The standardisation of OOXML is moving into a new phase. Most countries have now reached the end of their public comment gathering period, and are beginning to formulate their vote for the DIS 29500 ballot which closes on September 2.

Lobbying

Here in the UK there has been much lobbying and petitioning from all sides, and there is a stack of (mostly identical) letters for BSI to sift through, both for and against OOXML. In general, the pro-OOXML camp seems to have been more effective at mobilising support from politicians, public figures and captains of industry; the anti-OOXML camp has, in contrast, the support of the “blog collective” as exemplified by the noooxml campaign — though here, as so often in the blogosphere, there is something of an echo-chamber effect as the same self-convinced arguments re-echo from site to site, without achieving wider penetration in the real world. If the standardisation of OOXML was an exercise in lobbying (which, thankfully, it is not – at least not in the UK), then the OOXML proponents would have won with a knockout blow.

Internationally, is it interesting to note the influx of new members to SC34, the ISO sub-committee administering much of the standardisation process. Last week alone saw the announcement of three new P-members to the committee: Bulgaria, Côte-d’Ivoire and Sweden, as well as three new O-members: Greece, Mexico and Sri Lanka. As Ken Holman, our esteemed SC34 committee secretary wrote, “we warmly welcome their input”.

The UK Vote

Although public comments are no longer being taken, BSI’s technical panel will continue to scrutinise and comment on the DIS 29500 text until 24 July. The panel itself will not decide the UK’s position (though members will have a chance to express their views), but will pass the mass of accumulated evidence up to the standing technical committee IST/41, whose task, in turn, is to recommend a UK position. This recommendation (or, if consensus cannot be reached, a bunch of disparate views) will be passed up to the topmost ICT/-/1 committee, and it is here that the UK’s voting position is finally decided. As is apparent, BSI is a deeply hierarchical bureaucracy though – to its credit – one which appears well fortified against undue influence from interested parties. Evidence, not interests, will determine the UK vote.

While the UK position is undecided, and it would be foolish to try and second guess it, it seems to me personally that one position is ruled-out: a straight unqualified “yes”. This is because, leaving aside wider considerations, the text is often simply so poorly written. My colleague, and fellow BSI panellist, Andrew Sales, singled out this paragraph from page 1912 of the spreadsheetML specification as a particularly telling example:


Such gobbledegook has no place in an ISO standard; seeing this, one can only wonder at the rigour of Ecma’s standards process, and conclude that something is wrong that needs fixing.

The End Game ...

When the votes are in and counted, the standardisation process moves into a phase known as “ballot resolution”. I have been diligently attempting to keep the standardisation section of the Wikipedia OOXML entry accurate in describing this, though what I do not do there is ask the many unanswered questions there are about the precise sequence of events.

The process is governed by the ISO/IEC JTC 1 Directives. On the surface this reads like a thorough and detailed legalistic document, but on closer reading it – like a poor standard – admits questions and ambiguities which are difficult to resolve. Maybe I need to study it more, but I’m not alone in my puzzlement and have been engaged in email exchanges with several ISO standards veterans who are unsure of some of the implications of the text.

Such questions are:

1. Will there be a ballot resolution meeting under all circumstances? i.e. even if every country votes “no” will the standardisation process continue with a ballot resolution meeting (BRM) which is expected to assess national comments and reach a further, final, conclusion?

2. What changes can be agreed in this stage? Normally a text at the DIS stage must not be subject to any substantive change, as this will trigger a re-ballot. But does the Fast Track differ in being “no holds barred” for the BRM? Could the nations request DIS 29500 to be completely re-drafted during the meeting?

3. How long will the BRM take? Judging by the volume of comments from the UK alone, and the usual pace of decision making in ISO meetings, even a one week meeting is unlikely to scratch the surface of the consolidated comments. Assuming the meeting needs to adjourn and reconvene, then at the natural ISO pace this could end up being a “Crawler Lane”, rather than a Fast Track, standard.

... And Beyond

If DIS 29500 becomes a full International Standard, then the question of its maintenance arises. Interestingly Microsoft have been making public statements (for example at a recent public meeting) which suggest they envisage the maintenance of the standard continuing outside of Microsoft or Ecma, and that it will fall under control of JTC 1. This has been written up by John Scholes, a BSI panellist (I should enter the caveat – and as John makes clear on his site – he is coming very much from his own, anti-Microsoft, position).

If this happens, then it’s likely the current controversies will rumble on in one form or another. But, given that these file formats may find themselves out in the open under ISO control, perhaps ODF proponents would do well to take a longer-term view on document formats than that short-sighted one of the current “noooxml” campaign.