Standards Best Practices Blog

I’ve been away from the blog for a while, for the holidays and while working on some other projects, one of which is the revision of a paper for publication. While working on the paper I’ve been ruminating on the actual value of accreditation or recognition of the standards organization as it relates to the standard itself. By this I mean whether the organization developing and/or approving the standard is recognized or accredited in some way, usually by one of the international organizations ISO, IEC, or ITU, or by one of the national bodies. These organizations gain their authority either directly from government or through government recognition of their work. They in turn will recognize or accredit other industry organizations or their work, and will further approve work coming out of these accredited organizations. This accreditation is generally based upon a recognition of the quality of the work coming out of the organization or of the process used to create the work, and generally results in increased adoption/implementation of the work as the accreditation is a measure of quality among potential implementors.

But is this always the case? And can an unaccredited organization produce work of similar or even better quality? Probably not, and most certainly yes.

In general, work coming from an accredited organization can be counted on to be of high quality. This I define as, among other things, being a good technical solution for a particular problem or one that meets a specific need, that is implementable, is interoperable between various implementations, free of bugs, and has been developed under an open, consensus-based process where all interested parties have had the opportunity to participate.

But can only accredited organizations develop standards that meet this measure of quality? Absolutely not. As I’ve observed before, there are several unaccredited standards organizations, sometimes referred to as consortia or SSOs, that would qualify for accreditation if only they saw the benefit of applying for this status. And on the other hand, there is work coming out of accredited organizations that is never adopted — a good sign of poor quality.

The example of the day is, of course, the OOXML specification which was developed internally by Microsoft’s engineers then submitted to Ecma, where it was quickly approved without review or further work, then submitted by Ecma, using their qualifications as a fast-track submitter, to ISO/IEC JTC1 SC34 for approval as an ISO/IEC international standard. About 3000 comments where submitted during the short comment period, and after the first SC34 approval ballot failed a Ballot Resolution Meeting was scheduled for next month to see if the national body delegates can resolve the comments and find a way to approve the specification. During the voting there were allegations of vote stacking and disenfranchisement within various national bodies, and during the review process there was no opportunity for the public to review and comment on the specification, and discussions were open only to members of the subcommittee with no public examination of the deliberations. This is a worst case of how things can happen under an accredited, recognized standards development and approval process.

The equivalent standard coming from an unaccredited body is ODF, which was initially developed by Sun Microsystems then submitted to OASIS where it was further developed by a technical committee for about three years before approved as an OASIS Standard. During this time multiple public comment periods were held, and all deliberations of the technical committee were open to public viewing. ODF was then submitted by OASIS to SC34, the same body currently working on OOXML, for approval as an international standard, and the vote passed without much if any opposition. This is the best case of how things can happen at a consortium.

So what’s my point? I’m not sure, but perhaps it’s something along the lines of accreditation not guaranteeing quality, and that work coming from unaccredited organizations is no worse than that coming from accredited organizations. Note, though, that OASIS did recognize the value of international recognition when it submitted ODF for ISO/IEC approval.

So where does that leave us? What is the value of accreditation by a recognized standards body? And when it comes down to it, what does “standard” really mean? Nobody owns the word, so anyone is free to call anything by that name. A standard is really only a standard if people recognize it as such. Certain groups of people, known as organizations, have developed rules for how they will use the word, and they have applied the word to specifications they deem worthy according to their own defintion. It is up to other people, the potential implementors and users of the standard, to accept the organization’s definition of the word, and decide for themselves whether work developed and approved by that organization is of the quality that they require to solve their problems. If they don’t like the quality of the standard, there’s probably another one to choose from. So it’s up to the standards organizations to make sure that they have the best development and approval processes that they can build, and that they administer those processes in a fair, competent, and professional manner. It’s their reputations that will ultimately be the deciding factor in whether their standards get adopted.