Microsoft and modular are two words I wouldn't normally associate with one another, so I thought maybe I'd misunderstood it. Heaven only knows, patent applications are generally written to confuse, not illuminate, and so I sent it to Dr. Stupid to ask if he'd please explain it to me. He did, and his explanation was so interesting, I asked if I could share it with you.

As best as I can understand it, it's not an attempted patent on a modular system per se. That obviously wouldn't fly. As he points out, it's not new. The patent relates to a method of delivery of an operating system where you start off with a very basic operating system, a kind of crippled starter edition, and then you pick and choose (and purchase) additional functionality, with DRM used to make sure you don't self-help. It's like modular copyleft, turning the advantages of GNU/Linux -- modularity there increases what you can do and what you can add and how well everything works -- and instead turns the concept on its head by using modularity plus DRM to restrict and contain and enforce.

The result seems a bit scary to me, and I ask myself, why would anyone agree to something like this? Perhaps I can chalk it up to Microsoft innovating again. They must be testing the outer limits of what a customer will put up with before bolting to Linux, certainly a valuable scientific study from my point of view.

So, using our imagination, let's extrapolate -- how could such a business model, whether the patent is granted or not, be used in an imagined Microsoft Brave New Modular World? Of course, Microsoft doesn't need a patent to use this business model, nor does a patent application prove it will use the model. However, as attorney Michael Geist phrases it, Microsoft's Vista already "seemingly wrestles control of the 'user experience' from the user." And this patent application brings to mind a model that takes it a further step. Another issue: if Microsoft doesn't need a patent to pursue this model, then why apply for a patent? To corner the market in restrictive, controlling operating systems? Well, speaking for myself, Microsoft can have that corner of the market to itself, by all means. When would having a patent on such a system be an advantage? I'll leave all such possibilities to your imagination.

But here's how Dr. Stupid understands the patent and how it could play out as a business model:

******************************

Introduction

The patent is not interesting for its technical content -- all the building blocks of the described system have been used for some time now -- but for the glimpse it offers into the business model envisaged by the applicant. Therefore, I shall not delve into whether the patent should be granted at all in this discussion.

The application relates to a method of delivery of operating systems where one initially obtains a very basic operating system (for example, one that can do little more than display a web browser restricted to one site or domain) and thereafter is presented with the option to select (and purchase) additional functionality.

Some of the examples given echo aspects of the cut-down "Starter Editions" of XP and Vista:

10. The operating system of claim 1, wherein the at least one add-on module corresponds to a number of concurrent windows.

11. The operating system of claim 1, wherein the at least one add-on module corresponds to a number of concurrent applications active at a time.

Now, having a simple starter system to which additional functionality is added and upgraded via software modules is not new: this is the traditional installation approach used by Debian, and the idea goes back further than Linux distributions themselves (one could argue that it is part of the central UNIX philosophy, for example.) What distinguishes the proposed system's approach is the following:

1) Retrograde steps in base functionality.

Some of what is described as "additional functionality" is actually providing functionality which ought to be (and previously has been) standard in the operating system. For example, any modern general-purpose operating system, even a very basic one, should support running as many simultaneous applications as the user wishes, limited only by the constraints of the platform hardware. To impose a limit of, say, 4 simultaneous applications is an arbitrary restriction. More pointedly, preparation of the artificially-limited version requires further engineering effort; there is no "sweat of the brow" argument I can see to differentially price the limited and unlimited versions (in contrast, a multi-processor capable OS kernel does represent additional engineering effort compared with a uniprocessor kernel, and as such could be justifiably offered as a paid-for upgrade.)

This use of an explicitly defeatured system as, one presumes, a loss leader item which the producer hopes the user will pay to upgrade will be familiar to readers as "demoware" (or more unkindly "crippleware"), but to my knowledge this is the first time such an approach is being extended to encompass an operating system as opposed to a few low-cost handy utilities.

The patent application amusingly tries to dress up the removal or impairment of functionality which currently all users enjoy as standard as a benefit:

[0029] The peripheral category may include .... the number of peripherals allowed. A computer 110 functioning with the basic kernel operating system 202 may have limited peripheral support. Add-on modules may allow users to select the types, speed, and number of peripherals allowed and supported.

The purchaser of a new computer might be entitled to expect to be able to plug in all his or her peripherals from the old computer.

[0030] The communication category may include network interfaces, ... Add-on modules for communication may be used to support ... for example, DSL speeds up 500 kilobits per second, or speeds up to the maximum supported by the available hardware. Communication may also be limited by type, such as support for Internet browsing separate from peer-to-peer networking.

