On July 1st, 2010 a new set of rules (directives) took effect in ISO/IEC JTC1 including new processing and voting rules for JTC1 Fast Track submissions. If these rules had been in effect back in 2007, OOXML would have died after its initial ballot.
Let’s take a look at some of the specific changes that were made in reaction to the events of 2007-08.
First, we see the elimination of the contradiction phase in Fast Track processing. If you recall, under previous rules, a Fast Track begin with a 30-day NB review period, sometimes called the “contradiction period”, where NBs were invited to raise objections if they think the Fast Track proposal contradicts an existing ISO or IEC standard. This was followed by a 5-month ballot. The problem was that the word “contradiction” was not defined, leading to various irreconcilable interpretations. In the case of OOXML 20 JTC1 National Bodies (NBs) raised contradictions. Evidently, the passage of time has lead to no progress on defining what exactly a contradiction is, so the contradiction period has been eliminated entirely. Instead, looking for “evident contradictions” (still undefined) is given to JTC1 administrative staff, which is the surest way of guaranteeing that we never hear of contradictions again. The Fast Track DIS ballot remains at 5-months, so net-net this accelerates processing by one month.
Next, we see some clarification around how NBs should vote on Fast Tracks. Back, during the OOXML ballot, Microsoft made a huge effort to convince NBs to vote “Yes with comments” if they found serious flaws in the text, with the promise that they would all be addressed at the BRM. Well, we now know that this was a big lie. Very few issues were actually discussed and resolved at the BRM. And most of them were addressed by merely saying, “Sorry, no change”. At the time I argued that the rules were quite clear, that disapproval should be voiced by a “No, with comments” vote. Well, we now see another small slice of vindication. The revised rules now state:
If a national body finds an enquiry draft [ed. A Fast Track DIS is an ‘enquiry draft’] unacceptable, it shall vote negatively and state the technical reasons. It may indicate that the acceptance of specified technical modifications will change its negative vote to one of approval, but it shall not cast an affirmative vote which is conditional on the acceptance of modifications. (ISO/IEC Directives, Part I, Section 2.7.3)
I assume this is clear enough now.
Another change is that if the DIS ballot fails to get sufficient votes, meaning less than 2/3 approval of ISO/IEC JTC1 P-members, or more than 25% disapproval overall, the proposal dies at that point. It doesn’t go on to the BRM. Game over. If this rule had been in place back in 2007, OOXML would not be an ISO standard today.
Finally, we see the requirement for a Final DIS (FDIS) text for review and approval by NBs. Back in 2008 I was quite vocal about the absurdity of having NBs vote on a text that they were not allowed to read. Several NBs lodged formal objections at the time as well. All this was dismissed by JTC1 staff. But reality struck when NBs reads the actual published version of OOXML, and saw that it did not contain all of the changes mandated by the BRM. So belatedly, but better than never, the rules have been changed. Fast Tracks now require an FDIS text for NBs to review, along with a 2-month ballot on it.
There are also smaller, less substantial changes. For example, the dedication to Jan van den Beld, the former head of Ecma, for his “unwavering dedication to the development and evolution of the JTC 1 procedures”, has been removed. Ironically, both Ecma and Microsoft have indeed made long-term contributions to the evolution of Fast Track in JTC1, but probably not the way they intended.
The new ISO/IEC Directives are posted online. Note that one document expresses the common rules for ISO and IEC, while another is a set of supplemental rules which apply to only ISO/IEC JTC1. Evidently, we’re supposed to consult both documents and mentally merge them whenever trying to determine what the rules are.
Hi Rob,
Thanks for the narrative of these complicated documents.
I think the clarification on when to vote “no with comments” is good – Denmark has consistantly been using this check-box when we wanted to apply maximum pressure to get things changes after a ballot.
However – I do think that the possibility of having the DIS die at this point might have contradictory consequences. Denmark has repeatedly been voting “no with comments” to proposals where we kindof liked it – but that there were things we would like to have changed before voting yes. With the changed rules, we might end up not having a second vote if enough feel like we do.
