The Standards Blog

The OOXML Contradictions Disclosed

OpenDocument and OOXML
Someone was kind enough to send me the package of materials distributed by JTC 1 earlier today to its members (I have authenticated these materials through a second knowledgeable source as well). The package contains each of the responses filed during the ISO Fast Track Contradictions period for Ecma 376, the specification based upon Microsoft's OOXML formats, as well as the responses prepared by Ecma to those responses.   As you'll recall, I had been told that the majority of these responses were critical, while Microsoft had downplayed them, suggesting that most or many were either neutral, or in fact "laudatory." As they are to be made public shortly, you will be able to perform your own comparison, but here are brief extracts from each of the 20 responses received by JTC 1 (the package confirms, as I had earlier reported, that a total of 20, and not 19, responses were received). While these are only extracts, I have reproduced short out takes that are sufficient to allow you to now see the actual mix between favorable, neutral and negative responses. As you will see, 14 of 20 responses were clearly negative, two indicated divisions of opinion, three were inconclusive or neutral, and one offered no objections. I think that it's fair to say that the over view that I had previously offered has proven to be accurate.
What happens next? The files sent to me also include the JTC1 transmittal note, which indicates that after internal consultation, next steps will be communicated to the National Bodies "in the very near future."

Here are the extracts, divided by category, taken directly from the original responses filed by the 20 national bodies. 

Updated 10:50 EST 3/3/07: I have not felt comfortable posting the full documents that I have received.  However, I will provide links to them as others get copies and post them on line.  You can find the summary Ecma document, with Ecma's proposed actions, through the following link that appears in a March 1 article by Eric Lai at ComputerWorld, or in this link from a March 3 Groklaw article by Mathfox.  The same article by Mathfox includes the full text of the French response.

Favorable responses ( 1 ):   Romania (brief response): "We agree with the project as it is."   Neutral or inconclusive responses ( 3 ):   India (brief response): "It is requested to extend the contradictory period by one month."   Italy (brief response): "Italy has been unable to assess the existence of contradictions of [OOXML] to JTC 1, ISO or IEC standards. The term "contradictions" does not appear to be defined, not even at the level of examples."   Netherlands (brief response): "The Netherlands Standardization Institute is changing its reaction on [OOXML] [from? This is not indicated]. The Netherlands Standardization Institute votes "abstain." [Updated 3/2/07:  I have heard from multiple sources that the Netherlands experience replicated the American National Body experience.  See my comments and those of some visitors at the end of this blog entry for further details.]    Divided responses ( 2 ):   Hungary (short response): "[Hungary] abstains as a result of differing views in the National Committee about the necessity of the Fast Track handling of the subject."   Norway (short response): "The document is very comprehensive, and responding on whether there are contradictions to JTC 1 and other ISO and IEC standards is difficult within the short timeframe….There are things in this document that is [sic] probably contradiction to for example ISO 8601 and ISO/IEC 8632….On the other side; we recognize that this document covers areas that [ODF] does not, and these areas are very important for many users with considerable legacy in office documents."   Negative responses ( 14 ):   Australia (two page response): "Australia proposes that this document be referred back for discussion within SC 34 before it goes to [Fast Track Ballot], given the many significant issues that need to be clarified."     Finland (two page response): Abstains, objecting to the short amount of time provided to review such a lengthy specification, and requesting that OOXML be removed from the Fast Track process.   Canada (short response): "Canada perceives that there are contradictions between" OOXML and other named ISO standards. Canada "believes that all options need to be considered, up to and including cancellation of the Fast Track ballot. Canada does believe that the perceived contradictions should be resolved before the Fast Track Ballot does proceed. Vote: Disapprove With Comments."   Czech Republic (short response): States that 30 days is "too short for the preparation of comments to this document…We are not convinced that there do not exist contradictory provisions with other ISO and ISO/IEC standards….For this reason the Czech Republic suggests using the standard procedure for the development of ISO/IEC standard" rather than the Fast Track process.   Denmark (short response): Notes "perceived contradictions" with five ISO standards and other difficulties, but does not see a problem with advancing to the next step in the Fast Track process.   France (short response): Indicates a lack of consensus on what constitutes a "contradiction," but (unlike the US response), Afnor (the French NB) provides a response based on the two opposing viewpoints expressed by its members. Specifically, it finds that OOXML would not "prevent the coexistence" with ODF (this is the Microsoft definition of a "contradiction"), but is inconsistent with existing ISO standards (this is the definition subscribed to by the contributors to the Groklaw Wiki, among others).   Germany (short response): "The German NB [National Body] prefers to have a harmonized Standards [sic] over having two Standards serving basically the same user requirements." The response goes on to ask that the ODF-based ISO standard be examined to determine what is lacking to permit backward compatibility for existing Office documents, and offering OASIS the opportunity to augment the ODF standard.   Japan (brief response): "[OOXML] seems to be competing and incompatible with [ODF]." Japan appears to be proposing that the two standards be harmonized. It also has IPR concerns.   Kenya (twelve page response, plus two page annex): "The reason why [OOXML] is so large is because is mainly because it does not leverage on any other well known specification in ISO. As such there are large contradictions with the preexisting international standards that are already mature, with wide and independent implementations….Due to its size, it should be reason enough why this documenet [sic] should not be allowed to be "Fast Tracked…. As can be seen from the above [the] proposed [standard] violates in many ways the current JTC1 Directives and therefore need be withdrawn form the Fast-Track Process."   Malaysia (four page response): "…Malaysia has identified the following contradictions to JTC 1, ISO and IEC standards for consideration:"      New Zealand (brief response): "The National Body for New Zealand objects to [OOXML] proceeding via the JTC 1 Fast-Track process because its adoption would be in breach of JTC 1 Resolution 27, 2000 "Consistency of JTC 1 Products."   Singapore (three page response): "The Singapore National Body has identified the following contradictions with existing ISO/IEC standards"   Sweden (short response): "SIS [the Swedish NB] is not in the position to make a statement on whether this document should be passed on for a Fast Track ballot or not. There are, however, concerns from some of our members that two ISO/IEC standards for open document will be perceived by users as confusing and maybe also as a source of interoperability problems."   United Kingdom (three page response): "…the UK does not believe that ECMA 376…is an appropriate candidate for the fast-track procedure."