So even if your modem, line, and ISP support 8Mbit/s, your internet browsing may be deliberately throttled by the OS unless you obtain an additional "module".

[0031] .... a power user may want specific window and background themes with associated styles, a significant number of concurrent windows, and an unlimited number of concurrent applications, up to the capacity of the hardware.

I would tentatively offer the suggestion that *any* user would want the OS to use the full capacity of the hardware at all times, even when performing simple tasks.

... Users of business support applications may benefit from a support pack including high speed disk access...

The placement of high speed access in a module implies that the standard offering might deliberately throttle back disk performance to create an artificial reason to upgrade.

2) DRM to control system expansion

DRM, in various forms, is used (a) to ensure the new functionality has been legitimately obtained (i.e. paid for) and (b) to only allow "approved" additional modules.

3. The operating system of claim 1, wherein the at least one of the plurality of add-on modules further comprises a certification, wherein only add-on modules with the certification are activated by the operating system....

12. The operating system of claim 1, wherein the at least one add-on module comprises digital rights management corresponding to at least one of a number of uses of the add-on module, an expiration date of the add-on module, or a cumulative uses of the add-on module.

Of these two aspects, (a) can be defended as an anti-piracy measure. (b), on the other hand, serves the main purpose of obstructing third parties from filling in the functionality holes in the basic offering themselves, or (to borrow a popular phrase) removing the computer's owner's "freedom to tinker".

There is a clear signal that the ability to install even non-approved applications (let alone OS components) is controlled:

5. The operating system of claim 1, wherein the at least one add-on module enables installation of a non-certified application program.

Since the ability to install non-certified programs is placed in an "add-on module", the natural question to ask is: will the computer's owner have to pay a surcharge to install the software he or she wants on the hardware he or she owns?

3) DRM to selectively enable and disable functionality

[0039] ....The add-on module 300 may be examined for digital rights indicia .... When digital rights indicia are present and the criteria required are satisfied, the yes branch ...may be followed and the add-on module started .... Should the digital rights criteria fail... for example, if an expiration date has passed.... a notification may be presented to the user indicating the add-on module was not started. The notification ... may further include an opportunity to correct the failure, for example, by purchasing an additional usage allocation.

While not spelled out in the application, the fact is that the module's DRM attributes are checked periodically and on every system start allow remote termination of access rights for reasons other than expiry of a subscription. Thus, the entity administrating the digital rights framework (the operating system publisher or an appointed 3rd party) will be able to "punish" the user for actual or suspected transgressions of a EULA by disabling the relevant functionality.

Modularity

All operating system vendors are moving toward a more incremental model of OS delivery. Recent IT press coverage of the Vista release has had as one of its themes the widely-held opinion that Vista will be the last OS of its kind (a monolithically-delivered package) from Microsoft. Here's but one example.

Speaking generally, there are undoubted benefits to both user and OS provider in modularity. From the users perspective, modularity allows him or her to mix and match components, choosing from several providers for each coarse functionality area. Unnecessary components can be disabled, reducing system resource consumption and improving system security and reliability. As previously mentioned, this is not a new technology. From the OS provider's viewpoint, modularity allows the OS as a whole to be sold on a more flexible basis and arguably improves overall OS quality by reducing unnecessary coupling between components.

There are areas, however, where what appears as a benefit to the end user is not a benefit to an OS provider, or at least one trying to maintain a monopoly market position. By its very nature, modularity eases replacement of system components. Other things being equal, one would expect the market to naturally foster the appearance of 3rd party replacements for most if not all the OS components. The operation of the free market would also prevent artificial functionality restrictions (ones where straightforward functionality is being turned off, rather than new functionality being offered through genuine innovation) from being viable as a business model. For example, if the basic window management component of an OS only allowed for 8 top-level windows, then a 3rd-party replacement component would soon appear which removed this restriction.

In the monolithic model of OS delivery, a provider with dominant market position can use the practice of "bundling" to undermine competitors. If the basic installation of an OS has component X pre-installed to deliver functionality Y, this component enjoys an incumbency effect and thus reduces the likelihood of competing components W & Z being investigated, let alone obtained and installed. Switching to a model whereby component X has to be actively chosen (and possibly purchased) by the user would, on the face of it, improve the opportunities for W or Z to be chosen instead. Modularity - real modularity - militates against bundling.

