Couldn't put it better myself
Much has been said by many people about Microsoft's attempt to make their new document format an ISO standard, regardless of the pre-existing standard for office documents ISO 26300. First they rushed it through Ecma (a standards "rubber stamp" body), then moved it onto ISO on a fast track process. The 6000 pages that are Ecma Office Open XML (as it is correctly known) are moving through a process that is 9 months long from start to finish.
As DIS 29500 (DIS=Draft International Standard) some serious people from national standards bodies all over the world have looked at it. To counter the criticisms these serious people have made, Microsoft put forward some very not-serious people; marketing staff and senior officers from Microsoft partner companies. Now, a note of caution to readers making the assumption that my spin on this draft standard is coloured by my deep-seated mistrust of them as a company; while there may be some truth in that point of view, some of the errors in the draft standard are glaringly obvious. It is also useful to understand the three voting positions available to a national body during the fast track process: approve (approve the standard as it is), abstain (generally used when the national feels it has insufficient technical resource or time to participate, also used when the alternative votes could not be agreed upon) and disapprove with comments (where if the comments are addressed then the vote changes to approve i.e. a conditional approval). So when these shiny new Microsoft inspired members of national bodies vote to approve the standard as is, it is quite clear that one or more of the following conditions hold true:
(i) the people voting were technically incompetent and unable to use the web to overcome their abject incompetency;
(ii) they were technically compentent but didn't actually bother to review the standard or the critical material available;
(iii) they just voted they way they were instructed to by Microsoft.
So back to the serious people's perspective. BSI had a wiki in which they gathered over 600 comments, many of which were editorial amendments, 100 or so were technically significant and a few which were show-stopping. But the neatest summary of the reason to vote disapprove with comments comes from Farance Inc., a member of the US standards body technical commmittee:
-------------
Farance Inc. finds the current circumstances in the development of the US position on DIS 29500 troublesome for the US.
(1) Too much standards wording in too short a review cycle. A 6000+ page document was submitted for JTC1 fast-track ballot with just 6 months for technical review and approval. Even documents one tenth the size have taken years to review and revise according to the normal quality for ISO/IEC standards. We are concerned that many INCITS EB members do not acknowledge that more pages mean a higher risk (i.e., less likely) that the document is acceptable. Furthermore, the 6-month review has not permitted adequate review for such a large document. None of the reviewers in the INCITS EB or in INCITS/V1 have had enough time to completely review the document in detail. The technical comments compiled by the US represent merely a "first cut" at a review and, considering the National Bodies only have one review cycle to comment upon a fast-track ballot, the best that can be done with a limited amount of time.
(2) Poor quality. The document was recently developed and is relatively immature, as evidenced by the hundreds of significant technical problems. Had the document been hardened by competing, independent implementations and years of significant industry *interpretation*, as is typically the case for National Body standards submitted for DIS fast-track, many of these obvious errors would have been discovered, resolved, and corrected.
(3) Poor conformity. We have pointed to the significant problems with the conformity wording, which is significantly flawed and ambiguous. Given these flaws and ambiguities, it might be possible that only the creator of this technology would conform to this proposed standard. This is especially problematic for the marketplace of product and services as the buyers of the technology cannot be certain what they are buying (due to flaws/ambiguities in DIS 29500 conformity wording) and the sellers of the technology are at risk because of potentially false/misleading claims about their conformity (due to flaws/ambiguities in DIS 29500 conformity wording).
(4) A possible trade issue. These conformity flaws/ambiguities are particularly worrisome because INCITS EB members (including US Government and US industry) find it acceptable to approve such a document. It is not a question of whether or not the US supports DIS 29500 becoming a standard (Farance Inc. supports 29500 becoming a standard after the technical issues have been fixed), it is a question of whether or not certain technical issues are "must-have", which would be indicated by a vote of "Disapprove with Comments, Conditional Approval (as per JTC1 Directives 9.8)". Given the weaknesses of the standard, it appears that possibly only one US company could implement the standard as presented. Would the US be worried if another National Body submitted a standard for fast-track that appeared to be poised to take advantage of weaknesses in the standard such that *only* companies from that National Body could supply products and services? The US will have less credibility in making such arguments against other countries who take such actions. This stance is a strategic flaw and it is certain to have a negative impact across all US technologies for years and decades to come.
(5) A poor exemplar of standards principles, practices, and processes. In INCITS/V1, of the 15 V1 members (of a total of 25) voting for the recommended US position to be "Approve with Comments", 12 of those 15 "Yes" votes joined INCITS/V1 3-6 weeks prior to the vote deadline
(many of them newcomers to INCITS and JTC1 standards). In the INCITS EB, we discovered that 2 EB members were letting a third EB member control and direct their vote. Since when has the US National Body been accepting of one member having three votes? We discovered that it became impossible to understand the rationale of certain EB members votes because they said other EB members were controlling their votes. We've heard secondhand that there will be new P-members of JTC1 joining just in time for participating in the DIS 29500 vote at the JTC1 level. We've heard from the 29500 Project Editor (a participant in V1 and an participant in ECMA) that the BRM will only be 5 days -- an impossibly short time to review, discuss, and properly resolve these hundreds of comments. Furthermore, the 29500 Project Editor states that he believes there will be no other face-to-face meeting for National Bodies to discuss the resolution of these comments (note: given the complexity of these issues, an E-mail reflector, wiki, or on-line electronic forum would not be an acceptable method for resolving the issues).
(6) A poor standard for ISO/IEC, a poor impression about ISO/IEC standards and their processes. Given the "first cut" of the now-understood technical flaws (and others to be known over time), given the shortness of the BRM, and given the rush to get DIS 29500 approved, the next several years will be spent processing defect reports on 29500, which will hamper the adoption of 29500 as a standard and give the world a poor impression of ISO/IEC standards. While "ISO 9000 Quality" is world-known "brand" and connotes an improvement in quality, the failure of 29500 (regardless of approval as an ISO/IEC standard) will have its own "brand" but will have a negative connotation -- a negative connotation known at every place someone uses a computer and every time someone purchases a computer system or software. In conclusion, we are extremely worried about the DIS 29500 ballot process, the US position on this ballot, and the processes surrounding the standards process. We still believe "Disapprove with Comments, Conditional Approval (as per JTC1 Directives 9.8)" is the appropriate response -- regardless of the US comments reminding NBs that it can change it vote. We are opposed to the "Approve with Comments" response. Reluctantly, we accept the "Abstain with Comments" response due to lack of time in formulating the US position.
-------------
See here for the official entry.
Posted at 10:09AM Aug 25, 2007 by Chris Puttick in General |