Mastodon
Where is there an end of it? | Microsoft Fails the Standards Test

Microsoft Fails the Standards Test

The second anniversary of the approval of ISO/IEC 29500 (aka OOXML) is upon us. The initial version of OOXML (Ecma 376 1st Edition) was rejected by ISO and IEC members in September 2007, and it was only after extensive revisions and a bitter standards war in the following months that a revised format was finally approved on April 2, 2008.

The key breakthrough of the revision process was the splitting of the specification into two variant versions, called “Strict” and “Transitional”. The National Bodies confined all the technologies they found unacceptable to the Transitional format and dictated text to be included in the standard intended to prohibit its further use:

“The intent […] is to enable a transitional period during which existing binary documents being migrated to DIS 29500 can make use of legacy features to preserve their fidelity, while noting that new documents should not use them. […]

This annex is normative for the current edition of the Standard, but not guaranteed to be part of the Standard in future revisions. The intent is to enable the future DIS 29500 maintenance group to choose, at a later date, to remove this set of features from a revised version of DIS 29500.”

I was convinced at the time, and remain convinced today, that the division of OOXML into Strict and Transitional variants was the innovation which allowed the Standard to pass. Enough National Bodies could then vote in good conscience for OOXML knowing that their preferred, Strict, variant would be under their control into the future while the Transitional variant (which – remember – they had effectively rejected in 2007) would remain purely for the purpose of accurately specifying old documents: a useful aim in itself.

Promises and reality

Just before the final votes were counted, Microsoft made some commitments. Mr Chris Capossela (Senior VP, Microsoft Office) wrote an open letter promising what would happen if the Standard passed. Two years on, we can fill out a report card for a couple of these promises and determine how well Microsoft is doing …

Microsoft's promise on standards support in products


“We've listened to the global community and learned a lot, and we are committed to supporting the Open XML specification that is approved by ISO/IEC in our products.”

On this count Microsoft seems set for failure. In its pre-release form Office™ 2010 supports not the approved Strict variant of OOXML, but the very format the global community rejected in September 2007, and subsequently marked as not for use in new documents – the Transitional variant. Microsoft are behaving as if the JTC 1 standardisation process never happened, and using technologies (like VML) in a new product which even the text of the Standard itself describes as “deprecated” and “included […] for legacy reasons only” (see ISO/IEC 29500-1:2008, clause M.5.1).

Knowledgeable experts present at the Ballot Resolution Meeting, knowing what Microsoft planned, have publicly repeated the International consensus position in alarm. XML Standards guru Rick Jelliffe (an Australian delegate at the meeting) wrote:

“If [Microsoft’s] default format is OOXML Transitional, then they have abandoned support for an Open Standards process: OOXML was only made a standard because of the changes that were made at the BRM. The original ECMA version of OOXML (which is the basis of Transitional) was soundly rejected, let no-one forget.”

And Danish expert and BRM delegate Jesper Lund Stocholm, running an analysis of Office 2010 files wrote:

“It has been the fear of many that Microsoft will never, ever care at all about the strict conformance clause of ISO/IEC 29500, and my tests clearly [are] a sign that they were right.”

Microsoft, however, takes a different view to the independent experts. Their representatives will argue (with some justification) that terms like “legacy”, “deprecated”, and “new document” are tricky to define, but then this argument extends to the bizarre assertion that the Strict variant need never be supported. I believe, however, countries expect a more reasonable, plain-dealing approach to their clearly expressed intent – not this kind of wheedling sophistry. Mr Capossela writes that Microsoft has “learned a lot”; but on the evidence before us now, this was wishful thinking.

Microsoft's promise on standards maintenance


“We are committed to the healthy maintenance of the standard once ratification takes place so that it will continue to be useful and relevant to the rapidly growing number of implementers and users around the world.”

It all started so well – defect reports came in from many national bodies and (via Ecma) from Microsoft themselves. A number of useful improvements were made to the text correcting obvious defects, and (in the Transitional variant) fixing some of the evident mismatches between what the Standard said, and what legacy documents actually contained.

But as time has gone on, the situation has deteriorated. At the recent Stockholm meetings corrections agreed at the February 2007 Ballot Resolution Meeting were still being implemented, and while fixes which were evidently required for Office 2010’s headline conformance behaviour have been given the red carpet treatment, some other reports from National Bodies have been left to languish. Unusually, in Stockholm one of SC 34’s working groups (WG 2) recommended to the plenary that the OOXML maintenance group (WG 4) be reminded to answer overdue defect reports – in the ISO world that counts as a diplomatic incident!

Most worrying of all, it appears that Ecma have ceased any proactive attempt to improve the text, leaving just a handful of national experts wrestling with this activity. It seems to me that Microsoft/Ecma believe 95% of the work has been done to ensure the standard is “useful and relevant”. Looking at the text, I reckon it is more like 95% that remains to be done, as it is still lousy with defects.

Ironically, the failure to resource maintenance properly is only going to damage Microsoft Office in the longer term. The simple validators developed by me (Office-o-tron) and by Jesper Lund Stocholm (ISO/IEC 29500 Validator) reveal, to Microsoft's dismay, that the output documents of the Office 2010 Beta are non-conformant, and that this is in large part due to glaring uncorrected problems in the text (e.g. contradictory provisions). It is also a worrying commentary on the standards-savvyness of the Office developers that the first amateur attempts of part-time outsiders find problems with documents which Redmond’s internal QA processes have missed. I confidently predict that fuller validation of Office document is likely to reveal many problems both with those documents, and with the Standard itself, over the coming years.

So – while maintenance is happening, I think calling it “healthy maintenance” would be over-optimistic given the current circumstances.

Someone has blundered?

Microsoft has many enemies who will no doubt see the current state of affairs as proof that Microsoft never even intended to be good standards citizens. Indeed standards and XML veteran Tim Bray, writing shortly after the standard’s approval, made a prediction which could now seem impressively prophetic:

“It’s Kind of Sad • The coverage suggests that future enhancements to 29500, as worked through by a subcommittee of a subcommittee of a standards committee, are actually going to have some influence on Microsoft. Um, maybe there’s an alternate universe in which Redmond-based program managers and developers are interested in the opinions of a subgroup of ISO/IEC JTC 1/SC 34, but this isn’t it.

I suppose they’ll probably show up to the meetings and try to act interested, but it’s going to be a sideline and nobody important will be there. What Microsoft really wanted was that ISO stamp of approval to use as a marketing tool. And just like your mother told you, when they get what they want and have their way with you, they’re probably not gonna call you in the morning.”

For me, the puzzle of it is that in many respects, Microsoft does appear to get it. Senior management seems to want standards conformance, as Mr Capossela’s letter demonstrates – indeed strategically, playing fair by standards has always seemed like the most obvious way for the corporation to extract itself from the regulatory thickets that have entangled it over the past decade. Microsoft employs many eminent and standards-aware people of unimpeachable record – they also obviously “get it”. And on the ground in the standards committees there are many delightful, talented and diligent people who seem fully-signed up to a standards-aware (dare I say “non-evil”?) approach—as the SC 34 meetings in Stockholm again recently evidenced.

And if we look elsewhere within Microsoft we can see – for example from their engagement with HTML 5 and work on MSIE – that they can move in the right direction when the will is there.

So why – given the awareness Microsoft has at the top, at the bottom, and round the edges – does it still manage to behave as it does? Something, perhaps, is wrong at the centre — some kind of corporate dysfunction caused by a failure of executive oversight.

But whether Microsoft senior management have directed the company to behave badly, or whether they have failed to control a bad corporate impulse, is ultimately of no interest or concern to the National Bodies engaged in Standardization: for them, the effect is the same. Some responses will, however, be necessary.

Moving forward

If Microsoft ship Office 2010 to handle only the Transitional variant of ISO/IEC 29500 they should expect to be roundly condemned for breaking faith with the International Standards community. This is not the format “approved by ISO/IEC”, it is the format that was rejected.

However, it is foolish to believe they won’t ship it as is – and before long the world will be faced with responding to that release. In my view moving forward from there will be difficult …

  • Governments, corporations, other large entities – in fact, anyone – procuring office systems with a requirement for standards-conformance need to have their eyes very wide open about what precisely they will be getting with systems which create new documents which are extended Transitional ISO/IEC 29500.
  • Microsoft Office 2007 (the current version) reads and emits unextended Transitional ISO/IEC 29500, and so – strangely – may represent a high-water mark of Microsoft Office standards conformance. Anybody wanting to work just with documents which (modulo defects) are fully specified by Standards wholly under International control may want to stick with this version of the software.
  • Microsoft should make a public open commitment to support OOXML Strict fully. A service pack bringing this support to Office should be developed as a priority.
  • JTC 1 explicitly created the Transitional variant with the intention they would “at a later date, […] remove this set of features”. Now is the time to start the wheels in motion for this removal (the text will of course remain available for the perfectly good reason that the legacy needs to be documented).
  • Any assurances Microsoft has given to regulatory bodies (such as the EU Commission) about standards conformance must be looked at very carefully giving full consideration to the circumstances of this release.
  • Ecma need to commit adequate resources to standards maintenance and pro-actively seek to improve the text, working together with SC 34, if there is any appetite to improve the Standard to the point where it can be a trouble-free, or even good, basis for interoperable office applications.

In short, we find ourselves at a crossroads, and it seems to me that without a change of direction the entire OOXML project is now surely heading for failure.

Comments (196) -

  • Jesper Lund Stocholm

    4/1/2010 3:38:14 PM |

    Hi Alex,

    What a good post! I largely agree with most of your considerations, but I do find it silly (or unrealistic) to think that Microsoft Office 2010 would support OOXML< S >. The time frame from publication of 29500 until launch of Microsoft Office 2010 was simply too narrow, and I would never expect Microsoft to pull the plug on release of Microsoft Office 2010 (even though one might wish they had done so) to get full support for OOXML< S >.

    I also find it puzzling to deal closely with Microsoft in this area – because as you point out, it is a like half of Microsoft “gets it” and the other half doesn’t. It can be really frustrating at times.

    And finally – I completely agree that Microsoft should step forward and publicly commit to supporting OOXML< S >. When pitching this thought to Microsoft employees I usually get the answer that Microsoft will never talk about future products. That’s fine – I don’t expect them to – but this is “just” the file format. It won’t reveal anything about any new features of Microsoft Office to commit to supporting OOXML< S >. Microsoft has once shown that they are willing to break their internal rule about “not introducing new features in service packs” when it matters to them (ODF-support came with SP2 to Microsoft Office 2007). If they care about OOXML< S > I’m sure they can do it again.

  • Alex

    4/1/2010 3:56:57 PM |

    @Jesper

    Yes, it may have been unrealistic for anyone to expect MS would  implement Strict in Office 2010 -- but ultimately this is a question of corporate will. Is the will there? I think not.

    Having taken the (no doubt commercially advantageous) route of targetting Transitional, MS now have to be prepared to take the hit for have shirked their responsibilities in regard to Standards support. After all, we as standardizers can't exempt implementors on the grounds that doing the right thing would have been too costly!

  • Rob Weir

    4/1/2010 8:49:12 PM |

    The interesting thing, I think, is how the situation is very different in the browser world, where it appears the Microsoft is taking standards seriously now.  Of course, in that area there is a single standard, and competition among the vendors to provide complete and conformant implementations.  I think the key thing was that there was a single standard (or family of standards) in a single organization, with many vendors involved.

    Imagine how it could have turned out differently if instead of having a common HTML standard, Microsoft had pushed through a "Microsoft Open Web ML" pseudo-standard, full of legacy crud and vague promises about future improvements.  And if no other vendor except Microsoft implemented Microsoft's Web ML.  And if Microsoft stuffed the committees that control the Microsoft Web ML standard.  And if they arbitrarily added or removed features to Microsoft;s Web ML based on what Microsoft's plans were in the next release of Internet Explorer.  And if they rejected or failed to act on feature proposals that they did not originate.  I wonder how that would have turned out.

    The situation with OOXML in SC34 is that Microsoft has no incentive to do anything more, and has sufficient dominance in that committee and in WG4 to ensure that nothing does happen. They have their ISO standard and the means to preserve it for their exclusive use.

    Your sensibility on this topic is around three years too late.  You don't have the votes to ask for a glass of water in SC34 if Microsoft doesn't want you to have it.  "Why buy the cow when you can get the milk for free?"  Anyone who believed otherwise was a fool.  It is too late for remorse now.

  • Bjorn Sveinbjornsson

    4/1/2010 9:46:54 PM |

    So. What will happen now?
    What do ISO procedure rules say.

    Could OOXML T be destandardized as mr. Jelliffe suggested some moons ago?

  • Gopal

    4/1/2010 10:06:58 PM |

    After a few lines into reading the post, I went back to check the date ( it was too close to April-1 ), and then finished reading.  Coming from the person who headed the BRM  this is pretty serious.

  • Alex

    4/1/2010 10:17:19 PM |

    @Bjorn

    It's not really an ISO issue. ISO makes standards and implementers may implement them (badly or well).

    Sometimes market considerations may feed back into the process. For example if a standard in never implemented that is a good reason for withdrawing it -- but that applies (now) more to the Strict than the Transitional variant of OOXML.

    Standards can be withdrawn at any time if that is the will of the Countries that participate ...

  • A. Rebentisch

    4/1/2010 10:41:14 PM |

    Come on! Didn't you notice that the EU committment settlement mentioned the ECMA version of the standard instead of ISO/IEC: ECMA 376? At least that committment is binding.

    We currently have a debate in the EU concerning recognition of consortial standards. It makes sense to think more about the fundamental difference between de jure and market standards.
    ec.europa.eu/.../

    As of ISO/IEC 29500 the process was incapable to fix the standard. Insufficient review was already given on the ECMA "fast track" level. The question on how to improve documentation can also be addressed on the level of productivity tools for standard development, review and compliance.

    Btw., in terms of hands-on: http://odf.cenatic.es/

  • Alex

    4/1/2010 11:03:30 PM |

    @André

    Ecma 376 is in step-lock with ISO/IEC 29500; they are effectively identical. In any case, I think the relevant clause is 17:

    (17) Office Open XML. The “.docx, .xlsx and .pptx” file formats used in the Office 2007 version of Microsoft’s Primary PC Productivity Applications shall implement the ECMA 376 Specification. This commitment shall apply to successor versions of Microsoft’s Primary PC Productivity Applications with respect to IS 29500.

    MS Office 2010 will be a "successor version" and so (by my reading), it shall implement IS 29500.

    On consortia standards - my understanding with that the EU was already minded to recognize standards from a variety of sources (so, IETF for network standards). But then there's also the question of European-level standards (e.g. from CEN and CENELEC). Frankly, the politics and complexity of this topic keeps me well away from it!

    I don't agree that JTC 1 maintenance process is insufficient - in fact I think JTC 1's maintenance processes are one of its strongest features. The problem here is not how the machine works, but that we need to be feeding a lot more into that machine ...

    Granada - nope, can't make that. The Brussels one, maybe (that's an easy trip for me).

  • Jonas

    4/1/2010 11:34:39 PM |

    Sorry for a stupid question beamed in from five years ago: But if Microsoft *wanted* to use a standardized document format (for regulatory, business or other reasons), why wouldn't they use ODF?