In this context, the ability of the core system to only allow approved OS components to be installed appears to be a mechanism whereby the OS provider can retain the benefits of modularity from its perspective, while denying the end user his or her corresponding benefits. At the very least, 3rd party replacement components could expect to have to pay a de facto royalty to obtain and maintain their "approved" status. The impact on FOSS-licensed replacement components is obvious.

A possible scenario

So let us try to envisage what the total end-user experience is likely to be in such a system. Some years in the future, the owner of a new computer turns it on for the first time and is welcomed by - let's call it "Blister Basic Edition." This OS has almost no functionality. It can run a "get connected to the internet" wizard and then displays an online store in a simplified web browser (the browser does not allow any other site to be visited.)

Although I refer to an online store, the first transaction does involve more user expense. He or she redeems a unique coupon code (received as part of the purchase) which allows him or her to install a number of additional modules. Compromises may have to be made: perhaps the user does not have enough "credit" to have both "enhanced" (i.e. normal) disk performance and connect two printers...

The DRM and certification features ensure that if the user wants additional or improved OS functionality (including security features) these can only be obtained from the OS vendor or a vendor-approved 3rd party.

Subsequently, the Blister Update icon appears at regular intervals to offer the user new, improved functionality. The OS updates are more frequent and less dramatic -- in this respect Blister behaves more like Linux or indeed (to an extent) Mac OSX. These updates are unlikely to be free. Doubtless one will be able to purchase them on an as-is basis, but the pricing structure will be such that it seems much more compelling to buy, say, 3-year subscriptions. One simple regular payment and the user can get all the upgrades he or she wants! However, the total payments spread over 3 years will probably be more than the original monolithic OS price -- otherwise the OS vendor would be making less money. Moreover, the user could now be required to pay for bugfixes, which traditionally have been free upgrades. The application explicitly proposes this and, with no little audacity, claims it as an end user benefit:

[0028] ....In the area of operating system maintenance, service packs, bug fixes, and patches have been included in the original purchase price of an operating system for prior art systems. The use of an add-on module for support of bug fixes, service packs, and patches allows users to selectively pay for only the support that is of interest to them.

Requiring users to pay for security fixes "of interest to them" hardly seems the route to greater overall security on the internet.

Rolling upgrades and upgrade treadmills

If Blister is kept upgraded on a rolling basis like this, then in one respect the user does benefit - no more "big bang" upgrades (rather like Debian, you only install it once.)

From the OS provider's view, the rolling model, as well as yielding a steady revenue stream, has two market position benefits:

i) It gets rid of the "jumping off points". Upgrading from one major version of an OS to another (e.g XP to Vista) is rarely a straightforward experience. It is a situation where a vendor can lose customers -- either because they decide to change platform instead of upgrading, or the upgrade trashes their system and they change loyalties as a result. By removing the "big bang" of a platform upgrade, there is no longer a "natural point" for the user to try any other OS.

ii) If a user pays for 3 year's upgrades in advance, the sunk cost fallacy will deter him or her from switching OS providers during that period. Of course, as the 3 years approaches the end, various tempting special renewal offers will be presented to him or her by Blister Update. There is an obvious analogy with mobile phone contracts here.

iii) Because additional modules have to be certified, the user cannot remove modules whose functionality is purely for the benefit of 3rd parties (e.g. targeted advertising, prevention of fair use of legitimately purchased content) with whom the OS vendor has a business relationship or agreement. The ability -- or threat -- of remote disablement of functionality may also be relevant here.

Conclusion

All in all, it fits with an overall goal of turning your home PC from an empowering tool into a "don't tamper" appliance - indeed, into a device that you effectively lease (rather than own) that is useful for little more than delivering approved partner content and taking part in those harmless online activities that don't endanger the OS vendor's revenue (or those of its approved partners.):

[0040] ... Lastly, service providers or system operators who may be providing computers on a pay-as-you-go or pay-per-use basis may be able to limit installation of potentially harmful applications [DrS: harmful to whom?] or hardware peripherals by restricting the installation of required operating system add-on modules.

A model where the level of functionality of the OS can be purchased on so fine-grained a basis has many other interesting (and disquieting) possibilities, which cannot all be covered in this space. Since all functionality enhancements have to pass ultimately through the gatekeeper of the base OS vendor, this vendor gains a power of patronage. For example, a blogger who passed favourable comment on one of the OS vendor's products could find himself rewarded with a free upgrade of a minor system component. The scope for mischief by a unscrupulous OS vendor is also vast -- not that I make any particular allegation here.