An article in ZDNet remarks on what the common citizen should really worry about in the wake of the ISO OOXML voting controversy.

Whether or not OOXML is a good candidate for an open standard is beside the point: there is prima facie evidence of voting in bad faith, without proper consideration of all the aspects of the proposal. This is not something that can be fixed later; there are severe implications for the industry in adopting a standard that has not been fully analysed. Those who vote without understanding what they vote for, or because they have primarily political or commercial reasons, are guilty of subverting the process.

I suspect everyone understands precisely what they are voting for. Atleast some of the reasons have nothing to do with the merits (or demerits) of the proposed specification, and there's the rub.

The questions of how a national committee, deciding on a vote in any standardization process, is constituted and who participates in the deliberations should be considered by the ISO. For ISO standards to continue to be treated with universal acceptance, it should take cognizance of what goes on in the standardization process outside its halls.

The Hungarian Standards Institution has been asked by a minister to reconsider its approval. Sweden has decided to abstain from the OOXML vote, reversing the earlier approval due to a technicality. I came across a blog post by Wictor Wilén on the original Yes vote :

When this day approached both camps, the pro Office Open XML team and the no-sayers both gathered their forces for the final battle. We all entered the meeting at the last possible minute and we all was signed in to the meeting.

Sounds less like a specification meet and more like a medieval gathering of Vikings. Sam Hiser writes that the French meeting degenerated into what sounds like a slanging match. The Inquirer covers more, to coin a phrase, un-standards-like behaviour.

Jason Matusow writes on his blog :

If Open XML is to be approved for standardization at JTC1, it needs to do so by the book... it is critical that these activities remain within the realm of ethical behavior as well as behavior defined by the rules for the JTC1 process.

The stuffing of voting bodies, amazingly enough, seems to be within the rules - ISO addresses the question of who represents a country, but is vague on how national voting committees are constituted. This should be corrected. However with all due respect to Jason, it seems to me that the realm of ethical behaviour is being gerrymandered. The ISO Code of Ethics states, amongst other things,

ISO members are committed to developing globally relevant International Standards by :

ensuring fair and responsive application of the principles of due process, transparency, openness, impartiality and voluntary nature of standardization

There are atleast three principles in there that have been violated in several countries, if reports are to be believed. A case has also been made for involuntary canvassing in the Inquirer article. It is only because the Internet has ensured openness that the allegations are receiving publicity.

Unfortunately, actions coming under scrutiny are those seeking to get the standard approved. OOXML supporters seem to be ignoring the fact that a "No with comments" vote will be converted into a Yes during the Ballot Resolution Meeting if objections are addressed. When the comments are distributed after September 2, ECMA has until February 2008 to either refute them or change the specification to nullify them.

Jason also writes in an earlier post that all comments (including those prefaced with a Yes) will be considered by ECMA. Apparently this kind of reasoning has also prompted some countries to vote Yes and append technical comments. The logic escapes me : There has to be an incentive for the proposer to respond adequately to comments. The incentive is converting a No into a Yes.

When a voter has already voted Yes, the voter gives up the right to satisfaction. All comments could potentially be addressed by calling them invalid, for example. I might be doing ECMA an injustice by exaggerating, but the possibility exists.

There is also the chance that a vote can be changed from a Yes to a No, but if the original Yes vote is tainted, what odds would a gambler give on a change happening? Brings me back to the original question of the process.

I live in a country where subverting processes is done more often than one cares to remark on. Most of us avoid a code violation ticket by speculating with the constable on what miracle pays for his tea. An ISO process is held to higher standards, I hope.

Tags : OOXML ISO Standards