    When the ISO standardization process was in full swing the answer was one of perceived pragmatism: They just won't and OOXML is much better than the old proprietary format, so let's not ask that question anymore.

    Well, Microsoft came around on TCP/IP, they came around on the web and HTML, they will come around on HTML5 and the very second important customers demand it they will come around on ODF. I wouldn't be surprised if the ODF filter from Microsoft (wasn't that unthinkable a few years ago?) produced code closer to the standard than what Office OOXML does respectively.

  • Rick Jelliffe

    4/1/2010 11:54:07 PM |

    Jesper: I don't know why it is unrealistic to expect Microsoft to accept and generate Strict OOXML in Office 2010: they have had more than two years already. Most of the changes are trivial or systematic or were well in progress (e.g., VML had mostly been dropped for DrawingML by 2007) or would be features required for better support of ODF/Open Formula anyway.

    Rob: You mistake Alex's sensibility, I think. SC34 in general and Alex in particular has treated Ecma and Microsoft professionally, fairly and in an unprejudiced way, regardless of any expectations or MS' track record or the shrill innuendos of some of MS' competitors. (Alex of course is not speaking for SC34 in this blog.) And SC34 (and Alex) will continue to do so, I expect, just as it will continue to also treat OASIS in a professional and encouraging way, even though in *both* cases this may sometimes involve negotiating around different expectations about maintenance responsibilities and technical content, that may look like weaseling out of commitments to some, for example.

    You seem to want SC34 to judge incoming standards on something else than technical content, but you know that isn't the way it is allowed to work. (You may reasonably disagree on the specifics of OOXML's technical content, of course.)

    On your comment that no-one else has implemented OOXML, why don't you try OpenOffice? This is an open source product that claims to have OOXML import.

    Alex: A standard can certainly be withdrawn if has been superceded by a new standard: that the old one has some market traction has never been enough to prevent new versions of standards. In the case of OOXML, it is just the withdrawal of a *part* that this is explicitly marked as not being suitable for new systems. (Though perhaps some NBs might prefer making Part 4 entirely non-normative or transfering it to a Technical Specification, rather than being entirely removed.)

  • Rob Weir

    4/2/2010 12:20:56 AM |

    Rick, I would have been very pleased if you, Alex and several others NB experts in SC34 had judged OOXML purely on its technical merits.  But you didn't.  You fell into the false promises from Microsoft, that if you held your nose on this one, and approved all the legacy crap, that if you averted your eyes just this once and opened the doors, that the standard would be put under international control and evolved in an open, transparent and vendor neutral way.  But that isn't quite what happened.  They played you for Useful Idiots.  At least you, unlike Alex, got paid for your services.

  • Rick Jelliffe

    4/2/2010 1:07:08 AM |

    Rob: What a vivid imagination.

    IS 29500 mark 2 was approved by a large majority of National Bodies, not 'me, Alex and several other NB experts.'

    If Microsoft was pulling all the strings, how come they didn't get the standard they wanted first time? Or even the second time: the standard they got, with the 'legacy crap' removed to the Transitional part and marked with a big sign (that Alex mentions above), is one MS are still pushing against, it seems.

    Your claim that SC34 or experts should have one rule for Microsoft and one for everyone else doesn't wash. It is unprofessional. But you know that.  

    (I hope no reader understands Rob's comment to mean that I was paid by MS for any lobbying activity at SC34 or Standards Australia or my blog, then or now, or to alter my opinions. I wasn't. Sticks and stones...)

  • Frank Warren

    4/2/2010 1:21:27 AM |

    Dear Mr Weir and Mr Jelliffe,

    My name is Frank Warren

    en.wikipedia.org/wiki/Frank_Warren_(promoter)

    Since the Joe Calzaghe fiasco, I need a big ticket event to get me back on track.

    Do you have any interest in being involved in a bout in Las Vegas later this year?

    Going by your long history of trash talk, it should be a mouth-watering prospect.

    Frank

  • Alex

    4/2/2010 1:25:43 AM |

    @Rob

    In many NBs the approval (or not) of OOXML was a policy decision, but even on technical grounds many NBs found the defects they had identified were satisfactorily addressed in the Fast Track process. Get over it.

    (ODF was, of course, the ultimate "approved for policy reasons" JTC 1 Standard -- since few, if any, NBs bothered even to glance at the text before voting for it!)

    OOXML does not have problem being "evolved in an open, transparent and vendor neutral way" (except of course, in the PR narratives of its enemies) -- the problem I am concerned with is the rate of evolution.

    And - sigh - you continue the ugly pattern of behaviour demonizing independent experts whose views don't sit well with the IBM line. Rick has always been completely open, honest and up-front about his activities and is known and respected (among standardizers and FOSS developers anyway) as both a FOSS and Standards champion.

    I don't know - I used the phrase "lack of executive oversight" about Microsoft in my post. When you get into your attack mode you are an in-person exemplification of the fact that IBM has the same kind of problem! -- and it's a shame, because in person you're a gent ...

  • Rui Miguel Silva Seabra

    4/2/2010 2:14:34 AM |

    «If Microsoft was pulling all the strings, how come they didn't get the standard they wanted first time? Or even the second time: the standard they got, with the 'legacy crap' removed to the Transitional part and marked with a big sign (that Alex mentions above), is one MS are still pushing against, it seems. »

    Oh, it's very easy to explain that. Here in Portugal they almost succeeded to get almost unanimous approval at first time. They stuffed a committe in a 7 against 1 ratio.

    When that 1 asked to enlarge the committee, they kept stuffing and eventually the committee refused SUN and IBM for «lack of chairs» which was a blatant and convenient lie. There were multiple people per Microsoft friendly entity and there was a much bigger room in the building.

    The number one reason they didn't get it as they wanted it the first time, is that they found much more resistance than what they be expected.

    They expected that just stuffing committees with friends would be enough to get a rubber stamp approval.

  • Jeremy Allison

    4/2/2010 2:21:46 AM |

    Mr. Brown,

    The outcome that many had predicted, yet you insisted would not occur, has now come to pass. Better late to the party than never, I suppose.

    I'm a little sad it took a lack of Microsoft following through on their promises (so easily given) for you to get to this realization. Had you listened to others who have had much more experience than yourself in dealing with Microsoft, maybe you wouldn't now find yourself in this somewhat humiliating position.

    What do you think should be done about this sorry state of affairs now ?

    Jeremy Allison.

  • orcmid

    4/2/2010 2:27:53 AM |

    I don't know what the reasoning was with regard to the pragmatic impacts, but my impression is that there is an amendment to IS 29500 somewhere in the works that will change the namespace used for entirely-strict OOXML documents versus thost that bear transitional features.

    While, technically, I think this is a good idea, I don't see how this is something Microsoft has had 2 years to deal with.  I'm fairly confident that this was not something that Microsoft perpetrated on itself via its alleged dominant control over SC34 WG4.  

    I would hope that whenever Microsoft is able to recognize strict arrivals and optionally produce strict outputs for transitional arrivals, the juggling of namespaces will also work out.

    I think caution is called for here.  I'm not that confident that the massive changes made to accomplish the separation of strict and transitional into separately-namespaced definitions has been achieved without new defects to deal with.  I just worked over a much smaller set of errata proposals for defects against ODF 1.0/IS 26300/ODF 1.1 and the mishap rate is surprisingly high.  Excruciating care is required and we don't seem to live in an era that has much tolerance for what that takes.

    Going forward, I think the establishment of dual namespaces will be beneficial, but the transition in products is non-trivial (though maybe easier now).  I would also expect there to be support for transitional at least as long as there is product support for "Save as ... 97-2003 Document" and for the same customer-serving purpose.  To consider otherwise strikes me as absurdly disconnected from reality.

    Expecting a way for a community of users to establish "Save as strict" as an option or even a configuration default or at least the default for updating an already-strict document is a different story.  I expect that this sort of thing will have to be accomplished carefully and delicately with adjustment over time as the community of users and further product releases/updates adjust around that objective.  I am not going to second-guess how Microsoft, or the communities that consider only strict as acceptable, will work their way to some accommodation en route to International Standard Valhalla.

  • Stephen Walli

    4/2/2010 2:38:49 AM |

    I suspect I know what is "wrong at the centre" with respect to implementation problems and the like.  Blogged it here: http://bit.ly/9G1oty  

  • Alex

    4/2/2010 2:43:03 AM |

    @Mr Allison

    > The outcome that many had predicted, yet you
    > insisted would not occur

    Oh? I don't recall making predictions about Microsoft's behaviour? URL please!

    > Had you listened to others who have had much
    > more experience than yourself in dealing with Microsoft

    What, you mean get some off-the-peg set of prejudices? I've noticed that's how some people prefer to proceed.

    Don't you think corporations change? Google from wide-eyed startup to the new Big Brother megacorp; Sun from centre of the technical solar system to bin-end bargain; IBM from evil monopolist market-abuser to ... no, wait ... Wink

    Standardizers should be skeptical of corporations, but even within corporations there are many different voices. I think the view that reduces corporate disputes to some kind of soap opera with "goodies" and "baddies" is reductive and unintelligent. And I think the view that holds every corporation to continual account for its worst past misdemeanor is an impediment to any kind of progress ...

  • Jeremy Allison

    4/2/2010 2:57:36 AM |

    Not some off-the-peg prejudices. 18 years of experience with them, dealing with compliance on network interoperability I'm afraid Smile.

    My opinions on Microsoft are actually more nuanced than you might think. After dealing with them for so long my view of Microsoft is rather similar to Churchill's view of the Americans : "Americans can always be counted on to do the right thing...after they have exhausted all other possibilities."

    I'm truly interested in what you think should be done about this.

    Jeremy Allison.

  • Gareth Horton

    4/2/2010 3:03:46 AM |

    @Jeremy

    As a member of WG4, I don't in any way see this as an "outcome". That implies a finality, whereas standards maintenance is a continuum (until they get stabilized/withdrawn).

    In Alex's view, as a long time standards participant, the progress and modus operandi in WG4 seems to have worked counter to his experiences in other SC34 working groups.

    As a relative newcomer to the standards work, it's difficult to know what the benchmark is.  I have worked with company internal people that have been less responsive and committed than the Microsoft people involved in WG4, so the problem is less clear cut than many are making it out to be.

    It's easy to take the opportunity to use Alex's post as a gloatfest - fine, enjoy your moment of smugness.  

    What I think is valuable from Alex's post is really trying to pinpoint the issues on this anniversary. As far as I am concerned it is a lack of resources, both from the ECMA/Microsoft side and the implementer side.  

    There are very few technical experts involved from implementers that can help fix and improve the standard, as well as provide more balance within the working group.

    The reasons for the known failures in validation are simple bugs, one in the text of the spec and one in the Office implementation (IMHO).  I know, as the testing I was doing against Alex's validator is the source of them. It is not, as far as I can see, due to abject apathy on the part of Microsoft.

    I assume Samba has or has had bugs at some stage in the development process right?

    So, the upshot of this is that Microsoft will be reminded of their commitments to the process and adjust their budgets accordingly, if they are indeed committed to the process.

    I strongly feel that greater non-Microsoft participation is essential, both from the technical input perspective, as well as any political dynamics.

    If people and organizations feel strongly about improving the standard and holding Microsoft to account, then they should participate.  Jeering from the sidelines is not going to change anything.  Dealing with ECMA/Microsoft as a member of a National Body with a vote is far more powerful platform.  

    Alex's post as a key participant of the process holds more weight than any moaning from Groklaw and the other NOOOXML'ers and might actually help improve matters.

    Gareth

  • Gareth Horton

    4/2/2010 3:08:56 AM |

    @orcmid +1 Dennis, as usual.

    @jeremy working well with them on Excel compatibility since 1991, so just a little less than your 18 years.  Maybe they just don't like you Wink

  • Jim Smithers

    4/2/2010 3:26:03 AM |

    "And - sigh - you continue the ugly pattern of behaviour demonizing independent experts whose views don't sit well with the IBM line. Rick has always been completely open, honest and up-front about his activities and is known and respected (among standardizers and FOSS developers anyway) as both a FOSS and Standards champion.

    Read more: www.adjb.net/.../...-Fails-the-Standards-Test.aspx;

    As a avid open source user, I know this for the smell it emits. Your and Mr Jelliffe name sank below the bottom of the barrel for your part in the ooxml saga. This article was a small step back up the ladder. But the comments in reply to it still show you and Mr Jelliffe are still Microsoft fools.

  • Alan Bell

    4/2/2010 3:31:40 AM |

    Well as the date on the article is not the 1st of April I guess this is serious. I can well imagine the frustration of all the people working on the committees to try to make this whole thing work for the benefit of everyone. It seems that the conformance is close enough for marketing purposes and that is where it is destined to stay. How is their ODF support these days? I seem to recall predictions that Microsoft with their impressive development resources might end up making a stricter more accurate implementation of ODF than OpenOffice.org or anyone else.

  • orcmid

    4/2/2010 3:40:46 AM |

    @Stephen Walli - +1 for sure.  I missed that the first time.  The historical sleuthing around Unix/POSIX/Linux, how Microsoft was blind-sided, and the Java missteps is wonderful.  

    [The only disconnect for me is that I saw OSF as an IBM-DEC Play against Sun dominance and a beautiful move regarding AIX.  Some DEC Hardware may have run a lot of Unix implementations, but I didn't know many folks who knew the name of DEC's version.  I knew people happily running (close-enough) Unix on Sun that resented IBM's disturbing arrival.  At the OSF announcement event I attended, there were many unhappy-faced old-guard IBM reps in the back of the room, too.]

  • Gareth Horton

    4/2/2010 3:41:26 AM |

    Alex,

    Is "Microsoft fool" a new one?  You need to let Jesper know and update the scores.

    @Jim Smithers - partying like it's 2007 with good old Roy Schestowitz I see.  You might want to check your open source software and rip out anything Rick might be responsible for - make sure your principles remain intact.

    Gareth

  • Alex

    4/2/2010 4:01:26 AM |

    [I've just deleted a comment. Jibes mentioning "thalidomide" cross the line.]

  • Jomar Silva

    4/2/2010 4:08:00 AM |

    Alex,

    Good to see that you're now seeing things that thousands had seen at that time... again, good to see that we're starting to agree more.

    As I explained to you in Stocholm, BR is very concerned with the transitional-to-forever OOXML, and this is a joke that isn't fun anymore.

    Worst than this, we now see WG4 reverting most of the decisions taken at the BRM regarding the Transitional... the ISO 8601 removal is the latest trick on that.

    BTW, at that time I've had a good conversation with a friend that folowed all the development of the video and multimedia standards, and he told me several histories about Microsoft behavour in TCs... some of those histories are been reenacted right now.

    I honestly don't understand why we need to still spend money and resources on a standard that was only sent to ISO to have the "status" as someone already comented... They still playing with us !

  • Alex

    4/2/2010 4:21:00 AM |

    @Jomar

    The odd thing about the ISO 8601 reversion is that it is really nothing to do with Microsoft's wishes (in fact, in some senses, it is quite the opposite).

    Check out Gareth's post here:

    aristippus303.wordpress.com/.../

    The push to revert this comes from the National Bodies who are concerned their users will get screwed by silent data loss if T files start to appear with ISO 8601 dates.