For further blog entries on ODF, click here

subscribe to the free Consortium Standards Bulletin

Comments

Permalink
Why does the French response refer to the US response when there does not seem to be a US response?

The US decision was not to respond, due to a failure to agree on what a "contradiction" is.  I wrote about this at length on January 31, but here's the  most germane information from that post:

Yesterday, I learned that the Executive Board of INCITS decided earlier in the day not to propose to ANSI that any contradictions need be identified between OOXML and any ISO/IEC standards, Directives or other rules.  The reason is that Microsoft, which has a member on the committee, has persuaded a sufficient number of members of the INCITS Executive Board to adopt a very conservative definition for a "contradiction" – that definition being essentially the one articulated by Brian Jones at his blog – that a contradiction arises only where a system could not run two products, each of which implements one of the two specifications in question. 

The alternative suggestion made by others on the INCITS Executive Board had been to present all perceived contradictions to INCITS, accompanied by an explanation that there was a division over whether a narrow or a broad definition of contradictions was appropriate.  This would have allowed the managers of the process at JTC 1 to decide what to record as a contradiction, and what to disregard.  Unfortunately, sometime between Monday and Tuesday of this week, the consensus behind that course of action evaporated, and the narrow view carried the day.  The US formally sees 

I've been told by two sources that a somewhat similar situation occurred in several other countries, sometimes successfully, and sometimes not.  I've heard from two sources that this is what caused the Netherlands to abstain.  Here's the email that I received from one visitor:

I like to give you some more info about the Dutch situation.  I got my info from

The initial response was that it is too much and too complex to go the fast track and they would vote against it. They actually told this to ISO.  Under pressure of Microsoft on Friday before the deadline they changed their stance. 

  -  Andy

Permalink

Hardly any concrete information about the nature of the supposed found contradictions. Only a one reaction suggesting reference to the ISO 8601 and ISO 8632.

Especially the 8632 reference is a total joke as OOXML does not contains a graphics meta file definition so it can never contradict that one of ISO. A completly ridiculous argument.

Also OOXML actually does use and reference ISO 8601 (or a subset of that format via the standard XML schema's that is uses). It does use other dateformats only in parts of the spec particularly for spreadsheet calculations.

