ACCU Home page ACCU Conference Page
Search Contact us ACCU at Flickr ACCU at GitHib ACCU at Facebook ACCU at Linked-in ACCU at Twitter Skip Navigation

pinConsensus

Overload Journal #80 - Aug 2007 + Journal Editorial   Author: Alan Griffiths
It should be obvious that the process of agreeing effective standards must be dependant on attaining consensus.

It should be obvious that the process of agreeing effective standards must be dependant on attaining consensus.Over recent editorials I've commented on the increasingly concerning activities occurring in the ISO PAS standardisation process. One of the cases I've been following (the C++/CLI standard) came to an early resolution as a result of the efforts of Herb Sutter (Microsoft's liaison to the C++ community, and a major contributor to the ECMA C++/CLI work). Despite his investment in C++/CLI after seeing the reaction of the C++ community he managed to steer a path through the rules and regulations of ISO and stop the progress of this purportive standard at the ballot resolution stage. Initially JTC1 [ISO's Joint Technical Committee 1] refused ECMA's attempt to withdraw C++/CLI, but after lining up support for this withdrawal from a number of national bodies a second attempt to cancel the Ballot Resolution Meeting was allowed.

I'm sure that as developers many of us have experience of projects that only progress in something resembling an orderly manner as a result of the heroic efforts of individuals. Indeed, it is part of the description of CMM Level 1. Despite it being common this really is no way to run a project. (Sooner or later any organisation will run out of heroes.) It is also no way to run a standardisation process!

In Overload 75 I wrote the following:

Every organisation develops a culture over the course of time that reflects the way it tries to work. And ISO is no exception to this. Most of its standards are of interest to a minority of those on the decision making panels (not surprising really, there are a lot of standards and very limited resources to pursue them). A consequence of this is that national bodies with no interest in a particular standard try to keep out of the way by automatically voting 'approve' to anything that comes up for a vote. For traditional standards this is justified on both the tit-for-tat principle that others will do the same for standards that do interest them and on the assumption that those national bodies forming the working group have been diligent.

In organisational terms the fast-track is a new thing and the existing culture of 'approve' by default is still operating. However, for a PAS submission there may be no national bodies interested in the standard - with the consequence that neither the tit-for-tat principle nor the presumption of diligence need apply. The effect of this can be quite alarming.

Given these concerns I was very pleased when a paper written by one of the national standards bodies was brought to my attention:

South Africa is concerned about what seems to be a growing number of standards submitted under the PAS process that, although publically [sic] available, do not seem to have any measure of regional or even national consensus. These therefore tend not to have been referred to any of the JTC 1 sub-committees, and have obviously not been discussed at [sub-committee] level.

Our experience is that the result of this is then a round of intense lobbying by various other stakeholders for us to vote negatively on the PAS. Often these other groups take the trouble to compile a list of contradictions that are also widely distributed in order to justify the request for the negative vote.

A recent example is the proposed PAS on Open XML/ODF.

It is our opinion that the submission of such 'standards' directly to JTC 1 via the PAS route, where the standard has not been discussed within the relevant SC, was never the intention of the PAS System. The fact that some consortium has published a document that they refer to as a standard does not automatically imply that it has any sort of widespread industry acceptance. The fact that the publisher might claim international usage or acceptance is not longer a valid reason in these days of large multinationals, and the SABS [South African Bureau of Standards] has previously been approached by local branches of multinationals to vote in support of such PAS submissions, even if we have no local industry involvement or membership in the appropriate JTC 1 SC.

As result of this, South Africa will tend to vote negatively on all future PAS submissions to JTC 1 where they have not been appropriate SC. We would like to ensure that proper consideration be given to the PAS by technical experts. If the standard is indeed well known within the industry then this process might be very short.

This will be a change from our previous tendency to 'abstain' where we had no direct knowledge of the submission.

I may be reading more into the timing than is strictly warranted, but a further standards body (this time DIN - representing Germany) has been comparing the various standardisation processes that occur under the ISO auspices. It has the following to say about the fast-track process:

In case of the Fast-Track process, the proposer does not need to prepare any of the required information for a NP (certainly he is free to do so to support the proposal). National Bodies have no influence over the acceptance or rejection of a Fast-Track proposal at this stage as there is no formal 'acceptance' ballot foreseen. Only when 'perceived contradictions' are identified can they raise concerns. NBs/SCs have to 'accept' the additional workload brought in by a Fast-Track proposal, which may also have an unforeseen impact on their prioritised work plan (Art 6.3.1.4). In addition there is no check if there are enough P-Members committed to work on the proposal and to thoroughly review the proposed specification.

Even if NBs can clearly demonstrate that the submitted specification duplicates an existing standard, the Fast-Track proposer can still request to enter the technical ballot phase. During the technical ballot phase, a formal rejection is only allowed based on technical arguments (see Annex G18: "If a national body finds the DIS unacceptable, it shall vote negatively and state the technical reasons." and "We disapprove for the technical reasons stated"). JTC1 and its NBs are therefore not able to reject standard proposals based on arguments like business needs, stability, maturity of the technology, lack of independent and interoperable implementations, etc.

The other purportive standard I've been reporting on is that of OOXML - an ECMA document format standard whose claimed scope is supporting documents produced by MS Office (there are contentions that this standard fails to address this scope adequately). There are now accounts [Groklaw] appearing on the internet of national body committees being 'packed' with delegates with the intention of forcing a 'yes' vote on approving this standard. Vis:

We've seen now reports from Italy and Portugal of what some are describing as a kind of ballot-stuffing on the part of Microsoft and supporters to get Ecma-376 approved as an ISO standard. Trust me when I tell you that you haven't heard the half of it yet. I feel safe in saying that you will never hear the phrase 'fast tracking' again, without remembering what you are about to read.

It has never been a surprise that powerful organisations exert their power in their own interests. But many may be surprised that there is sufficient value in standardising a document format for these forces to come into play. There are two reasons - there are some (usually government related) markets that will insist on conformance to standards (because standards promote competition and other good stuff) and Microsoft doesn't want to lose these markets, and also because competitors implementing the alternative international standard (ODF) can use the interoperability this enables to facilitate their own marketing.

Somewhere in the midst of all this manoeuvring the ideal of gaining consensus has been replaced with that of competition. Consensus could serve everyone - competition serves only one winner.

Reference

[Groklaw] http://www.groklaw.net/article.php?story=2007071812280798

Overload Journal #80 - Aug 2007 + Journal Editorial