    Microsoft had already written code in an Excel pre-release to use ISO 8601 dates. See Jesper's screen-shots at:

    http://is.gd/baaQJ

    So, in fact, this was an argument where MS were broadly pro 8601, and the NBs were (broadly) saying no -- hence the upcoming proposed amendment to remove ISO 8601 dates from T (not S), which I hope Brazil will support for the safety of Excel users everywhere.

    This is a real-world case which proves things are rather more complex than in Rob's fantasy-land of a working-group controlled by Microsoft!

  • bob

    4/2/2010 4:25:26 AM |

    All of this high minded talk is nice, but it misses the key point: money. MS Office is worth a heck of a lot of MS's profits, because they can charge more having a close to a monopoly on it. If they go to a strict standard defined file format they open themselves up to competition and will have to cut prices, at a minimum. And if competitors are multi-platform customers may stray from the other big profit center: the Windows OS.

    MS has always been brilliant at marketing. This allows them to claim the marketing plus of standards compliance without suffering from real standards and the competition that would ensue.

  • Stephen Walli

    4/2/2010 4:45:54 AM |

    @Gareth:  You said There are very few technical experts involved from implementers that can help fix and improve the standard, as well as provide more balance within the working group.  This is the real problem.  A standard with only one serious conforming implementation isn't a standard, especially when they have their own product difficulties with which to contend.  Because of the way Microsoft played the early marketing game with claims of all the others "supporting" the standard rather than actual conformance, I'm betting there is little reason for the likes of Apple to continue to play on.  Without real implementer participation it will be very difficult to get the standard on track.  (I grumbled about this two years ago:  http://bit.ly/abtyOY

  • W^L+

    4/2/2010 4:57:08 AM |

    You're dealing with a corporation, one that derives much of its profit from the difficulty of switching out their office products for competing products. You should not ever expect serious standards compliance until they face a loss of market share.

    If OOXML-Strict compliance is important to you, help StarOffice / OpenOffice or WordPerfect or iWork to produce more-compliant documents and then promote those products instead of Microsoft's products. If you succeed, the loss of market share will cause them to pay attention. (As a side note, the effort spent improving the products' OOXML compliance may also benefit the products' ODF compliance, which is what I personally care about.)

    Don't pretend that you did not expect this result. No one who watched the standards battle would believe you are that naive. Instead, focus on how to move forward from here.

    Finally, Rob and Rick: It is possible to disagree about nearly everything and still maintain personal respect. You both have a lot to offer here. The way forward is not "us" versus "them", but a much larger "us" finding areas of agreement. Can we start by acknowledging that much?

  • orcmid

    4/2/2010 5:04:25 AM |

    @Alex

    "The initial version of OOXML (Ecma 376 1st Edition) was rejected by ISO and IEC members in September 2007 ..."

    For me, there's something too strenuous in the emphasis on rejection. It is clear that the the submission was not accepted, but it was also in the ball-park for ballot resolution.

    Not being there, I don't quite know how to temper this but to observe that the BRM made changes that were acceptable enough to the objecting parties.

    To suggest that OOXML was strongly rejected seems too much like the claims by some parties here that the American people strongly and resoundingly reject the Health Care Reform legislation that was finally passed here.

    Having said that, I do recognize your concern for the pace of evolution toward a strict-OOXML world and even a time when there is a strict-OOXML opportunity.  The lack of a roadmap is not helpful.  On the other hand, maybe what's required is a little more sharpness about "support" in the Microsoft warrant that includes:

    3.3(c)
    (iii)  Microsoft shall Support the ECMA 376 Specification in the .docx, .xlsx and .pptx file formats used by Word 2007, Excel 2007, and PowerPoint 2007;
    (iv)  Microsoft shall Support IS 29500 in the .docx, .xlsx and .pptx file formats used by successor versions of Word 2007, Excel 2007, and PowerPoint 2007;

    We could save up our allowances and paper-route money and maybe purchase one of those warranties with a rider about strict-versus-transitional?

  • cheve

    4/2/2010 5:09:19 AM |

    Dear Mr. Brown :

    I find it 'very' interesting where you state that "..In many NBs the approval (or not) of OOXML was a policy decision, but even on technical grounds many NBs found the defects they had identified were satisfactorily addressed in the Fast Track process..."

    If I recall(based on reports) correctly, people were running out of the time to discuss the tech issues(and that there were many) -- A large portion of text were not even look at due to the lack of time.  So your stating that "on technical grounds many NBs found...were satisfactorily addressed" is, somewhat ishonest IMNSHO.





  • Rob Weir

    4/2/2010 5:25:03 AM |

    Re-reading Alex's post I think he misplaces the blame, at least in part.  The fact that maintenance has not been staffed to his satisfaction -- this is not a Microsoft issue.  It is not an Ecma issue either.  Neither of them are assigned maintenance of ISO/IEC 29500.  That was assigned by JTC1 to SC34.  Alex described how control had been handed over back in 2008:

    "JTC 1 have handed full responsibility for the standard over to SC 34"

    http://adjb.net/index.php?entry=entry080409-221633

    Is there some part of "full responsibility" for 8000 pages of crud that makes you a bit uneasy now, Alex?

    More from that post: "The passage of ISO/IEC 29500 has instituted a new era of standards activity in SC 34 related to document formats".

    Indeed.

    "SC 34 has a plan: it envisages taking control of OOXML"

    A cunning plan, my lord.  But as I said before, it is more like Microsoft took control of SC34.

    It seems to me that if the committee were not dominated by Microsoft (which I believe it is, but Alex denies) then there should be no problem staffing the work adequately without relying on Microsoft.  Maybe Alex can explain how Microsoft is now to blame for the slow progress of OOXML maintenance in a committee that has "full responsibility for the standard" and which had broad participation from non-Microsoft experts?  Does SC34 have responsibility or not? Does it have broad and active participation from non-Microsoft employees or not?  

    And if you cannot handle the work for projects already assigned to SC34, then why are you proposing new work items like an ISO ZIP format?

    I can appreciate that you think progress here is slow, but the blame appears to be entirely misplaced.  OOXML is your albatross now.

  • Alex

    4/2/2010 5:25:20 AM |

    @Dennis

    On "rejected" v "not accepted" - if the text had remained unchanged, the initial voting result would have stood (in that case no vote modification would have been permitted). So in terms of JTC 1 rules I'm comfortable with "rejected" being the right word for NB opinions on the Ecma 376 text as it was in September 2007.

    I completely agree that a "roadmap" would be of immense value to moving things forward!

  • Alex

    4/2/2010 5:27:26 AM |

    @cheve

    You have your "reports"; I have my first-hand knowledge and second-hand accounts from the people involved.

    If an NB felt there had been insufficient time to review the text, it was open to it to vote no. Some (e.g. India as I recall) did just that for this reason ...

  • pan

    4/2/2010 6:07:51 AM |

    The MSFT standards campaign was simply window dressing to prevent uppity governments from rejecting the Window DOC[x] monopoly as non-standards compliant.

    When the immediate danger that some activist government would emmulate Massachusettes, USA and similar municipalities, Microsoft dropped any concern for "standards".

    The whole Doc format standard campaign was simply meant to avoid rejection of the proprietary format.  It worked, the political will to punish MSFT for trying to patent text authorship has passed into a vague, and confusing bramble of standards.

    The unfortunate fellow travelers with Microsoft are now waking up to how simply they were co-opted by a monopolistic criminal organization.

    Sorry guys, but you got used.

  • TheOpenSourcerer

    4/2/2010 6:58:27 AM |

    I note with interest the (so far at least) silence from Microsoft.

    Alex, I am surprised, after all the allegations of stuffing, bribing and coercion of NB's that were, until DIS29500, uninterested in XML document standards, that you seem surprised at the outcome thus far.

  • eduardo

    4/2/2010 9:45:33 AM |

    Microsoft, being a corporation, does what it believes will earn it the most profits. For decades this has been maintaining a proprietary document standard. Apparently those with the authority to decide believe this is still the case.  If at some point in time it decides it would make more money converting to a standard it will do so.

    Your error, Alex, was to fail to look at Microsoft's promises from this perspective.  Remember, past history (i.e Kerebos, the Java contract) makes clear that Microsoft will lie quite convincingly if it believes it is in its interest to do so (or do you think I am wrong about this?)  It would be helpful if you described what Microsoft said to persuade you of its good intentions, and why you believed it to be telling you the truth.

  • Alex

    4/2/2010 5:14:13 PM |

    @Alan

    Well, a simple "Hello World" ODF file passes the Office-o-tron with no conformance errors, which is more than can be said for the same document from the latest OpenOffice.org ... so, watch this space in anticipation of more meaningful testing!

  • Alex

    4/2/2010 5:31:33 PM |

    @Rob

    What a strange comment.

    You need to read that blog post of mine a bit more - especially the section entitled "Bringing Ecma In".

    As you well know, Standards organisations ultimately rely on contributions from individuals and corporations to progress technical work. It is not a tap they can turn on and off. The fact that Microsoft provides resources through Ecma is to be welcomed. The fact that any corporations provide resources to standards bodies is to be welcomed - bravo I say! But I am greedy: I want more.

    In International standards committees (like SC 34) liaisons and corporations members have no power of decision-making: the framework is there to handle this well.

    The view (from certain corporations) that certain other corporations should not make contributions to standards activity is, in my view, very short-sighted and ultimately anti-standards.

    If WG 4 is not sufficiently resourced then we have a problem. We might find ourselves getting into a situation where - I don't know - maybe it would take us 5 years to get a modest update out of the door. Imagine that!

    As to ISO ZIP -- that would not be a WG 4 project and so the concerns of my blog don't really apply. If it can't be resourced it won't be done. However, since it's likely to be only 20-40 pages, and since - informally - I've already heard offers of help from several sources, I don't think that will be a problem. Perhaps an IBM guy might like to contribute? If so, I would again say bravo!

  • Alex

    4/2/2010 5:53:23 PM |

    @eduardo @pan @Sourcerer

    I think there's a misconception at large that NBs voted in favour of OOXML because they somehow "trusted Microsoft". As far as I'm aware this is a long way from being true. If anything, some positive votes may have been motivated more by a distrust of Microsoft!

    Remember that the maintenance regime of OOXML took decision-making away from Ecma/Microsoft and placed it in the hands of the National Bodies. The international community controls the OOXML standard (beware spin from MS competitors saying otherwise: they fear of course the same wrenching of control may happen to them), but the international community does not "pick winners" or necessarily even concern itself over-much with what Microsoft does. Those concerns are more properly dealy with by market regulators, not standardizers.

    For myself, I have always advocated taking a skeptical view to corporations and putting more faith in conformance testing. See for example what I wrote here before about interoperability testing:

    When it come to Office Document interoperability we do not need to rely on bitter blog exchanges, the warm words of press releases, or even on the success (or not) of workshops in Redmond. Questions of interoperability will be decided by the cold hard fact that certain bytes arranged in certain ways will betoken good behaviour; other arrangements will betoken bad behaviour. We, the users, can measure which is which, and we, the users, can improve the tests by making the standards that govern office document more thorough. If we deal honestly and standardise well, the optimal outcome for us is within reach.

    - www.adjb.net/.../Office-Document-Interop.aspx

  • A. Rebentisch

    4/2/2010 6:24:26 PM |

    Whatever the standard says, the simple task of office applications and office productivity tools out there is to ensure compatibility with the original format. In this sense it seemed important in the process to get certain deliverables and ensure the white angel gets taped when it is butchered by the dark knight. Despite Sohne in Kenya of course no one physically died.

  • Rob Weir

    4/2/2010 8:28:45 PM |

    Alex, I think you overestimate the value of the ISO imprimatur. I cannot think of any large corporation that would consider it a fair trade to hand over control of a critical part of their technology to a cabal of petty consultants, especially ones with such a poor reputation. Anyone who thinks the contrary is a fool.  No one is going to hand over control of their technology to a bunch of loose cannons that run alternatively hot and cold, who make up the rules as they go, that whine and moan and run to the press whenever they feel that their dignity as International Grand Poobahs is impugned.

    You seem to want to play ODF and OOXML off each other, in hopes that you can force concessions, hoping that continual threats against both standards will ensure that the midget in the middle can control both.  That might have worked before, but we're all on to your little game.  Remember, ISO approval is worth only a small amount of inconvenience.  And there always remains the possibility of both ODF and OOXML removing themselves from ISO, at which point SC34, as well as you yourself, would just descend back into its original and justifiable obscurity.  

    So whenever you start making demands on a large international corporation, I suggest you think more like a salesman, and tell them why this is to their benefit, what value they would get, why it is worth their trouble.  If the best you can do is to lead with idle threats about withdrawing their standard, then you are making a weak argument.  This hooliganism is also causing S34's future pipeline to dry up.  Microsoft was once going to send along XPS from Ecma for Fast Track processing in JTC1.  They backed off that, no doubt evaluating the cost and the benefit and deciding that working with SC34 on this is simply not worth the trouble.  I cannot find fault with that calculus.  

  • Laszlo Kurti

    4/2/2010 8:46:43 PM |

    Oh Microsoft really don't give a s...t for there own promises? What a surprise! Oh common. Considering the circumstances of OOXML voting it was more than obvious that MS want the ISO standard for marketing reasons only. MS feared of the growing impact of OO.o (ODF) and wanted there own standard in any price. If the OOXML would have been "standard ready" at that time than why were many strange activities within different National Bodies such as in Italy, Sweden, Hungary, not to mention many African countries which are never been near by ISO. Sorry to say but the story so far is if you have enough money than you can buy your own standard, from ISO as well.
    Did you really believe that MS just changed? You are really surprised that MS just left ISO (you) alone? Hey wake up, it's all about money, as the OOXML prove it clearly.

  • Olivier

    4/2/2010 11:38:00 PM |

    This situation needs fixing. An open standard with no reference open implementation is a problem. An "open" standard with not even one proprietary implementation from the standard's main promoter is a disgrace and a joke.

    Microsoft abused the ISO process, and the ISO committees let themselves be bamboozled.

    The lesson are:
    - MS is not to be trusted,
    - ISO is weak.

    The actions needed are:
    - repeal of the OOXML standard, or at least the transitional part, so that MS cannot benefit from its lies
    - perfect treatment of the remaining "true" OOXML format, including adamantly requiring a very clear spec and one open-source implementation.
    - official reprimand to MS
    - internal ISO restructuring to make any further such incidents impossible.

    Anything less would be still more hypocritical kow-towing to the lying bully.

  • foo

    4/2/2010 11:42:19 PM |

    "And if we look elsewhere within Microsoft we can see – for example from their engagement with HTML 5 and work on MSIE – that they can move in the right direction when the will is there."

    You mean... like the 55/100 result in the ACID test Microsoft proudly displays on their web site? Link:

    ie.microsoft.com/.../Default.html

    No, Microsoft is not interested in standards. They want to control the platform, and let everyone play catch up.

  • Olivier

    4/2/2010 11:50:04 PM |

    Regarding the "how come MS did not get the standard they wanted ?"... This question shows how naive you are.

    MS did not want a specific standard. They wanted anything that vaguely looked like a standard so that their salespeople could counter ODF. Actually they're very happy with the standard they got, with its 2 versions, one of them un-implementable by anyone else and the other un-implemented even by themselves.