Looking at such arguments it does look like some of the national bodies were severly influenced in their ideas by the sometimes anti-ooxml campaign suggesting that these were issues to oppose IS standardization fronted by IBM and Groklaw and also included this blog..

 

Looks to me more like Microsoft was applying as much pressure on as many countries as possible and continues to apply pressure and astroturfing techniques to try to brute-force it's product spec through the standards process. It's too bad they can't simply adopt existing ISO standards instead of incorporating the Microsoft proprietary implementations into the OOXML standard by reference and it's too bad that MS feels the need to subvert the ISO standards process through subversion and intimidation of the member countries.

I think opponents of MS put a lot more pressure on the national bodies than Microsoft ever did.

Also several of the ISO national bodies contains members that work for  Microsoft competitors like IBM.

I wonder how much of those exerted pressure on their fellow members to object agianst this standardization proces.

I think opponents of MS put a lot more pressure on the national bodies than Microsoft ever did.

Well of course they did. It's not often a standards national body receives letters from Real People rather than from the lobbyists of multinational organisations. Especially over something as arcane as XML-based document formats.

Also several of the ISO national bodies contains members that work for  Microsoft competitors like IBM.

I know people at Cisco and Juniper who serve on international standards bodies, and I know they would not put the interests of their employer before the interests of the standards body, and I also know that those two employers would never ask them to do this.

Why do you believe that IBM or those individuals who are employed by IBM would behave differently?

Usually the staff working on standards committees are extraordinarily talented -- no one wants to have to ship a product which implements a poor international standard. An employer asking an employee who serves on standards body to vote in a particular way is counter-productive: the employee would simply make such a request public and will have a new employer by the following morning. I'll note that no IBM employee has done this, so I think we are safe in assuming that IBM did not give any direction to its employees.

I wonder how much of those exerted pressure on their fellow members to object agianst this standardization proces.

Most "fellow members" would be outraged by such behaviour and would have no difficulties saying so in public. So names and dates please, otherwise we should assume people behaved honourably.

In short, your imputations lack likelihood and evidence.
Glen Turner, gdt at acm dot org.

Especially the 8632 reference is a total joke as OOXML does not contains a graphics meta file definition so it can never contradict that one of ISO.

EOOXML supports wide rage of popular graphic formats - but doesn't support  standard ISO8632 one. This is contradiction.

Also OOXML actually does use and reference ISO 8601 (or a subset of that format via the standard XML schema's that is uses). It does use other dateformats only in parts of the spec particularly for spreadsheet calculations.

Using "other dateformats" in specification is prohibited. This is also contradiction.

ISO's definition of contradiction views it from point of view of interoperability/harmonization/etc. IOW when you need to input a date you do not have to think "where I am - is it word processor or is it spreadsheet application? which date format shall I use?" This is all what standards are about: you can put the same date everywhere and standard complying applications would understand you w/o problems.

 


Permalink

It is embarrassing that the France reaction actually references the Grokdoc objections as a kind of definition for ISO contradictions.

The Grokdoc definitions page is actually protected against updating so the obvious flaws in the arguments on those pages cannot be corrected. I have set up a blog with some counter arguments on ooxmlhoaxes.blogspot.com although of course that has little influence.

It shows a very sad state of affairs amongst the national bodies in ISO if they start relying on the info of a known ODF supporting and anti MS and anti OOXML site for definitions for ISO contradictions.

Surely the french national body should look at the JTC1 committee for definitions of ISO contradictions.

The referral to percieved issues on Grokdoc show that the influence of the mailing campaign by Groklaw are very strong as it influences national bodies in there actions based on a totally onesided anti campaign.

The Wraith

Permalink

I apologize for the confusion created by my shorthand reference to the Groklaw Wiki.  No such reference appears in the French letter, and as I re-read the French response, it doesn't even make good sense as a shorthand reference.  I had used it simply to spare myself from doing more typing (the files sent to me are PDFs).  By way of amends, here is the exact text of the French response on the relevant points:

......Nevertheless, the Directives do not give any specific definition of what should be understood by the concept of "perceived contradiction".

AFNOR members are of the opinion that fundamentally, a contradiction occurs when either:

1.  the implementation of one standard prevents the coexistence with another standard in the same product or system, or

2.  both standards address a same set of user requirements (as an instantiation of this, when new standards are introduced to handle functionalities already handled with International standards)

...Considering element 1 of the definition no contradiction was identified by AFNOR.