Any thoughts on this?
Jesper, remember, a DIS ballot is not a ballot of technical committees or subcommittees. So this is not the preferred time for deep technical analysis of the draft.
If you look at “normal” progression (not Fast Track) you see a similar pattern. Committee Stage (CD) is “the principal stage at which comments from national bodies are taken into consideration, with a view to reaching consensus on the technical content. National bodies shall therefore carefully study the texts of committee drafts and submit all pertinent comments at this stage” (2.5.1). But at higher stages, for example when when JTC1 votes on the FDIS, you have the same restriction as a DIS, that you cannot submit conditional Yes votes (2.7.2).
I think the logic of the voting rules is clear. “No with comments”, means you wish the DIS to be defeated. “Yes with (or without) comments” means you want it to pass. If more than 25% vote disapproval, then there is no second chance for the proposal. It essentially “goes back to committee”, in the DIS case back to the submittor of the Fast Track. But this is no different than for a home-grown JTC1 draft, right? If you get too many disapproval votes in an FDIS ballot, you do not go to a BRM. It also reverts back to the subcommittee.
So what do you do in your hypothetical Denmark situation, where you see a DIS that you “kind of like” but have some technical issues? Well, what would you do for a non-Fast Track draft at that stage? I don’t think the intent of Fast Track and PAS is to have a different set of criteria for the technical contents of the drafts.
I’d appreciate any clarification you might offer regarding the relationship between the new JTC 1 Supplement and ISO/IEC JTC 1 Directives, 5th ed., rev. 3. The new supplement seems to speak to many of the same topics, but less than all. E.g., there is but a single mention of interoperability in the Supplement whilst the Directives supplied an entire annex of guidance on the topic.
It is not clear to me that the Supplement is intended to replace the Directives. The Supplement states at lines 117-120:
“Part 1 of the ISO/IEC Directives, together with this Supplement, provides procedural rules to be followed by ISO/IEC JTC 1. There are, however, other documents which provide further guidance, such as JTC 1 Standing Documents. Forms unique to JTC 1 are found in the Templates at
http://isotc.iso.org/livelink/livelink?func=ll&objId=8913214&objAction=browse&sort=name [.]”
The linked page includes a download link for ISO/IEC JTC 1 Directives, 5th ed., rev. 3. I’ve found no language in the Supplement thus far indicating that it supercedes the JTC 1 Directives. That’s an important issue to resolve because JTC 1 Directives provide in relevant part:
“This document replaces the ISO/IEC Directives, Part 1 – Procedures for the technical work, but has been developed to be consistent with them. Any differences between this document and the ISO/IEC Directives, Part 1 are dictated by the nature of information technology work and have been authorised by the ISO Secretary-General and IEC General-Secretary ( Secretaries-General) and Councils of IEC and ISO. Where differences between this document and the ISO/IEC Directives exist, the provisions of this document shall govern.”
Any enlightenment to offer here? Did I miss something in the Supplement?
The new JTC1 procedures consist of ISO/IEC Directives, Part 1 plus the JTC1 Supplement. The idea was to put the procedures common to ISO, IEC and JTC1 into a common document and then in the Supplement to express any peculiar JTC1 rules.
There is a transition period, until July 1st, 2011, where some in progress work can continue to follow the old rules, while new projects start under the new procedures. So I would not say the old Directives are dead, but they are only used for the in-progress work, and only for the next 12 months. That’s my read of it.
It is quite possible that things were lost in the move to the new documents. You mention the interop annex. I just noticed also that they lost the clause on how an organization can become a Category A Liaison like Ecma. They define the rights of such a liaison, but never talk about how one applies for or is approved to have that status.
It would be good to have a comprehensive list of the “documents which provide further guidance”. Surely one must be ISO Directives, Part 2, which provides the drafting guidelines for standards. Another one must be the PAS Management Guide. There may be others.