    That way the "open" check box on governments' and Big Corps' purchase-approval checklist are ticked, but they remain locked-in to MS's proprietary formats. Best of both world.

  • Alex

    4/2/2010 11:59:48 PM |

    @foo [not your real name]

    I did say "move in the right direction" - and MSIE is a Hell of a lot better than it used to be ... (MSIE 6 anyone?)

  • Sebastian Gomez

    4/3/2010 12:13:00 AM |

    Many previous posts said it was unrealistic to expect microsoft to implement proper support in Office 2010. I think what is unrealistic is expecting microsoft to implement any kind of standards.

    The only time they will implement anything that is standards compliant is when they have no choice. Think about IE. It took 15 years to get them to implement standards in IE (In IE9) and they only did so because Mozilla, Apple, Opera and Google forced them. Only after they lost significant marketshare against this companies that they implemented HTML5. And, remember, embrace, extend, extinguish. IE9 is only phase1 (Embrace). In a year or so, we'll see IE9 marketshare grow, and the proprietary extensions will start rolling. In a few years, It'll be 2001 all over again. IE15 will be as incompatible as IE6 was.

    This is microsoft. That's what they do. They won't change. They are the most hostile company I've ever seen. They blatantly attack the rest of the industry, and as long as people put up with it and buy their products, they have no reason to change their tactics. They've worked well for them for almost 3 decades.

  • Common Sense

    4/3/2010 12:13:12 AM |

    microsoft ditching open in favor of proprietary.... only fools expected otherwise. ISO got bent over like countless techies said they would. end of story. ISO now supports closed proprietary formats as "open".

  • Alex

    4/3/2010 12:24:07 AM |

    @Sebastian

    I'm not sure who was "expecting" Microsoft to implement OOXML Strict in Office 2010; though certainly many hoped they would. However what MS does (or does not do) is not the business of the standards organisations.


    @Olivier

    OOXML Transitional unimplementable? I think not. If you're interested in office suites then  OpenOffice.org will read it, and check out SoftMaker Office (http://www.softmaker.com/) - I know I will soon. (And of course these formats are used in the tool chains of countless back-office and enterprise systems the world over).


    @Common Sense

    This isn't an open/proprietary issue, it's an issue about standards variants and conformance.

  • Olivier

    4/3/2010 12:47:17 AM |

    "OOXML Transitional unimplementable? ...I think not...  OpenOffice.org will read it, and ... SoftMaker Office... (And ...tool chains...)"

    OO.org, plenty of competing Office software and other apps also import "regular" MS formats. Does that mean they are standards ? The point of an official standard is not to be able to kinda read/write a subset of a format for kinda OK results.

    A standard is a complete and unambiguous spec, with a reference (open if possible) implementation, more is better. Transitional is not a standard, it's much too vague for that. Yes, you can import and manipulate a subset of it. No, it does not guarantee interoperability. Nowhere near, actually.

  • Alex

    4/3/2010 12:59:27 AM |

    @Olivier

    Well, you said T was unimplementable - now you're on to a different topic.

    In JTC 1 there are no reference implementations, instead the emphasis is on the text of the Standard being entire unto itself.

    I think that document format standardisation is a fairly immature discipline and this is reflected in the standards themselves and in the level of interoperability between implementations of them. It's an interesting question whether sending out a .docx file (say) will result in greater fidelity when opened with MS Office/SoftMaker Office/OO.o as an equivalent ODF file would opened with OO.o/KOffice/MS Office.

  • Sebastian Gomez

    4/3/2010 1:05:11 AM |

    @Alex

    You are suggesting people to checkout other privative software like softmaker? All that'll do is perpetuate this problem.

    Emacs is all you need people. If you want formatting, use TEX. d

  • makomk

    4/3/2010 1:10:01 AM |

    Rick Jelliffe: Open Office doesn't actually support the OOXML ISO standard, or at least didn't last time I checked. Instead it supports the OOXML that Microsoft Office actually produces and reads. The developers concluded that support for the ISO standard was useless since nothing was likely to actually read or write it, and it'd just confuse people expecting to be able to exchange documents with Microsoft Office.

    All the non-Microsoft applications that might add ISO OOXML support are far better served by ODF already.

  • Alex

    4/3/2010 1:14:20 AM |

    @Sebastian

    I don't see why - I asssume you mean - commercial software will perpetuate any 'problem'.

    I somehow can't see the proposition of shifting a workforce of office document users over to TeX being particularly easy to follow-through on!

  • Olivier

    4/3/2010 1:23:01 AM |

    I'm trying to make the point that some level of interoperability ( your "OO.o and others open OOXML") does not a standard make. Hence my counter example of other formats OO.O and others can work with, which are NOT standards. A standard has to ensure almost perfect interoperability; When, as in Transitional's case, the specs are at best unclear, and on top of the there are 3rd party (let alone open) implemetations, there is a problem. Strict's case is even worse, with neither implementation nor much confidence in the specs.

    As for you cop-out that other standards are screwed, too... I can't politely react to that. I guess I'm gonna go steal some money now, but less than others, so that's OK...

  • Freedom

    4/3/2010 1:25:28 AM |

    Mr. Brown, you are a tool.  This is exactly the behavior one should expect from a convicted, but not punished, monopolist.  You and the ISO have been had, and have given Microsoft no reason to change their abusive tactics.

  • Ian Lynch

    4/3/2010 1:27:20 AM |

    MSFT has a commercial interest in holding out against interoperability as far as they credibly can and for as long as they can. They need to buy time because their desktop monopoly is going to die a death of 1000 cuts with migration to the web rather than competing desktop technologies. Problem is that it is difficult to see how they could replace the Windows/Office combination as a cash cow but everything has its day and change is inevitable. Really we should have one ISO document standard and that should be odf.

  • Alex

    4/3/2010 1:35:54 AM |

    @Olivier

    > A standard has to ensure almost perfect interoperability

    Yup - but dream on. ISO/IEC 26300 (ODF 1.0) has no spreadsheet formulas for example! As I say, immature standards.


    @Freedom

    ISO makes standards; it does not regulate markets. You confuse these activities.


    @Ian

    Why do I have this sudden sense of déjà vu? Wink

  • Doug Mahugh

    4/3/2010 1:39:47 AM |

    Hi Alex and others. Interesting post, and interesting discussion.

    I’d like to respond in some detail, and will do so in a blog post early next week.  Have a great holiday weekend everyone, and I look forward to your feedback next week.

  • Olivier

    4/3/2010 1:45:34 AM |

    At least ODF has an open source reference implementation, which is the second-best thing.

    And, once again, you're using others' issues as a cop out. Those "morals" are extremely easy to live by. I feel I'm hearing my kid explain that his E is pretty good, because someone else got an F.

    And by the way, that's an F you got.

  • Jamie

    4/3/2010 1:51:07 AM |

    For many who followed this from the beginning, I find it rather amusing you seem surprised by the current situation Alex.

    I've never previously followed an ISO standard, so perhaps all the ballot stuffing and irregularities are common; personally I found them shocking, and even more so how people involved such as yourself seemed to simply turn a blind eye to the whole mess.

  • Ian Lynch

    4/3/2010 1:52:22 AM |

    @Alex

    There will be a lot of deja vu simply because the issues are not going to go away and they will be constantly revisited. In the end the trend is to open standards and interoperability, particularly in document and web related formats. The discussion is really in how long legacy de facto standards and their control by particular interests can be maintained. Personally I'm finding I'm using WP a lot less and publishing direct to web pages, sharing spreadsheets on Google Docs etc. We encourage schools to teach more of this to prepare their learners for the future.

  • Peter

    4/3/2010 2:00:55 AM |

    Forgive my ignorance, but does this mean that OOXML's status as a recognized ISO standard could be withdrawn?

  • Rick Jelliffe

    4/3/2010 2:10:35 AM |

    Jeremy: It seemed to me that your position against a standard for OOXML at ISO was to some degree at odds with the struggles in the rest of your SAMBA career to get better more complete documentation, better IP disentangling, and to get competitors or other stakeholders at the table discussing technical details and how to get the rough edges and legacy kludges cleaned up.

    For years, we in the document industry have been demanding the same kinds of things, but in relation to Office file formats. When the EU made its call for MS to use XML and submit their formats for standardization, were we really supposed to say "Oh, after calling for this for all these years, we don't really want it or need it?"

    If MS had come to the SAMBA people 10 years ago and voluntarily submitted to internationally QA'ed open scrutiny on their documentation and its completeness, devoting substantial resources to put out so much documentation (remembering that OOXML went from around 2000 pages to around 8000 pages in response to requests for information by standards bodies) that people even began to complain it was too big, would you really have turned around and said "Oh, no, we didn't mean it, we would prefer the status quo where we have to be satisfied with whatever partial crumbs that they post on their website!"?

    I suspect most people would have said "Lets raid the chicken
    coop while the gate is open!"... Lets get that documentation vetted and published and the parties into some kind of dialog.

    Corporations always blow hot and cold with standards, and shop around for the most suitable standards bodies. Take Google's Ian Hickson's efforts with HTML5 and the WHATWG for example. Indeed, XML itself is the poster child for this, when Sun's Jon Bosak took the development of the SGML profile from ISO to W3C, taking many of us with him. I came back to SC34 by invitation, largely after becoming disenchanted with what I saw as domination by two large corporations (not Google or Microsoft or IBM sort of) in a particular W3C standard, in effect denying a whole set of use cases that didn't fit in with these companies technologies.

    A necessary but not sufficient tool for dismantling the grip of a monopoly or oligopoly technologies is to have a viable FOSS alternative, which is what SC34 did when making alternative schema languages available. It is not up to experts to decide that ODF is better than OOXML or OOXML is better than ODF; it is the market and time and procurers. We are better off with competition and real alternatives, not monocultures. We were better off having both TCP/IP and OSI rather than just one (imagine if it was OSI!)

    Many of us, I think, have had confidence that ODF would progressively be capable of providing that alternative, and that the ODF/OOXML kerfuffle was, as Alex memorable put it, a phoney war. I have every confidence that ODF will continue to thrive, and that it will completely displace DOC/RTF/DOCX for simple editable public documents; but this will happen not because ODF was standardized nor held up because OOXML has been also standardized, but when market decides that implementations have good enough interoperability (which *then* relies, in turn, on the base standard being good enough.)  [[Indeed, ISO SGML was mandated for DoD use in the USA but never successfully transitioned from aircraft manuals to the desktop because the applications were not available.]] All standardization does in this case is to help the documentation for both to get better QA, and to help some of the wrinkles in feature sets to get ironed out: they have to duke it out in real implementations.

    So the fact that Microsoft would sooner or later blow cold (if indeed it has, at least in its conformance for OOXML, which remains to be seen) is no surprise or mystery or embarrassment. The precursor to SC34 was largely supported by IBM, now some IBM-ers are not so keen, but in the future other personalities may find SC34 a useful forum again. The wheel will turn. We may see Google keen and participating at SC34 on some technology at some time: why not? There are lots of useful things to do in lots of areas where Google is getting an interest, and sooner or later SC34 may be an appropriate forum.

  • Mark Pellegrini

    4/3/2010 2:16:38 AM |

    @Gareth - "So, the upshot of this is that Microsoft will be reminded of their commitments to the process and adjust their budgets accordingly, if they are indeed committed to the process." - that was gracious of you to rebut your own comment and save everyone else the job of pointing out the obvious.

  • Rob Weir

    4/3/2010 3:21:30 AM |

    Thanks, Rick.  That takes the prize for the most idiotic thing I've read today.  It was a competitive day, but you managed to pull ahead.  Congrats.

    Now it is obvious to all that SAMBA is not based on a standard.  It is based on proprietary Microsoft networking protocols.  So the connection to OOXML is not apt.  I and many others have said that having technical documentation on OOXML was a good thing, but standardizing it was not.  

    You mention that IDABC asked Microsoft to standardize OOXML.  This is not quite true.  They "urged Microsoft to publish and provide non-discriminatory access to future versions of its Office document specifications" in 2006 but only asked them to "consider" submitting them for standardization.  In any case, if you look at the progress of Microsoft technical disclosures and the EC over the past 4 years, you see that they have not asked Microsoft to standardize anything else.  All of the agreements since then have been around private publication of their proprietary interfaces.  The EC appears to have progressed in their thinking in this area and learned from past mistakes.  Why haven't you?

    You (and Alex) also err when you suggest that standardization stands outside of the market and is not involved in picking winners and losers.  Baloney.  Look at ISO's own definition of a standard where it says "Standards should be based on the consolidated results of science, technology and experience, and aimed at the promotion of optimum community benefits".  If you are not making decisions, rejecting the bad ideas and promoting the good ones, then you are failing at standardization.  You are looking toward a plot of dirt filled with weeds and calling it a garden that the magical market will figure out.

    Take the VML example, that Alex is complaining about.  The W3C rejected it years ago.  Yes, they picked winners and losers, and the winner was SVG and the loser was VML.  But SC34 came about, with Microsoft apologists/contractors like you, and approved VML as part of an ISO standard. And now you have the gall to complain that Microsoft is still using VML?  Well enjoy your unweeded garden, Rick.  You deserve it and its fruits.

  • Scott

    4/3/2010 3:36:13 AM |

    @Alex:

    "I somehow can't see the proposition of shifting a workforce of office document users over to TeX being particularly easy to follow-through on!"

    Why not? There are nice programs available now that use it on the back end to serve as examples of how easy it can be. LyX, for example. It doesn't have to be any harder to transition to and to use than it was to move workers from typewriters to computer-based word processors.

    That was achieved through training and/or brute force, and I believe it was a much harder transition than this one would be.

    The key difference between using TeX and the usual practices of most word processor users is shifting from a formatting mentality to a semantic markup mentality. Initially it would be jarring to many people. But, they would quickly get over it when they realize that they don't have to think about the mundane details of fonts and margins any more.

    Today's word processors are, to put it bluntly, stupid. The work of formatting documents should have been delegated to the machine years ago. Instead, the typewriter was virtualized as the word processor, right down to leaving it to the typist to decide where on the page text should appear, what font and point size it should appear in, etc.

    It's funny that a technology from the 1970's holds the promise of correcting the blunders of today's technology, yet it is so easily dismissed as being to hard to implement. It's a good thing businesses didn't think that way about introducing computers to their work forces in the 1980's and 1990's, or this discussion would only be possible in the letters column of some print magazine or newspaper.

  • Rick Jelliffe

    4/3/2010 4:12:29 AM |

    Rob: My pleasure.

    So ISO standardization is involved in "picking winners" after all, is it? I thought you wrote earlier that Alex "overestimated the value of the ISO imprimatur" and its influence was only worth a small amount of inconvenience?

    Your comment "No one is going to hand over control of their technology to a bunch of loose cannons that run alternatively hot and cold" is particularly revealing on many levels: I thought we were supposed to be compliant apologists.

    Your comment "You seem to want to play ODF and OOXML off each other, in hopes that you can force concessions, hoping that continual threats against both standards will ensure that the midget in the middle can control both.  That might have worked before, but we're all on to your little game."  But the midget is a forum of the National Standards Bodies of multiple nations of the world.

    From what I can see, it is legitimate for SC34 to want for parties to stick to the agreements they made or have good-faith negotiations out if circumstances change. Last year it was with ODF issues, and it embarrassed you; this year it may be with OOXML issues, and it may embarrass Microsoft. You complain that SC34 is some kind of passive lapdog, and yet you also complain that it has its own bark and agenda independent of your large US corporations. Err, that rather the point of its existence.

