Open Standards are up there with Jolt Cola as a "good thing" in computing: much better to use a carefully crafted specification lovingly honed by groups of world class professionals than a homespun solution. For software purchasers, open standards seem to be the ideal for which to aim. Open standards imply multiple solutions from which to choose companies competing for business. This competition feeds demand, which in turn increases the size of market, and thus enables still further competition. It would appear to be all good, until a closer examination of the politics and realities of open standards, which suggest a much less ideal motivation. Here are some recent examples of what I mean.
The Adobe PDF standard, ISO 32000, was a fast-tracked standard carefully crafted to offer advantages to Adobe. It is a classic example of control via incomplete specification.
The ISO PDF specification tells you how a PDF document is structured and it helps Adobe to broaden market appeal and demand. However, there are lots of associated technologies for PDF, which Adobe firmly controls and profits from products such as Acrobat — that are undocumented.
Some of the features within the PDF specification are specifically constructed to allow Acrobat to provide special features, but only for people who make their PDFs using Adobe software. Parts of the PDF specification, such as XFA and U3D, are not actually part of the ISO specification, yet implicitly they are part of PDF. Adobe regularly publishes updates and clarifications that are not vetted by any industry body.
For developers, the ISO PDF specification is the standard. However, the commercial reality of the world is that if Acrobat displays it, then it is a valid PDF. There is no open standard for Acrobat. To quote Duff Johnson of the ISO committee: "There is still no alternative to Adobe Reader when it comes to validating third-party software. As the vice-chair of ISO 32000, that bothers me, and if you're relying on the idea that PDF is indeed an international standard in your organization, it should bother you, too."
Restricting information and incomplete specifications are all ways to control access to the market.
Microsoft put forth Microsoft Office Open XML as an open standard. Open XML is also a fast-tracked ISO standard and the specification is terribly Microsoft oriented. It is an example of control via practically impossible specification.
In theory, the ISO Open XML specification provides an entry to competition with Microsoft Office. However, the standard is a behemoth seven thousand pages of committee-oriented text. It is just the kind of thing that large companies like Microsoft are good at handling. It is just the kind of thing that groups of inspired and dedicated programmers their competition find completely unworkable. To quote Georg Greve, president of the Free Software Foundation Europe, "Open XML is unnecessarily bloated, partly because it packs in unrelated features that lead users to other Microsoft applications…This is a classic vendor lock-in strategy."
Nevertheless, the specification was fast tracked through the ISO process to a chorus of complaint and unhappiness about vote buying, pressure by Microsoft, and conflicts of interest. By ensuring standards are highly complex and tied into your existing products, companies can restrict access to the market.
It is not only corporations that want control governments do, too. It is now known that the NSA campaigned for years to try and keep encryption levels low to allow them to monitor worldwide communications. We now also know that the NSA actively campaigned to undermine the effectiveness of encryption algorithms standardized by groups such as the National Institute of Standards and Technology (NIST).
As a result, NIST is now recommending that certain algorithms not be used and is undergoing a complete review of all its standards.
Participants on standards bodies are often required to license their intellectual property in a reasonable way. But what is reasonable to IBM may not be quite so reasonable to a small startup. The people who are best placed to deal in these rights are the developers of the standards themselves companies that can efficiently swap and cross-license their various technologies in a manner unavailable to everyone else.
Professor Joel West at the Keck Graduate Institute has studied the ways in which companies use open standards to produce competitive advantage. In cases like the development of 3G standards, major players within the industry forced the adoption of a patent policy for the standard that allowed them to dominate the industry. Indeed, many of these firms filed increasing numbers of relevant patents during the standardization process a strategic policy designed to maintain control and deter rivals.
Patents can even be surreptitiously hidden in standards. Companies such as Rambus have business models that are crucially dependent on licenses from patented technologies and have in the past been accused of lacing standards with patents for later ambush.
Obstruction of the Process
Even if you don't intend to license your patents you can still use them to your advantage. Back in 2011, Opera developer Haavard Moen accused Apple of bad faith: "Apple is at it again. Another year, another attempt by Apple to block open standards using patents."
Apple had lots of relevant experience and intellectual property regarding the W3C touch events specification. It would have been only natural for the company to participate in the group setting touch standards. But it declined to do so. Each time the standards body came close to release, Apple would disclose a patent that might be relevant, triggering an inevitable and sizable delay.
Had Apple participated in the group, it would have been required to disclose all relevant patents early and offer free licensing for the work covered in the specification. So, while Apple did not participate in and could not control the touch standard, the effect was that it was blocked repeatedly by the company's actions.
Both PDF and OpenXML are now standardized via the ISO. Is there perhaps a reason for this? Certainly, the fast-track procedures to accept existing de facto standards can be seen as little more than a rubber stamp on corporate standards; that is, ISO as a brand rather than anything substantial.
To be fair, the ISO does not refer to its standards as being open they are just standards. The ISO requires that members be prepared to license their intellectual property in a reasonable way, but it makes no specifications about what is reasonable.
Standards bodies like ISO churn out standards without too much reflection on whether they are useful or needed or practical or any good.
Ultimately bodies like the ISO provide an impression of openness rather than any guarantee it's more style than substance which is probably exactly what corporations like Adobe and Microsoft want.
Everyone wants to have open standards and no one has a monopoly on the term, so most people will call their standards "open" even if that only means that they are free to download.
The crucial aspect of standards to companies are that they have to make their standard closed enough that they maintain control of it, but open enough that it becomes adopted. These things are not black or white open or closed rather, shades of gray.
Do not expect open standards to be 100% wholesome goodness. Companies often abuse this process: They have agendas far beyond and far more closed than making their technology accessible. Be aware of this when evaluating how standards should be selected and used.
Jos Vernon is the CEO of WebSupergoo, which makes and sells image and PDF manipulation tools.