A Question of (Open) Standards
Published 14:23, 06 September 12
As long-standing readers will know, alongside ACTA, the other main theme of this blog over the last year or so has been the battle for the soul of open standards, which culminated in the UK government's consultation on the subject. We don't yet know what the outcome there will be, but whatever it is, the issue of open standards will only increase in importance.
That makes this independent study on "the costs and benefits of introducing an open standards policy for software interoperability, data and document formats in government IT requirements", commissioned a few months ago by the Cabinet Office, extremely useful.
As far as I know, it's the first rigorous, in-depth look at this whole area, and it makes for fascinating reading (well, it does if you are one of those sad people like me that finds this subject interesting.). Most of it will be familiar enough around here, but it adds plenty of links to original papers that provide further food for thought. And along the way, it brings out a few specific points that are worth emphasising - like this one, contrasting the impact of copyright and patents on software interfaces, which lie at the heart of standards in the computer world:
The law has recognised that software interfaces should be exempted from copyright protection. The law on software patents has developed separately and there is no exemption for interfaces. There is evidence that the concept and implementation of software patents is flawed, does not incentivise innovation and could restrict the operation of standards and interoperability. This raises the question of whether there is any economic justification to encourage patent protection of interfaces and whether there is a failure of the market which has not been corrected by the adoption of SSOs of FRAND policies.
That's an early indication of the report's healthy scepticism about the value of open standards that allow FRAND licensing - something that is explored at greater length later in the report.
Then there is this:
Setting standards creates an exclusive market position which can easily be abused. The co-operation between competitors and the setting of detailed specifications can exclude competing technologies. For this reason the behaviour of firms participating in standard setting can infringe Art 101 and 102, competition law provisions of the TFEU [Treaty on the Functioning of the European Union].
That's an important point: that the very act of working together on a standard runs the risk of encouraging anti-competitive behaviour. It's a reminder that standards aren't a magic wand that can be waved over real-life situations to make them instantly better; instead, they need to be chosen with care or they can be worse than having no official standards at all.
I also rather liked this characterisation of the current procurement model:
The "bargain then ripoff" model associated with vendor lock-in is avoided as open standards avoid lock-in to a particular supplier. This could result in a decrease in the discount offered on the initial contract as the supplier would not be able to charge a premium when supplying further software or services. This would increase the initial cost of switching to new software but, as future procurement will remain competitive, there should not be an increase in the total cost of ownership.
Again, that makes an important point: that in the wake of open standards, prices may rise, apparently negating the benefit of such standards. But in the long term, the overall costs should fall as competition on a truly level playing-field kicks in.
The report also includes this useful summary of the famous Munich migration to open source:
The City of Munich is said to have achieved cost savings of over €4m in one year by switching from a proprietary supplier to open source Linux. A third of the IT department's budget was saved by switching to Linux and OpenOffice. It is reported that not only did they avoid the cost of purchasing new proprietary software and upgrading systems, at a saving €15m but they also avoided another €2.8m for licence renewal. Support calls to help desks also fell from 70 a month to 46. In addition the City continued to use its existing hardware as Linux did not "stress" the system. Extending the life of hardware not only avoids cost, but can also have environmental benefits by avoiding waste.
Here, too, there's a neat little insight: another way you save money by moving to GNU/Linux is that your hardware can be used longer because the requirements are lower. Note, too, the interesting fact that GNU/Linux-based systems actually require fewer support calls than proprietary alternatives - contrary to the usual the FUD.
And here's the report's final summary of the situation regarding FRAND and RF licensing for open standards:
Adopting FRAND can be justified on the ground that it gives the most choice, but there are considerable risks as the owners of the IPR will have rights (magnified by being part of a standard) which could conflict with the users' interests. The use of RF standards where available and commercially viable has advantages and should be encouraged. Policies that might be adopted include preferring RF standards and/or open source software solutions, or a more general policy of encouraging open standards and/or open source suppliers in public procurement through dissemination of information and pilot projects.
This is still a very live issue, as the recent announcement of something called the OpenStand principles shows:
Five leading global organizations—IEEE, Internet Architecture Board (IAB), Internet Engineering Task Force (IETF), Internet Society and World Wide Web Consortium (W3C)—today announced that they have signed a statement affirming the importance of a jointly developed set of principles establishing a modern paradigm for global, open standards. The shared “OpenStand” principles—based on the effective and efficient standardization processes that have made the Internet and Web the premiere platforms for innovation and borderless commerce—are proven in their ability to foster competition and cooperation, support innovation and interoperability and drive market success.
That all sounds unexceptionable to the point of blandness. But the devil is in the details, and so it proves here.
Visiting the Principles page, we find under "Availability" the following:
Standards specifications are made accessible to all for implementation and deployment. Affirming standards organizations have defined procedures to develop specifications that can be implemented under fair terms. Given market diversity, fair terms may vary from royalty-free to fair, reasonable, and non-discriminatory terms (FRAND).
So there we have it: FRAND welcome here. That in itself might not be of note, were it not for the fact that the W3C has been one of the staunchest champions of RF-only licensing for open standards; its presence here is then a hugely retrogressive step - I wonder what Sir Tim Berners-Lee thinks about it?
As ever, the best commentary on what this all means can be found on Andy Updegrove's Standards Blog. Here's what he has to say on the choice of FRAND:
The background for this position is not difficult to guess: the IEEE has over 500 active working groups, most of which are in technical and industrial areas that are accepting of standards that require payment of license fees. It would not have been likely that IEEE (or, for that matter, the other SSOs) would have agreed to a statement limited to royalty free standards, and I have been able to confirm that this was, in fact, the case.
Some will say that given such a split, it would have been better to say nothing at all than to include the statement made with respect to IPR. Others will say that this is a battle to be fought elsewhere, and that the W3C can be excused for not being intransigent on this point.
The larger issue is why bother with this very vague statement about open standards that aren't even really open? Again, Updegrove has some good background analysis:
Many governments have traditionally shown a preference for, or indeed restricted themselves entirely to, including standards in procurement decisions that were created through the traditional, global standards infrastructure. For example, until recently European governments have engaged in a sort of NeverNever Land exercise of refusing to alter their procurement rules even as they make wholesale use (like everyone else) of technology that is rife with standards developed by consortia. Indeed, most of the Internet and Web is enabled by standards developed through these organizations.
Europe is currently reevaluating this position... Moreover, the International Telecommunication Union (ITU) has for many years tried to insert itself into control of standards enabling the Internet and Web, and those efforts are continuing. Finally, increased attention is being paid to the open standards criteria laid out under the World Trade Organization’s Agreement on Technical Barriers to Trade. Under that Act, signatory countries are barred from using “home grown” standards to bar, or impede, entry of products conforming to globally adopted standards. Proponents of standards developed by traditional standards organizations frequently assert that consortium-developed standards should not be regarded as meeting this threshold.
I’ve confirmed that one motivation behind the release of the Principles at this time is to publicize the view that “non-traditional” SSOs such as the founders are not only capable of creating world-class standards that can result in vast benefits to humanity, but that the processes that they employ are the equal of, and perhaps superior to, those of their more traditional peer organizations.
So, basically, there's some high-level politicking going on here between national governments, traditional standards bodies, newer consortia like the W3C, the International Telecommunication Union and the World Trade Organisation. Against that portentous background, the aims of the OpenStand are laudable, but my fear is that the explicit mention of FRAND in the principles will be seized upon by the increasingly desperate supporters of this anti-open source licensing approach. We shall see.