  • Jeremy Allison

    4/3/2010 4:30:48 AM |

    Rick,

    I wasn't opposed to Microsoft documenting OOXML, I was opposed to them getting ISO to declare it a "standard". You can have one without the other you know.

    I once compared OOXML to "good MSDN documentation, but not in the same realm as a standards doc.". I still stick to that assessment. I would have been really happy with the OOXML spec. published on msdn, just like the "Workgroup and Server Protocol Docs.". I even help report bugs in them to Microsoft. Not a standard of course, but then they never claimed them to be.

    As Alex has now discovered, having worked on getting whatever document was necessary to get the marketing "ISO standard" stamp of approval, Microsoft has *no* interest in implementing whatever it is, and could care less. Try explaining the difference between "strict" and "transitional" to any government body and watch their eyes glaze over. In fact try explaining it to anyone but a standards wonk.

    No, all they care about is being able to say "What Microsoft Office emits is an ISO standard, just like OpenOffice". The truth of it doesn't matter.

    You and Alex helped them do that, after being told many times that this would be the outcome, and now we all have to live with your choices. I'm unhappy about that, and I have a right to be.

    Jeremy.

  • Mark Pellegrini

    4/3/2010 4:44:01 AM |

    "Forgive my ignorance, but does this mean that OOXML's status as a recognized ISO standard could be withdrawn?" - I, for one, am waiting with bated breath for Alex to answer this question.

  • TGM

    4/3/2010 5:04:43 AM |

    Let's take the facts here - no media involved.
    Microsoft, by fair means or fail, get an ISO standard in one of their file formats.
    They then proceed to fail to properly implement it as their promise to ISO (all subsequent versions...)
    ISO recognize this.

    Does this mean ISO have learned that Microsoft need a lot more attention apposed to other companies when it comes to standards? If Microsoft brought out a new "open" file format tomorrow, and wanted fast-track, would they get it? Or would ISO force Microsoft through the proper channels? I hope for all our sakes further caution is used in future.

    Can you promise us that Alex?

  • Dean Edmonds

    4/3/2010 5:46:52 AM |

    You seem puzzled by Microsoft's reluctance to move more swiftly toward Strict compliance, saying:

    "playing fair by standards has always seemed like the most obvious way for the corporation to extract itself from the regulatory thickets that have entangled it over the past decade"

    The costs to Microsoft of those "regulatory thickets" are nothing compared to the enormous rents it continues to extract from customers locked into its proprietary formats. All they needed from ISO was a stamp of approval. Having gotten that, it remains in their best interest to delay the implementation of a fully open standard for as long as possible. The surprise is not that they are staffing the standardization effort so poorly, but that they are staffing it at all.

  • Chas Bloom

    4/3/2010 5:54:34 AM |

    Your first mistake was believing Microsoft was some sort of decent company.  Your second mistake was granting them a standard.  Your third mistake is refusing to immediately revoke this standard.

    You are tools.

  • A. Rebentisch

    4/3/2010 6:09:51 AM |

    @RobWeir: IDABC was an administrative programme of the EU-Commission DG Informatics for pan-European egovernment services. Very low-level, without real "political governance", but quite efficient in terms of deliverables and pragmatism. The project which "forced" them to standardize was < 100 000 EUR/anno.
    http://ec.europa.eu/idabc/en/document/3428/5890

    IDABC is succeeded now by ISA, also just 164 Mio EUR, focussed on interoperability, openness, reuse and so forth. What made IDABC very famous was the disproportionate lobbying effort dumped on a harmless programme which was mostly about writing some broschures and guidelines, or development of administrative software e.g. for document management. The primary reasons was that they wanted to do "open source", what apparently the company saw as a bridge head that required a deterrence strategy.

    Other EC market organization interventions are so to speak from a totally different department, on the real political level or from competition authorities, e.g. DG Competition in the Samba case as a competent competition authority, of which players from oversees failed to understand that it acts like a Court, so parties are expected to show respect and do what they are told.
    And the standard policy portfolio is with DG Enterprise, laid doen by laws.

    IDABC/PEGSO on the other hand had formally no authority to command anything.

    Btw, maybe you want to contribute to the Interoperability Strategy consultation?
    http://ec.europa.eu/idabc/en/document/7854

  • Rob Weir

    4/3/2010 6:10:01 AM |

    Rick, I said standardization is about picking winners and losers.  This is any standardization, not just ISO.  The fact that SC34 does it so poorly makes that venue less useful and the imprimatur less valuable.  

    But let's take the contrary view and play with it for a moment and see where it leads. The comments on the OOXML DIS ballot, approximately 1,000 unique ones, were for the most part rejected by Ecma, and the rejections were approved at Alex's BRM on that controversial batch ballot on the final day.  Wasn't that picking winners and losers?  What if we didn't do that, and instead took all suggestions equally?  What would that lead to?  Or are you just in favor of standardizing bad ideas if they come with their own cover sheet?  OK.  What if we took all 700+ rejected comments, and added a cover sheet to them?  Would you favor approving them as an ISO standard?  And then doing the same with every comment on every amendment and corrigenda that has come up?  Anyone who disagrees with a proposal gets their own standard.  So no winners, and no losers?  And then do this not only for OOXML, but for ODF as well.  And bring in other legacy formats, WordPerfect, Ami Pro, WordStar, etc.  Every file format becomes an ISO standard and every time someone doesn't like the direction one of them is going, we fork it to create another standard.  If you are not willing to pick winners and losers that is what you end up with.  

    This is not "consolidated results of science, technology and experience, and aimed at the promotion of optimum community benefits".  This is just making extra employment for standards consultants and editors like you and Alex.  But is it a good thing for anyone else?  I don't think so.

    The point is every time you make a non-unanimous decision, a standards body is picking winners and losers.  You can't avoid this.  You just either do it well or poorly.  Picking winners and losers within part of standard, between revisions of a standard, between conformance classes in a standard, or between standards, it is all the same thing.  You are selecting what provisions will be standardized, and deselecting what will not.

    As for SC34, I'll just note that there are two kinds of committees: those designed to get work done and those designed not to get work done. There appears to be a disagreement among participants exactly which class of committee WG4 was designed to be.  I would not automatically take the large number of Microsoft employees in WG4 as evidence that it was designed to accomplish great things.

    You call SC34 a "forum of the National Standards Bodies of multiple nations of the world".  I call you bluff and pull aside the curtain and instead of the Great and Omnipotent Oz, I see mainly Microsoft employees, partners and consultants.  Of course, I no longer count you since Australia is no longer a member of SC34.

    Remember, the fact that Alex is complaining on his blog rather than getting a resolution in SC34 last week suggests that he has insufficient support to raise the issue there.  If he had the votes, he'd be telling us how this was resolved, not just whining about it.

  • Jean-Bernard

    4/3/2010 6:10:37 AM |

    Looking at Mr Brown's résumé, I'm not that surprised at the process and end result. No insult, but Goody-two-shoes engineer 0 - Wicked businessmen 1.

  • Mike Brown

    4/3/2010 8:04:59 AM |

    @Alex,
    Standardizers should be skeptical of corporations ... I think the view that reduces corporate disputes to some kind of soap opera with "goodies" and "baddies" is reductive and unintelligent



    I appreciate that treating the likes of Microsoft with a Jeremy "why is this lying bastard lying to me?" Paxman approach is a luxury not available to a neutral body such as ISO. Even if that had been your thoughts at the time, you still had to treat all parties fairly and equally.  It's like how we vote for politicians based on what's in their manifestos, even though we know that they probably won't keep too many of their promises once they're in power.  (Just occasionally though, they do come through, e.g. Obama's Health Bill.)

    By the same token, however, you didn't have to bend over backwards to help Microsoft get their standard approved either.  For surely, that is what you did.  Rather than following your own (very sound) advice to be "skeptical of corporations", you displayed an unbelievable naivety in trusting that Microsoft would make any serious attempt to fix its problems once the company had obtained the coveted ISO stamp of approval.

    By any objective view of the text, as it existed at the time of the BRM, the standard should have failed.  Five days was never enough time to fix its mountain of problems.  Game over.  Do not pass Go.  Do not collect $200.  Fail.

    But by splitting the standard into Transitional and Strict versions, the BRM found a way, perhaps the only way, of getting it through.  You also proposed the "all or nothing" vote, when it became clear that there was not enough time to discuss all the NB's issues on an individual basis.  Did ISO rules require you to come up with this?  I think that they did not.  You appear to have interpreted your role as to find the best way to get the standard passed, come whatever.  That was *your* choice and ultimately it was your failure.









  • eduardo

    4/3/2010 8:05:19 AM |

    @Alex: I think there's a misconception at large that NBs voted in favour of OOXML because they somehow "trusted Microsoft". As far as I'm aware this is a long way from being true. If anything, some positive votes may have been motivated more by a distrust of Microsoft!

    Remember that the maintenance regime of OOXML took decision-making away from Ecma/Microsoft and placed it in the hands of the National Bodies."

    The whole immense effort was completely pointless unless Microsoft actually went ahead and followed the standard. That you put in so much time and effort indicates you thought it would. And as I recall, you were scornful of those who thought it wouldn't.

    Alex, are you certain that Microsoft top management intended from the beginning to follow the standard, or do you think it is at least possible it was a fraud from the start?

  • Rick Jelliffe

    4/3/2010 12:07:15 PM |

    Rob: In fact, SC34 has a tradition of providing standards to compete against the currently 'winning' technology. Take SGML for example: it worked against first the presentation-oriented markup languages (troff, TeX etc), then the efficient binary formats (ODA, DOC etc), then the WYSYWIG formats and pioneered linking, separate stylesheets and so on, in each case 'losing' the mass market (until XML, HTML, ODF and OOXML put it into the centre.)

    It was a useful and successful standard even though it never made it to the desktop or the magazine covers, and was limited to a niche area in industrial publishing large document collections (legal, reference, technical, aerospace, weapons systems, pharma, IC documentation, etc.)  

    Of course standardizing existing technology (the normal case) involves deciding which parts are in or out or need tweaking, but "picking winners" as a goal, in the sense that only technologies that get mass acceptance. A standard does not need to be a 'winner' it just needs to be useful and get uptake in its particular niche.

    The point is to ensure breadth, where any niche area of interest of interest to NBs has an appropriate technology, by not to pick single 'winners'. The point has never been either to pick winners nor to be contrarian: the point is breadth.

    Schematron, RELAX NG and even ODF can be seen in this light. OOXML is the exception not because it involves "picking a winner" to exclude others, but because it involves providing breadth using a technology that could reasonably be the mass-market dominator as well as being useful for niche requirements (such as SGML-style automated document production.)

    But arguments based on "this will entrench Office" are rather meaningless when it is already so deeply entrenched in the mass market: the lazy tactic of trying to change the status quo of document formats by merely repeating variations of "They are EVIL" has failed in the past, is failing now, and has slim chances of succeeding in the future. It is because of standardization that we can hold OOXML and ODF implementers up to objective scrutiny about their actual support for open formats: in particular using the objective metric of validation.

    And what are the best tools for validating documents? SC34's DSDL schema languages (RELAX NG, Schematron, NVDL, etc) in most publishing-oriented cases, and in some niche cases W3C's XML Schemas (and in some niche cases OASIS CAM, or OData CSDL, etc). The DSDL schema languages are certainly not the winner in the commercial market: major corporations invested deeply in XSD, and many subsequently have regretted it. According to Rob's logic, I think, either SC34 should not have standardized any of the DSDL languages, because W3C XSD was the winner, or it was wrong-headed to standardize DSDL, because it has not become the 'winner'. But it is just corporatist nonsense: standards don't have to be about the next big thing, nor even the next little thing: they need to ensure a breadth of viable technologies, as far as standardization is capable of doing this.


  • Rick Jelliffe

    4/3/2010 12:31:18 PM |

    Jeremy: What standardization gives us, which merely publishing things on websites does not, is the position we are at now: where the international community can objectively say "This is what you promised, this is what we were willing to live with, and this is what you have delivered" and be on objective grounds using tools like validation.

    I certainly agree that standardization without effective review with broad participation, without ongoing maintenance, without irksome formal procedures to prevent controversial discussions sinking into rabble-rousing, and without objective (and machine-validate-able) conformance criteria is not much better than just putting information up on a website.

    And I agree (and commented so at the time) that it would have been better for ODF and OOXML to be Technical Specifications rather than International Standards: SC34 had a practice that it provided 'enabling' technoligies (infrastructure) because end-technologies (e.g. specific schemas) were asking for trouble. SC34 did not ask for ODF and it did not ask for OOXML, remember. SC34's (and its officers') primary obligation was fairness, which it certainly fulfilled.

    But the option of being a Technical Standard was not the one on the table, was it? If the choices were A) no standard, B) a standard with ongoing maintenance now, and C) a technical report at some vague stage down the track with no commitment to maintenance, then the National Bodies chose B, which I think was the reasonable decision, but of course reasonable people can differ.

    Standardization of OOXML is a win for the community regardless of which way Microsoft goes: if they implement OOXML Strict thoroughly and as their default save format, it is a win because the worst carbuncles as raised at the BRM will have been dealt with; if they don't implement OOXML Strict it is a win because communities with requirements for Open Formats can then insist on ODF compliance. Either way, the situation is an improvement on the status quo of a couple of years ago; just not on the timetable that some ODF boosters pretended was possible.

  • Mike Brown

    4/3/2010 1:51:04 PM |

    @Rick,

    if they [Microsoft] don't implement OOXML Strict it is a win because communities with requirements for Open Formats can then insist on ODF compliance

    That's assuming that those "communities" care or even know about the difference between Strict and Transitional.

    And if Microsoft only supports OOXML Transitional in Office 2010, as is now looking likely, you think that the product won't have "produces ISO-approved, standard documents" plastered on the side of its box, regardless?

    I mean, who's going to stop Microsoft from doing that?  Does ISO have the will or even the authority?

  • Ian Lynch

    4/3/2010 3:46:54 PM |

    It is exactly because those in political power are unaware of the difference between formats and the importance of open standards that those that do know and are involved in the process have to act with ethics at the forefront. A standard has to be fit for purpose and undue influence by specific commercial interests makes it unfit for purpose by definition.  

  • grouch

    4/3/2010 4:40:17 PM |

    It appears Mr. Alex Brown and Mr. Rick Jelliffe have selective memory of the history of 29500. For refreshing those memories, or doing other research regarding OOXML, I recommend:

    www.groklaw.net/.../index.php

    and for the (so far) 5 years of OOXML's history in news and blogs:

    www.groklaw.net/.../index.php

  • A. Rebentisch

    4/3/2010 5:24:03 PM |

    "...those that do know and are involved in the process have to act with ethics at the forefront."

    @Lynch: I was impressed by the IEEE ethics:
    http://www.ieee.org/about/ethics/ethics_code.html
    ISO/IEC should enter that in their standards directives, in particular: "to seek, accept, and offer honest criticism of technical work, to acknowledge and correct errors, and to credit properly the contributions of others;"
    So to speak the antithesis of what happened on the national level, with a few exceptions such as the UK.