Considering element 2 of the definition, AFNOR noted [a partial overlap with ODF, as well as interoperability, cultural and linguistic inconsistencies with three more existing ISO and ISO/IEC standards].

  -  Andy

 

I can't speak to the inconsistencies, but they could have been brought up by anyone that wanted to do so.  The only difference between what OOXML is going through and what ODF went through is that OOXML is coming in throught the Fast Track process, in which contradictions are brought up first, and all other comments second.  In the PAS process, all comments are brought up at the same time.

I recall that there were a number of objections brought up with the fact that ODF did not enable all language needs (such as languages that read from right to left, rather than left to right, which I assume were addressed), but I haven't gone back to check whether the specific ones you mention were brought up.  If you look back at my blog entries for early May 2006, you should find a link to all comments (I think).

  -  Andy

Permalink
The Netherlands Standardization Institute is changing its reaction on [OOXML] [from? This is not indicated].

The Dutch standards body initially wanted to vote against OOXML and file rejections, but they got a phonecall from Microsoft a day before the deadline and changed their position to abstain. That was all over the news (and widely criticized) here in The Netherlands and also posted on Groklaw and LXer around that time. Microsoft probably "educated" them to the Microsoft definition of contradiction.

No, it is not the rumour. I have seen the official vote that was cast. Quote: The Netherlands Standardization institute is negative about the start of a five month DIS ballot on 29500. There are many contradictions to JTC 1, ISO, and IEC standards, for example to: ISO 8601:2004 Data elements and interchange formats -- Information interchange -- Representation of dates and times ISO 639 Codes for the representation of names of languages ISO/IEC 8632 Information technology -- Computer graphics -- Metafile for the storage and transfer of picture description information ISO/IEC 26300:2006 Open Document Format for Office Applications (OpenDocument) v1.0 Detailed information about the contradictions you can find in the attached document. Because we are seriously concerned about accepting the ISO/IEC DIS 29500 for a five month DIS ballot, we have included more information to support our position. Furthermore: the Netherlands Standardization Institute thinks it is almost impossible to decide on a Fast Track review in only one month, on a large and complex document as N8455.

Here's a thread on LXer about it: http://lxer.com/module/forums/t/24600/

The first comment in that thread links to the original news article in the Dutch newszine "automatiseringsgids". Sadly their articles become members-only after 30 days so the original article is no longer available (and Google didn't cache it). Searching for NEN and OOXML in Google brings up more references.

Permalink
<p>The full contradcitons and response are published now. </p> <p>I posted a link to them here </p> <p><font face="Arial">ooxmlhoaxes.blogspot.com</font></p> <p>The Wraith</p>

I have read (some of ) the ECMA responses - assuming the document at http://www.computerworld.com/pdfs/Ecma.pdf is genuine.

One of the arguments made is that ODF and OOXML could not me harmonised  in much the same way that HTML and PDF could not be harmonised. I'm not an expert ion this but surly that is not right. Perhaps a community project maybe on Grokdoc could define a namespace which would allow all the elelments of OOXML that can't yet be captured in ODF to be drafted... Perhaps that could cut off that particular argument.

just a thought!

Harmonisation would require that for instance OASIS and allow would allow for massive changes additions to the ODF spec which is unlikely to happen.

On the other hand the OOXML spec could be altered in future version to allow for more standardisation simular to ODF. Also it could move deprecated features to an annex (slugging 1k pages of the specs).

So I think that OOXML has potential to move towards ODF but the other way around I don't see that happening as the creators of ODF have particular interests in opposing Microsoft and are not interested in moving towards MS at all. Microsoft on the other hand could be interested in improving OOXML and make it more interoperable as that makes it's product more sellable.

The Wraith

ooxmlhoaxes.blogspot.com

 

Wraith, I've read your OOXML hoaxes blog and you should rethink some of your arguments. You seem to be unaware of just how Microsoft will be able to leverage their standard to their sole advantage. For example, you correctly state that (the legacy MS proprietary compatibility) parts of the OOXML specification are optional, but ignore the obvious fact that any application that doesn't include them will be severely disadvantaged (if not altogether stillborn) in the market place against Microsoft's all-inclusive products.