win business. Microsoft gave Newham free consultancy, IBM also offers this to the public sector, and IBM and Sun are both funding OGC open source pilot studies. One can however reasonably doubt the OGC's status as any kind of driving force in this area - it's at least as likely that it's just accepting free stuff and issuing press releases claiming the credit. If a council or other public sector body goes to Sun and asks it to fund a pilot study then, Andrews concedes, Sun will almost certainly fund it, and this certainly happens already.

So you can see the war settling down into the large-scale clash of tanks funded by the big boys, with government itself acting largely as a confused and irrelevant spectator while congratulating itself on negotiating price cuts (allegedly) and free stuff. But there are a couple of problems with this, and Netproject's Bleasdale articulates them well.

The government policy gap

If the IT big boys are willing to fund pilot studies, then independent consultancies like Netproject cannot compete. Netproject has acted as a catalyst in several cases already, and is most usually seen as being in Microsoft's crosshairs - but free stuff from Sun and IBM is just as threatening to its business, and to the business of numerous lower-profile outfits who're already quietly doing business in the public sector.

The readiness of cash-strapped local government to accept free stuff is to an extent justifiable, but central government's espousal of it as a matter of philosophy (think public private partnership) is positively dangerous. Conforming to this philosophy, the OGC sees it as a positive benefit if it can get Sun or IBM to pay for pilot studies, without appearing to grasp that successful pilots will most likely result in sales for Sun and IBM. So you swap hegemonies, and Bleasdale is irate that the OGC appears not to see this, and sees nothing wrong with accepting the free pilots.

Obviously, he's right. The British government is effectively abdicating IT strategy to the major players and confining itself to attempts to play them off against one another in order to obtain lower prices. But it isn't considering overall strategy, and it isn't seriously opening up the competitive landscape. Nor can it do this without putting serious funding behind its own studies, ones that could conceivably be described as independent.

Bleasdale questions the nature of the open source offerings of Sun and IBM, pointing out that both companies, having paid for the open source publicity ticket on the door, will sell you a hell of a lot of proprietary stuff as well. You would of course expect that, because although open source does benefit when either company wins contracts, in both cases open source is a tactic, not a religion.

But we don't need to get into a religious argument here about whether open source should be totally embraced, or whether it can happily exist as an adjunct to proprietary systems, because the current government approach is clearly faulty. One of the things open source is about is opening up the competitive landscape. It makes it easier for small outfits to build systems using freely-available components, and thus allows them to compete more effectively against large organisations. But by working with large companies who position themselves as 'taking the risk out of open source', government is missing the point, and basically treating open source as just another proprietary alternative.

You could reasonably doubt that government could ever have the imagination to do anything else, but without some major change in the landscape, the future will be something like this. Microsoft will lose desktop share, but will be able to consolidate its holdings in some areas. It will face a substantial decline in revenue from the desktop franchise, and the franchise period itself will come to an end as its advantages of ownership of the turf are slowly eroded. The market percentage points will depend on how quickly and effectively the rival camps can get their acts together. Microsoft will probably look more damaged than it really is, particularly if it keeps making such a big deal of battles like Newham and Munich. Open source will benefit to the extent that it is deployed as part of Sun and IBM systems, but open source will not have won. It will not have been invited to the war, really. ®

Related links

MPs call for Gov.uk to switch to open source (maybe)
Public Accounts Committee report
Savage discounts from MS flush OSS desktop from London council
Open source not ready for desktop, IBM told UK government
UK gov aims for 'level playing field' with open source pilots