  • Ian Lynch

    4/3/2010 5:39:39 PM |

    The snag in the UK is that although the tide is turning and government officials are becoming more aware, implementation of policy is slow. This is why I say the long term outcome is inevitable. Open standards will become the norm in mature technologies. The issue is time scale. The quicker we get there the less expensive it will be. That is why I think there is a need to maintain the pressure on companies that clearly have self-interest in maintaining the status quo and obfuscating the rationale for change. ISO as an organisation has a specific responsibility in all this which I think was badly executed with OOXML. The end result was predictable at the time.

  • Alex

    4/3/2010 9:39:13 PM |

    @Olivier

    > At least ODF has an open source reference implementation

    News to me. What’s that then?


    @Peter

    > does this mean that OOXML's status as a recognized ISO
    > standard could be withdrawn?

    Any International Standard can be withdrawn if that is the will of the nations. I don’t expect anything to happen to OOXML any time soon.


    @Scott

    I sympathise with your TeX enthusiasm, but I can am confident in saying TeX has had its day.


    @Jeremy

    > Try explaining the difference between "strict" and "transitional"
    > to any government body and watch their eyes glaze over.
    > In fact try explaining it to anyone but a standards wonk.

    Not sure it’s as clear-cut as that. The Danish govt for example has been flirting with the idea of specifying OOXML Strict for procurement purposes. Besides, standards activity can hardly be expected to proceed purely based on second-guesses about how ignorant government officials might be! (though marketing activity might).

    As it happens, I have raised in WG 4 the possibility of requiring better “conformance labelling” for applications that claim to implement OOXML – so that consumers know precisely if they’re getting strict/transitional, extended/unextended, standard/custom, etc.


    @TGM

    > Can you promise us that Alex?

    How can I promise what the combined 86 member countries of JTC 1 will decide? It’s their consensus that decides everything. The nature of International standardisation is such that (as Alexander Pope had it) “whatever IS, is RIGHT”.


    @Rob

    > Alex is complaining on his blog rather than
    > getting a resolution in SC34 last week

    As you surely well know, JTC 1 is in no position, at any level, to pass judgement on claimed implementations of its standards.


    @Mike Brown

    You should remember the purpose of the BRM was not to “pass” or “fail” OOXML, but to improve the text. My role was to facilitate that. Most people seemed reasonably accepting of the outcome until the overall result one month later, and then I got it in the neck from the camp that “lost”. I predicted this childish behaviour before  the BRM:

    Tweedledum believes their DIS is so good that only a “failure of process” can thwart it; Tweedledee, however, is convinced that the DIS is so deeply flawed that only a “failure of process” could allow it to become standard.

    And so, while I have up till now thought that a solid grounding in the JTC 1 Directives and meeting procedures would be a good education for convening the coming BRM, I am coming to believe that in fact the best preparation is being a father of two small children, both of whom are sometimes prone to intemperate outbreaks of sibling-rivalry. Inevitably, when this ends with one of them feeling they have “lost” a dispute, the complaint will be “it’s not fair”. I am fully expecting something equivalent when this standardisation process produces a result.


    - http://adjb.net/index.php?entry=entry071211-013357

    My children have matured past this stage now. Shame that's not true for the OOXML complainers.

    Speaking to others who have been in similar situations, I believe it’s a useful truth that if you chair a controversial meeting, you should expect to have shit thrown at you :-(

    The ISO and IEC members subsequently decided whether they wanted OOXML as an ISO/IEC standard, or not. And it turns out they did. End of.


    @eduardo

    I have no special insight to the collective thinking of Microsoft – if they have any collective thinking, that is. (I think the nickname MS have of “The Borg” is so wrong – it seems to me they are a highly federated company and that lack of unity of purpose is more a characteristic.)

  • Rob Weir

    4/3/2010 10:57:51 PM |

    Rick, you are making a false dichotomy and beating up on a straw man of your own invention.  I never said that there were no areas of standardization where it is advantageous to have multiple standards.  I don't think that is a tenable position.  But I do believe that there are some areas of standardization where a single standard is the optimal outcome.  And I believe that it is not a respectable position to suggest that there are no areas of standardization where a single standard is optimal, which is what you seem to be stating.  Discernment is required, and is in practice applied most every day by standards committees when they pick the winners and losers.

    And save your "SC34 tradition" for the history books.  SC34 today is mainly Microsoft employees, business partners and consultants, and has been so since 2007.  The new tradition in SC34 is to do whatever Microsoft wants.  SC34's best days, like Ecma's, are behind it.  It is at the periphery of the markup world.  List the top 20 markup standards in use today.  How many are from SC34?  Aside from ODF, OOXML and maybe Relax NG, the market relevance of SC34 is minute, and the pipeline is even smaller.

  • MSSUX

    4/3/2010 11:19:12 PM |

    Hmm me thinks that all of you miss the point. MS is all about standards - thier standards; and interoperability - as long as everyone interoperates with their products.. all shall be well - according to the management of Microsoft.

    It's all part of the grand plan to continue their monopoly and to make everyone vertically integrate into it.

    To the stooges who pushed MS's agenda for them, and the idiots who went along with it - well what does anyone expect? Bullshit begets bullshit - and the ISO committee is sufficiently stupid enough to prove it.

  • Peter

    4/3/2010 11:52:32 PM |

    "Any International Standard can be withdrawn if that is the will of the nations. I don’t expect anything to happen to OOXML any time soon."

    Thanks Alex, for addressing that question. I'm trying to understand this better and I've got rather naive questions (sorry to bore the experts). But here's where I'm confused:

    The ISO/IEC 29500 standard is actually *two* standards? That is, the "transitional" and the "strict"? Or have I misunderstood? If I haven't, how can we call two different things one thing? I also don't understand how a standard can be considered a standard if it exists only for "legacy" documents. That seems contradictory.

    Will the ISO/IEC 29500 standard eventually be split into 2 documents - one describing an approved standard that is appropriate to implement for the foreseeable future and a second document describing an old (but now unapproved) ISO/IEC standard? It would seem that if this "innovative" split is to work, there needs to be a reasonable deadline to remove the "traditional" strain from the process entirely. This deadline should be set regardless of what Microsoft does with their products.

    Then again, splitting the standard seems like a questionable decision in the first place. Or have I misunderstood something?

  • Alex

    4/4/2010 12:04:06 AM |

    @Peter

    Oh dear - you asked for it Smile

    OOXML is actually four standards - or, to be more precise it is a four-part multi-part Standard.

    Part 1 is the description of the core markup languages for office documents

    Part 2 is the description of the packaging mechanism (OPC, or Open Packaging Convention), a ZIP-based format which is shared with XPS

    Part 3 describes extensibility mechanisms

    Part 4 is a set of additions and overrides for Part 1 for "Transitional Migration Features".

    So when we talk of OOXML Strict we means Parts 1, 2 & 3.

    Wehn we talk of OOXML Transitional we mean Parts 1, 2 & 3 taking account of the additions and overrides of Part 4.

    The Transitional variants of OOXML could therefore be withdrawn by withdrawing Part 4 of the Standard.

  • Rick Jelliffe

    4/4/2010 12:57:25 AM |

    Rob: Sour grapes?

    On the contrary, WC34 WG1 (Schema and markup), WG2 (Information Presentation), WG3 (Information Association), WG5 (Document Interoperability, and WG6 (ODF) are not stuffed full of Microsoft employees, as far as I am aware: are you claiming MS has a plurality in all or any of those WGs?  WG4 (OOXML) indeed has had many ECMA and Microsoft people involved, as is only to be expected: is it a plurality? If we exclude MS partners, who you find so sinister, should we also exclude IBM partners? The resolutions from the plenary lists participation by "Brazil, Canada, China, Czech Republic, Denmark, Finland, France, Germany, Japan, Korea (Republic of), Netherland, Norway, Sweden, UK, USA, Ecma International, OASIS, W3C and XML Guild". Which of these NBs had a plurality of MS employees in their delegations at the plenary?  Or is it the mirror committees at the domestic NBs?

    As far as SC34 standards not having the top 20 markup standards, err that is because SC34 doesn't try standardize end-user markup standards in that sense, as I mentioned. SC34 is not OASIS or W3C because it has never wanted to be. (Nor would it be very good at it.) Indeed, SC34 sometimes even sloughs off technologies that get too close to applications or have viable maintenance prospects in more agile organizations: we handed over maintenance of the public entity sets for special characters to the W3C MathML WG for example. XML is another example.

    SC34 has the reverse of an empire-building mentality, which must puzzle a hyper-competitive corporatist.

    But when you look at the number of important standards that are defined using RELAX NG (e.g., ODF and multiple SC34 recommendation), or the impact of the OpenFont standard, then certainly SC34 has very successful standards. A little standard like Schematron keeps planes from colliding over Europe, checks criminal data in the UK, vets transactions coming into the greatest insurance market in the world, and checks naming rules in the US NIEM schemas: to say these standards lack market relevance shows a phenomenal but ignorant aggressiveness.

    This is probably difficult to grasp for an inveterate factionalist, but SC34 is not about standards that "win" in the sense Rob seems to put it  at all. That Alex is talking about de-standardizing Part 4 of OOXML, which has tremendous 'market relevance' and is, in broad strokes, the 'winner' shows that being number 1 is not what SC34's focus has ever been about, at least in my sporadic but long experience of it.

  • Rob Weir

    4/4/2010 2:00:28 AM |

    Rick, you say "not stuffed full of Microsoft employees, as far as I am aware".  But the problem is you are not aware.  Have you even attended an SC34 Plenary since Microsoft ended their contract with you?  Australia's participation in SC34 has disappeared altogether.  If SC34 is so relevant, then why is your own NB not even an O-member?  

    But don't take my word for.  Sneak into an SC34 Plenary sometime and ask others.  At the last two I've been approached by delegates from various NBs dismayed by how many Microsoft employees were there.

    To your other point, if you suggest that SC34 is all about picking losers, then touché, Rick.  My point merely was that SC34 doesn't have what it takes to make market-relevant standards.  It fails predominately because it lacks the confidence and participation of major vendors, without which it can only eek out a meager existence on the fringes of the markup world. Even with a large number of Microsoft employees present and voting, SC34 is not exactly their preferred venue for standardization.  They tolerate it.  And honestly, if they decided they had enough with the nonsense in SC34, and withdraw OOXML, I'd need to consider recommending the same for ODF.

  • Scott

    4/4/2010 2:19:51 AM |

    @Alex

    "I sympathise with your TeX enthusiasm, but I can am confident in saying TeX has had its day."

    My enthusiasm is not for TeX, so much as it is for a semantic work flow that de-focuses on formatting details. TeX may have had its day, but semantic markup and semantic work flow have yet to have theirs. I'm just saying that the way today's document preparation systems are typically used leaves much to be desired and that much can be learned from TeX and from programs such as LyX.

    Your dismissive attitude contributes nothing to the discussion.

  • Alex

    4/4/2010 2:36:05 AM |

    @Scott

    I'm all for semantic markup, and I agree that from a conceptual viewpoint office documents are fairly dumb. However I simply don't believe we can expect the disciplines of semantic markup to be broadly adopted in maintstream document workflows - not for documents, and especially not for spreashseets and presentations.

  • Alex

    4/4/2010 2:55:48 AM |

    @André

    I covered the i4i patent in a recent post:

    www.adjb.net/.../...mat-Standards-and-Patents.aspx


    @Rob

    Now I'm intrigued. How, procedurally, would a corporation like Microsoft go about withdawing OOXML?

    Interesting to see you'll be taking your cue from what Microsoft does -- when the vendors close ranks we can be sure we're on the right track!

  • Rob Weir

    4/4/2010 5:40:37 AM |

    @Alex, there are ways.  Remember, SC34 is only one of many ISO/IEC SC's, and ISO/IEC is only one of four places where International Standards are maintained.

  • Scott

    4/4/2010 6:49:52 AM |

    @Alex

    "I'm all for semantic markup, and I agree that from a conceptual viewpoint office documents are fairly dumb. However I simply don't believe we can expect the disciplines of semantic markup to be broadly adopted in maintstream document workflows - not for documents, and especially not for spreashseets and presentations."

    I cannot agree with you here. If all that mattered is how a document appears when printed, I could. But increasingly, documents target multiple mediums, and direct formatting falls apart quickly when that's the case. I think we can expect that as more experience with this is gained, that the benefits of semantic markup will become clearer and more desirable. It's still too early to declare it too hard to implement.

  • Mike Brown

    4/4/2010 7:39:54 AM |

    @Alex,

    Thanks for the explanation of the differences between Strict and Transitional.

    Is there a point in time at which Microsoft (or anybody else for that matter) can no longer claim compliance with the ISO standard if they're only adhering to the Transitional rather than the Strict version?

    Surely, the Transitional version must come to an end at some point.  Otherwise, it's not really transitional, is it?

  • Jack

    4/4/2010 7:58:33 AM |

    As far as I Known no application supports either the Strict or the Transitional part of the standard.

    If anybody knows of one please tell.

    Since it's not implemented by anyone and it will induce in error those who think that it has any merit (like governments) wouldn't be correct and honest to retract its ISO certification?

    Not doing that is not only dishonest but criminal!

    It's sad hiding behind consensus when there's none.

    The only power that ISO has is that of revoking ISO status; if not what's the point of having a standard.

    Anyone can and will claim that they have when they in fact don't. Like Microsoft. But also, those that Alex Brown mentioned. Not one supports either standard.

    The ISO certification was farcical.
    The implementation of the standard is also farcical.

    It's time to put an end to this!

    We already have a standard for documents, why are we wasting our time with this?

    Doesn't ISO or the people working for them care about their credibility?


  • Jack

    4/4/2010 8:06:04 AM |

    ...we can be sure we're on the right track!

    Do you really though that you were on the right track anytime during this process? Really!?

    A redundant format that nobody was interested in implementing, that a majority of independent technical people found lacking.

    You are on the right track to waste a lot people a lot of time and money that could be spent doing more important things.

    Being involved in a "Document File Format Soap Opera" is not something one should be proud of.

    The fact that nobody besides Microsoft cronies comes to the meetings is telling. Don't you ever asked yourself: What the hell am I doing here? What the purpose of this? Travel expenses? Like The British MPs?

  • Peter

    4/4/2010 2:14:53 PM |

    Thanks Alex, for explaining in detail the Transitional vs. Strict parts of the standard.

    Yes, the process of removing the transitional part of the standard should occur immediately. Office 2010 should not be allowed to ship claiming it adheres to an ISO standard.

    What do you think the likelihood is of getting the transitional part removed before Office 2010 ships?

  • Alex

    4/4/2010 6:27:02 PM |

    @Jack @Peter

    I was careful in my choice of words about Transitional ("T"), in writing that it is time "to start the wheels in motion". I don't think we want an immediate withdrawal of T.

    T is a very important standard! It is (apart from a small and shrinking number of defects) the format of MS Office 2007, and of many back-end enterprise and office software systems the world over. There are millions of desktop installations of applications which support it, and hundreds of millions of T documents out there.

    SC 34 is working to ensure that T does accurately represent the existing corpus of these legacy documents - thus it is valuable as a "field manual" for people having to work with these formats. And that includes FOSS people too.

    The complaint of this blog posting is not about OOXML itself, but of Microsoft's decision to implement T (only) in a new Office product. T should be a dead language, like Latin. The International Standards community did not want it be to used for new documents.

    What might be useful for T is a phased deprecation plan, to make its purpose more obvious to everyone. But right now I think the correct approach (for us standardizers anyway) is to introduce conformance labelling provisions in the standard so that implementations had to make it very plain exactly what kind of OOXML they supported. People procuring systems, armed with the information, will be empowered to make more a informed choice.

  • Rick Jelliffe

    4/4/2010 9:36:41 PM |

    Rob: Yes, all standards groups need to avoid domination by individual corporations or cartels or particular nations etc. Where one company or one sector can dominate (i.e. too many SME representatives, too many US corporations, too many users, too many govt reps) it is suboptimal. Indeed, as you know, the US law has case law (Allied Tubemakers) about how standards bodies must not become a mechanism for a cartel of vendors to exclude an unwanted competitor's technology.

    "Remember, SC34 is only one of many ISO/IEC SC's, and ISO/IEC is only one of four places where International Standards are maintained."

    Well, documents don't fit into the scope for IEEE or ITU or IETF etc, so I think you may be well be stuck with ISO/IEC: information technology at ISO/IEC is allocated to JTC1, and the SC for Document Description and Processing Languages is SC34. If you can get enough NBs to vote on a new specialist SC merely because you don't like the representatives that NBs and liaison groups send, feel free to try.

    Oh, except it is the same NBs who already participate in SC34 who would have vote, so the result would be a new group with exactly the same NBs involved, who would be sending exactly the same delegations. Same irritating midgets attempting to be both fair and critical on drafts from any source.

  • MSSUX

    4/5/2010 8:50:01 AM |

    I actually was seriously involved with the entire process of MS rigging the approval of THEIR standard as a global standard via the ISO committee. In a nutshell - it was an effort to undermine and head off the competition, and continue market dominance.

    Unfortunately for MS, they have delivered enough shitty products, with great features like a complete lack of backward compatability, the lack of interoperability etc., etc., etc., to make them an expensive and unpalatable proposition.

    Consider one of the managements latest moves - selling Windoze 7., I think they were selling it online and in packs for around $100 US, and they were selling it in other countries around the world for around twice the price.

    They undertook a campaign of stopping people outside the USA from buying from the USA, by blocking the sale via IP address and by delivery address.

    Since the management of Microsoft is that stupid and that greedy - to not only try on that crap, but when Bill Gates was asked on UK television why the British had to pay twice as much as people in the USA, he said, "It's to allow for currency fluctuations".

    Sure the UK pound is going to devalue by 50% overnight.....

    Idiot.

    That is what the Microsoft people are like - they are full of shit.

    Lying, scheming and ripping people off - and this goes back to the same scams with the ISO certification and way back beyond that.

    I used to kind of like using Microsoft software, except for the many "Who is the idiot who thought this one up?" types of observations I had when using it.

    Now I am only using some Microsoft products as legacy applications - as I am now a full time Linux (Ubuntu) user....

    There is no more buying rebadged versions of the cash cow - or what the management of MS call upgrades.

    That is all the Microsoft have left to stay in business - the deceitful art of giving last years model a coat of paint and a set of fresh tires......

    And it just gets worse, and more insane and I and many others just groan more and refuse to buy / upgrade / update and switch to Open Source because we are sick of being served up crap and getting ripped off in the process.

    For instance, with DSKCHK - it does a typically pox scan of the hard drive surface, and then hides the results many layers deep in a file - with no references or obvious links between the process and the result.... "I mean who is the idiot that thought that one up?" - it's just more of that never ending Microsoft bullshit.

    So I don't care what MS does any more - their products have cost me way more in lost time and resources that what I paid for them.

    They seem to have the idea that even pirated versions of their products are worth having....

    (Bin please - and CD frisbees a go-go)

    And the Open Source products, generally speaking - are sooooooo much better.

    I just feel so relieved at the idea of NOT having to have Microsofts naziware products in my computer and nothing but their naziware in my computer.

    I am just so grateful for all the people who have helped to make the open source software that we have today - and I am pleased and proud to be using it.

    So much so I want to spend some of my time, contributing to some of it's development as well.

    For all the value of this discussion - I'd only accept Microsoft's submissions for anything, if it was printed in non carciogenic ink, on huge rolls of soft and gentle to my anus toilet paper....

    Ubuntu, Ubuntu, Ubuntu - traaa laaaaa, laaa, laa.....

  • orcmid

    4/6/2010 2:23:40 AM |

    "What might be useful for T is a phased deprecation plan, to make its purpose more obvious to everyone. But right now I think the correct approach (for us standardizers anyway) is to introduce conformance labelling provisions in the standard so that implementations had to make it very plain exactly what kind of OOXML they supported. People procuring systems, armed with the information, will be empowered to make more a informed choice."

    From my (office) armchair, that looks promising.

  • MSSUX

    4/6/2010 7:51:05 AM |

    What might be even more useful is to harvest all the Microsoft lackeys who rigged the ISO certification, and all their stooges who went along with it - and to drag them outside and give them all a good kicking.

    The pro Microsoft - created this mess., if any of the tossers who declare themselves to be "standardisers" had any balls, they'd pull up the fact that the pro Microsoft were and are corrupt and kick them and their corporate agendas out of the ISO completely....

    All you morons - having MS and it's greasy palmers, is like allowing pick pockets into a conference, and then allowing them to stay - is like giving them the keys to the cloak room.....

    Well Duh - unethical conduct and rigging things and typical Microsoft scamming - begets nothing but problems and problems beget more problems...

    Wipe the slate - Kick Microsoft and their stooges out of the ISO certification process.

    That is unless of course your too stupid, spineless and that much of an arse licker yourself, to be able to grow a brain and speak up and kick them out.....

    The ISO is full of crawlers anyway......

  • Stephen Walli

    4/6/2010 7:51:35 AM |

    @Alex:  
    But right now I think the correct approach (for us standardizers anyway) is to introduce conformance labelling provisions in the standard so that implementations had to make it very plain exactly what kind  of OOXML they supported. People procuring systems, armed with the information, will be empowered to make more a informed choice.

    Why was this conformance language NOT included in the first place if that was indeed the intent?  Microsoft already flaunts the conformance requirements in their marketing (http://bit.ly/abtyOY).  Strict language would have made things clear.  

  • Rick Jelliffe

    4/6/2010 9:12:40 AM |

    MSSUX: You seem to have a problem.

    But what you are asking for is not possible. JTC1 cannot ban delegates based on hearsay: it has to accept which ever delegates a National Body sends to an SC meeting.

    And while National Bodies may have rules for making sure there is a spread of representation or to give preference to national rather than international organizations, most (in democratic countries at least) would be unable to ban Microsoft unless there has been some legal/regulatory incrimination of them that was directly material.

    The laws/regulations on these issues in many countries are strongly against cartelization: I have mentioned the Allied Tubemakers case in the US, whose reasoning seems likely to be applicable in many countries including Australia, but if you followed EU competition csarina Neelie Kroes comments at the time, she also mentioned the cartel issue. So NBs cannot exclude a company merely on the say-so of its rivals. The UK case where British Standards' activities were challenged was thrown out of court.

    In the case of the Standards Australia delegations to the BRM, delegations were chosen by seniority: people with the most experience. (Indeed, I had to sign a contract that I was not being paid by any foreign organization in any way for my time at the BRM. In fact, I originally made it known that I would not be available for the BRM and only accepted when asked by SA after the initial delegates became unavailable for family reasons at the last minute.)

    Editors and Invited Experts are a different case, under the ISO rules (Google for "ISO Annex SE"). They are there by appointment or invitation per meeting and whether they are asked or heard on any issue depends at any meeting is up to the meeting and its chair; certainly they don't participate in any plenary votes since they are by NB delegation. Experts act in a personal capacity.

    So the only feasible chance for excluding Microsoft is to vote No on a particular standard that they think would be harmful. However, the standard is not Microsoft, and indeed since many participants think that the lack of information from MS is one of the problems with MS, doing anything to discourage more disclosure would be self-defeating.  (And many NBs have a strong disinclination to take notice of any non-technical issues apart from market relevance, which OOXML clearly has.)

    So talk of growing brains to kick them out comes from someone who writes with no knowledge of what is actually possible in the situation. Just as with the "contradiction" claims, where newbies decided what contradiction must mean apparently without bothering to ask what had actually been found to be a contradiction before (it is quite limited). The same pattern happened repeatedly: the purpose of the BRM and the conduct of the BRM: people make up rules to let themselves win on technicalities then get offended when it is pointed out that the procedures are quite different.

    ISO is very fair: demands for special treatment just won't swing; and claims that ISO's inability to allow special (negative) treatment for the dastardly Microsoft are actually special (positive) treatment don't wash.

    So given there is no technicality to prevent a technology that comes from Microsoft becoming a standard, and given OOXML had market relevance, and given it had come from a liaison group with submission rights and had been worked on for some time, and given that it was the kind of thing that many SC34 participants and document people had been calling for, and given that Microsoft clearly has a large and legitimate ecosystem of developers and integrators and competitors who could legitimately be rallied to support IS29500 in NBs, and given that options such as conformance options such as Strict/Transitional were possible that would allow some kind of win/win between NBs who wanted documentation and NBs who wanted trimming, it is not suprising that DIS29500 failed first then succeeded after the Ballot Resolution Meeting put in necessary changes. You cannot blame that on scamming.

  • MSSUX

    4/6/2010 9:36:41 AM |

    Well you opinion is greatly appreciated - but much like Microsoft rigging the prices for Win 07 everywhere else in the world to be twice the price of the USA and Bill Gates telling the people of the UK that they did it because of currency fluctuations...

    Oh he means that $200 US in the UK vs. $100 US in the US and A, is to account for the pound dropping 50% in Value overnight?

    Sure...

    Just like your statement:

    "So the only feasible chance for excluding Microsoft is to vote No on a particular standard that they think would be harmful. However, the standard is not Microsoft, and indeed since many participants think that the lack of information from MS is one of the problems with MS, doing anything to discourage more disclosure would be self-defeating."

    Microsoft is a corporation run by people - these people have consistently made many decisions to monopolise everything they can and to crush the competition through unfair and dishonest means.

    Now your saying "OH Microsoft is playing games with communication and us" and "To kick these leeches out, would be making it even worse".

    I have one question for you, "Were you born stupid? Or does arse licking to arseholes give you some kind of inner warm fuzzy feeling?"

  • MSSUX

    4/6/2010 9:40:07 AM |

    Re phrase:

    I have one question for you, "Were you born stupid? Or does
    crawling to arseholes with "peculiar distortions of reality" give you some kind of inner warm fuzzy feeling?"

  • Rick Jelliffe

    4/6/2010 11:00:46 AM |

    MSSUX: As I wrote, it was just not possible to "kick these leaches out" from the ISO/IEC JTC1 process: it is not constituted to do so and could get into legal trouble if it did kick out anyone on non-technical grounds. The same is true for many national bodies. You can froth at the mouth all you like. (You can use as many double quotation marks as you like too.)

    Rather than waste "your" emotional energy, why not channel it to open source projects that don't benefit any large US corporation but people in general? That is what many SC34 participants actually do.

  • Rick Jelliffe

    4/6/2010 11:23:46 AM |

    Stephen: In the standards eco-system, ISO is not involved in conformance testing: that is for organizations like NIST or National Bodies or organizations certified by a registration authority. For the US, you should lobby NIST. I believe the EU has a test plan under way and are developing tools: Alex has better info on this I am sure.

    All that the ISO process can do is set up categories and schemas and perhaps abstract test suites: the ISO 9001 example is a good one: refer en.wikipedia.org/wiki/ISO_9000#Certification

    But where there is no certification or testing regime set up at the NIST level, as a lack of interest by the organization, there is nothing that ISO can do.

    In the case of SGML/XML documents, we are in a better place, because the schemas do provide an objective way to test documents; however, the valid/invalid result that most schema languages return is inadequate for real use: a testing regime needs to have some sense of which errors are significant and which are trivial for example. Most validation technologies have an implicit production process that an invalid document should be rejected and sent back to the creator for fixing by hand, which is patently not appropriate for consumer word processing documents!

    What IS8879 SGML (from year 1986) did was have a feature list (the System Declaration) that applications had to tick off. All applications had to support the base, but could choose which features to select. Customers could select based on this. (Documents had a similar list, the "SGML Declaration, so you knew whether a document would be processed by a particular system.) In fact, few vendors bothered to do this. (ODF 1.0 had a similar kind of idea with its non-normative table of features that various applications should support, though that table seemed to match the feature set of a particular product, so some people had mixed feelings about it.)

    I suspect that where legislation requires conformance to open standards, procurement people are now on much stronger grounds for demanding actual conformance, however they will need some institutionalized objective conformance testing to back themselves up by.

    So the issues for IS29500 OOXML (and IS26300 ODF) is whether the standards currently have good enough categories to allow a testing body to discriminate between good and poor implementations. Contra Alex, I think there is currently a workable level of conformance language--the sentence on not using Transitional for new documents couldn't be clearer (congratulations to Rob and the other US delegates who drafted it): the issue is the level of knowledge and the backbone of the testers and procurers.

  • SGJ

    4/6/2010 11:35:02 AM |

    Alex, you wrote: Looking at the text, I reckon it is more like 95% that remains to be done, as it is still lousy with defects.

    The obvious question is then: why was it approved as an ISO standard in this state, much less via the fast-track process?

  • Stephen Walli

    4/6/2010 1:19:00 PM |

    Hi Rick:  I completely understand that ISO can't (and shouldn't) be responsible for certifications and warranties.  People that care need to put their money where their mouth is and organize their own certifications.  For POSIX.1 in the bad old days, it was certainly NIST on behalf of the U.S. government procurement organization.  For UNIX it was X/Open, on behalf of their collected vendor membership trying to demonstrate their own conformance.  But both certification efforts were based upon the conformance language of their respective specifications.  Microsoft was flaunting the OOXML spec in it's marketing two years ago by claiming lots of implementations that "support" the standard.  Any tightening of the conformance language would help (eventually) when someone cares enough to put a certification mechanism in place.  

  • Alex

    4/6/2010 1:58:11 PM |

    @SGJ

    Why was it approved? Well, the simple answer is: because the ISO and IEC member bodies wanted it approved. I think some reasons included:

    - Specific technical issues NBs had raised had been dealt with fairly well by Ecma (and the way the voting forms work, many NBs commit to alter their vote if their comments are addressed)

    - The quality bar for this kind of standard had been lowered to the floor for ODF, establishing a precedent

    - No standard is perfect, and given that maintenance had been transferred to International control, the potential for improvement was there

    - MS Office technology is pervasive and likely to remain relevant to large consituencies of users of it for a long time, so their voices will have been asking for a standard (see e.g. the public/governmental institutions voting for it in the U.S.)

    - Given the choice between having this technology inside the standards tent, or outside, the former was seen as preferrable

  • Mike Brown

    4/6/2010 5:16:43 PM |

    @Alex
    >> Why was it approved? Well, the simple answer is...

    Whatever.  What's done is done.  (I can't wait to see Rob Weir's response to your comments about ODF though!)

    The question is: what does ISO do now?  Is it going to allow Microsoft or any other company to claim compliance with the standard indefinitely, even though their products are producing documents in the Transitional and not the Strict version of the standard?

  • MSSUX

    4/6/2010 7:32:01 PM |

    Mmmmm I think I will shut up after this....

    I read stacks of material on Microsoft's fast tracking of the ISO certification with their own specific OOXML (?) format over the ODF (?) specification.

    I read the committee reports, and the forum fights for and against the ODF and the OOXML formats.,

    Almost all of the flack in the air was about Microsoft stacking international committees with their own members - magically enmass and generally out of thin air.

    Microsoft are not my friends, "they" are a corporation, driven by profit - manifested by greed and unethical practices; in order to do one thing - To generate revenue for themselves.

    They do this one the basis of two tactics.,

    1. To make sales; and

    2. To repeat those sales; by

    a) Product upgrades; through

    b) Product lock in's.


    The Microsoft standard and it's "certification" by manipulation through the ISO process, was to generate it's OWN standard, as a form of regime, that everyone else, they hoped would comply with and toe the line too.

    The big problem with the monopolistic practices of Microsoft is that the harder they squeeze people into product lock in, the more they escape into the open source product lines.

    For instance Open Office (.org) in October 2009, achieved 100,000,000 downloads.

    http://www.openoffice.org/news/

    No doubt that would incurr many upgrades of repeat consumers, but it also reflects the uptake of new software by people and entire economies and whole countries.

    While Microsoft had the power of it's mass of it's user base and inertia of it's slow footed and dim witted corporate size; it's unethical practices to achieve that transitional point, it's worsening and indeed openly deviate practices to keep a desperate grip onto the peak once headed into that unyeilding arc of fall, are coming back to haunt it.

    The outrage surrounding the market dumping of Vista, the ripp off scam of double dealing the retail prices of Win 07, with one country getting one price and every other consumer in every other country paying double - with blocking of international sales by Microsoft.

    Then Microsoft copped the worlds largest fine for anti-competitive practices from the European Union:

    European Union Microsoft competition case
    en.wikipedia.org/.../European_Union_Microsoft_competition_case

    This is a tale all and by of it's self with much corruption and stories with in stories.

    As far as Microsoft's rigging of the ISO certification process, the web is thick with corruption and the rigging of the approval process:

    I cite just 2 of such cases in the following.

    So before I hand you over to them and sign off for good, "Microsoft's OOXML standards? Who cares? I don't I am now a full time Ubuntu Linux and Open Office user.

    Microsoft can jam their corruption and theft, right up their collective arses.

    The articles:


    http://government.zdnet.com/?p=3741

    'Scandal' over Norway's vote for OOXML

    Posted by Richard Koman @ March 31, 2008 @ 9:13 AM

    Categories: Government technology

    Tags: ISO, Norway, Committee, OOXML, Register, Iso standards, Process Improvement, Quality, Business Operations, Richard Koman

    We hear stories of “voting irregularities” around the world, not least of all right hear in the USA, but surely the ISO is different. No? The Register reports that delegates have been “complaining loudly about alleged heavy-handed tactics and misdeeds in the voting process” on Microsoft’s OOXML standard.

    The Register says that Microsoft appears to have locked up the election after a number of countries’ standards bodies made abrupt switches to favor the format.

    Norway’s Geir Isene says the country sees a booming business in fixing “such a broken standard.

        The meeting: 27 people in the room, 4 of which were administrative staff from Standard Norge.

        The outcome: Of the 24 members attending, 19 disapproved, 5 approved.

        The result: The administrative staff decided that Norway wants to approve OOXML as an ISO standard.Their justification: “Standard Norge puts emphasis on that if this [OOXML] becomes an ISO/IEC standard, it will be improved to better accommodate the users’ needs.”

        This translates to: “Yes, we know the standard is broken, 79% of our technical committee have told us. But we hope that it someday will be repaired by someone. And we’ll be happy to help if someone can give us the resources.”

    What happened in Norway is nothing less than a scandal, says Wium Lie of Opera Software, according to Computerworld.dk (translation via GrokLaw)

        “This is a scandal! I am shocked. I am speechless. 21 members of the committee say no, while MS manages to win through its position anyway. It is incomprehensible,” says Wium Lie angrily.

        The meeting on Friday began first with a general committee in Standards Norway, which thereafter was reduced to key individuals in the organization plus five representatives who were in attendance at the ISO meeting in Geneva. At the end, even these were dismissed, and three individuals from Standards Norway made the decision.

        “Those who made the decision at the end are not those who know the most about this. They are not qualified to make this type of decision and do not necessarily have national interests at heart. They do not represent the Norwegian committee,” says Wium Lie.

    The committe head, Steve Pepper, is similarly distressed:

        “I am deeply shocked that Standards Norway has chosen to ignore the large majority in the committee. It is the bureaucrats in the committee who have said yes, not Norway. It is a win for MS and a great loss for the rest of the world,” says Pepper.

        “So if the majority in the committee were against OOXMl, why do you think that Standards Norway said yes. They have been targeted of enormous pressure from one market interest which has use of great resources, and they have most likely been more preoccupied by their own interests as a standards organization than by the end users interests,” says Pepper.



    boycottnovell.com/.../

    09.30.08
    OOXML Scandal, Part 1001: ISO/IEC ‘Only’ 6 Months Late, Still Secretive

    Posted in ECMA, IBM, ISO, Microsoft, Open XML at 5:49 pm by Dr. Roy Schestowitz

    ISO Sold Out to ECMA

    A FEW DAYS after IBM protested [1, 2] against the abomination which is ISO/Microsoft (the latter had captured the former) there’s some response but not in the form of a statement. That evidently-corruptible ISO is instead reminding people of what it calls “World Standards Day”:

        World Standards Day is celebrated each year on 14 October to pay tribute to the efforts of thousands of experts worldwide who collaborate within IEC, ISO and ITU to develop voluntary International Standards that facilitate trade, spread knowledge and disseminate technological advances.

    OOXML on the trash canShown on the right is the man who, along with Microsoft, denied the endless manipulations involved. They care not for gruesome evidence but only for their own ever-sinking reputation. And over at Honeymoon Island, one can now see Patrick Durusau and his new friends from Microsoft, who might be having a jolly good time, and not for the first time, either.

    More important, however, is the following leak, which reveals the official ISO circulation (still internal) of the ISO-OOXML specification. According to the directives, these specifications ought to have been circulated back in March, not the end of September.

    So, magically enough, ISO can pass shoddy specifications in a matter of 6 months (fast track) but is still unable to pass a copy in less than 6 months. How come? 6 months late, available to members only? Appended below are those details of interest. Is ISO operating a stamping shop? More countries and companies should follow IBM’s lead and let ISO dwindle in the darkness where it already operates anyway.

  • hAl

    4/6/2010 10:06:01 PM |

    As could be seen in Jespers posts on Office Open XML Microsoft has already been removing more and more of the transitional stuff from the Office 2010 implementation.
    idippedut.dk/.../Moving-towards-OOXML3cS3e.aspx

    It is regrettable if Office 2010 does not yet support the strict schema but frankly in a world where 95% of document are still created trough the old MS binary format this is not a disaster.

    Let's be more realistic. Document format change is very slow in pace. It just takes Oasis about 5 years to create ODF 1.2 specs from ODF 1.1 (which Rob Weir originally roadmapped for ISO standardisation by september 2007).

    We should keep pushing Microsoft towards generating the strict schema towards the future.

  • MSSUX

    4/7/2010 1:10:50 AM |

    We?

    "We should keep pushing Microsoft towards generating the strict schema towards the future, said hA1....

    Ummm you mean the people who manage Microsoft in their typically dishonest and unmanageable fashion, and according to you, can't get their shit together enough to continue to rig the stooges in the ISO to create a fake standard all by themselves?

    So you now take it upon yourself to speak on behalf of everyone, that WE have to keep doing it for them........

    Who in the fuck is we?

  • Philip

    4/7/2010 4:03:57 AM |

    If you have >5000 pages of documentation it is quite normal that you have lots of mistakes Smile

  • MSSUX

    4/7/2010 4:29:02 AM |


    Doug Mahugh, employee of Microsoft, in his blog writes:

    blogs.msdn.com/.../...or-iso-iec-29500-strict.aspx

    "I’m humbled by the level of expertise that WG 4 members bring to the table, and also by the commitment of those who volunteer large amounts of their own time to work toward improving the standard. I know I speak for every member of WG 4 in saying that we’d love to have even more participants involved."

    "Calling all toadie$$$$...., Calling all toadie$$$$...."

    And the next time you get that "locked in" idiot ribbon on the worlds most hacked and insecure office suite, on the shittiest of operating systems that doesn't come with any malware protection - you know who to thank.

    But what the fuck? It is certified as ISO compliant - from one of the most rigged and bribed committees that Microsoft's money can buy!


    (Hey Doug Mahugh - Nice article - great piece of spin doctoring)

  • Mike Brown

    4/7/2010 5:52:05 AM |

    @Doug,
    >> In Office 2010, we’re providing read/write for
    >> Transitional and read-only support for Strict.
    (from your own blog post)

    Which will be of use to who, exactly?

    Who is going to generate all these Strict documents for Office 2010 to read if Office 2010 itself can't actually produce them?

    OOXML is, essentially, Microsoft's standard.  Nobody else is going to lead the way on this while Microsoft itself is seen to be hanging back.  Third parties will just copy whatever Office produces, same as they've always done.

    A cynic might say that this is exactly what Microsoft wants.

  • Jomar Silva

    4/7/2010 5:57:03 AM |

    @Doug,

    Just in case you don't approve my comment on your blog, I'll cc it here:

    Read what no one writes... you guys are entering the Monty Python area again, hmm ?

    Cheers,

    Jomar

  • MSSUX

    4/7/2010 7:19:00 AM |

    Yes Doug - your publicly publishing my email address simply shows you up for what you are...

    Another scumbag Microsoft employee...

    "More bullshit OOXML standards anyone?"

  • Mike Brown

    4/7/2010 7:23:09 AM |

    @Doug,

    >> would you prefer that we prioritized writing Strict
    >> over reading it?

    Are the two mutually exclusive in some way?  To put it another way, do you know anybody that can write but can't also read?  The former implies the latter, does it not?

    So on that basis, yes, you should have prioritised writing Strict documents.  It would seem that Alex Brown thinks so too, or did you miss the title of this blog post?

    You've also dodged my main question: who/what is going to produce all these Strict documents for Office 2010 to read if Office itself is incapable of doing so?













  • MSSUX

    4/7/2010 7:52:01 AM |

    Yes Doug - Since OOXML the Microsoft standard, rigged for user unfriendly use, was printed out in over 5000 pages around 2 years ago....

    Well when Mike Brown put this question to you:

    "You've also dodged my main question: who/what is going to produce all these Strict documents for Office 2010 to read if Office itself is incapable of doing so?"

    My first thought was..."When"...

    "If it was about 2 years ago that MS got the ISO committee rigged enough to make a fake ISO standard for themselves, why didn't they start working on their bug infested standard 2 years ago?"

    And

    "How come people like you Doug - do not as yet have it finished?"

    I mean did you get your university degree in Wankanomics from a $10 certificate farm, a mere day before applying for a job at Microsoft?

  • Doug Mahugh

    4/7/2010 8:51:39 AM |

    @MSSUX, it's even worse than you think -- I've never set foot in a university classroom, actually. But I hear they can be a lot of fun, so maybe I'll do a Rodney Dangerfield "Back To School" thing some day ... when a stern bookwormish professor lectures on how to engage with the community through social media, I'll hold court on what you *really* have to put up with when you stick your neck out online, and the kids will scribble furiously in their notebooks. It'll be hilarious!

    @Mike Brown, I can't answer your specific question about who will be generating those documents because I don't know product plans of other implementers. We decided that reading Strict documents was a priority to assure good interoperability with other implementations, including ours down the road. If you're eager to see Strict broadly used, as you seem to be, then I assume you'd agree that it's in the best interest of the broader IS29500 ecosystem for us to provide read support for both Strict and Transitional, just as we will be providing write support for both formats.

  • MSSUX

    4/7/2010 9:12:35 AM |


    Yes Doug....

    "We decided that reading Strict documents was a priority to assure good interoperability with other implementations, including ours down the road......"

    How far down the road Doug, your've decided that waiting 2 years so far is just fine.....

    So what is it going to be before it actually "gets read" and then 'acted upon" and then "delivered"?

    3 years... 4 years... 5 years... 10 years... 20 years...



    So you as the representative of Microsoft make haste not.

    While entire countries say, "Doug and his Microsoft standards?"

    "We will take Open Office thanks."

  • MSSUX

    4/7/2010 9:21:26 AM |

    Methinks Doug Bullshiteth Much.


    Google Search:

    microsoft loses sales to Open Office

    Microsofts thieving and lying and scumbag antics have come back to bite them on the arse.

    So now tell us all Microsoft Doug - exactly what is this ummm standard for and when?

    Perhaps thou buttplug bindeth too much.

    Etc.


    When Doug? When and what Doug? What's that Doug? What?????
    www.boygeniusreport.com/.../

    Microsoft loses patent appeal, Office and Word 2007 sales to be halted January 11, 2010

    Microsoft has lost its appeal in its ongoing legal battle with small Canadian company i4i over XML code within Microsoft 2003 and 2007. The loss upholds the previous decision which requires Microsoft to pay $290 million to i4i and forces Microsoft to pull Word 2007 and Office 2007 from the shelves effective January 11th, 2010. Panic and mass hysteria will not ensue as Microsoft has stated that it is in the process of removing the offending code and will have a Word 2007 and Office 2007 version ready for retail by the injunction date. Microsoft also confirmed that Office 2010 is not affected by this ruling and is expected to launch on time in mid-2010. The winner, i4i, was much more jubilant in its response stating that it “couldn’t be more pleased with the ruling”. A $290 million windfall from a few lines of XML? We would be well pleased, too.Read


    www.appleinsider.com/.../..._office_by_jan_11.html

    techblips.dailyradar.com/.../

    Etc.

  • Mike Brown

    4/7/2010 9:50:23 AM |

    @MSSUX,

    >> Mmmmm I think I will shut up after this....

    That's sound advice.  It's a pity that you didn't follow it.

    Foule-mouthed rants and name-calling don't help anybody. You lose the argument immediately, no matter what the merits of your case might be.



    @Doug,

    >> We decided that reading Strict documents was a priority to
    >> assure good interoperability with other implementations...

    Which is the big joke, of course.  There aren't going to be any other implementations!

    There is no incentive for "other implementers" to produce docs in the Strict version of OOXML until Microsoft Office does the same.

    Seeing as Office 2010 is not even out yet, and the gap between Office versions is usually 3 to 4 years, we're looking at 2013 at the earliest before that happens, are we not?









  • MSSUX

    4/7/2010 10:08:23 AM |

    @
    Mike Brown

    Ugghhh well - even you, as a add on to a committee that chased it's own tail and ended up counting for nothing. are entitled to an opinion.

    Exactly what is your point?

  • Rick Jelliffe

    4/7/2010 10:16:56 AM |

    MSSUX: http://www.linux-mag.com/cache/7439/1.html

    "I may make jokes about Microsoft at times, but at the same time, I think the Microsoft hatred is a disease." Linus Torvalds

    +1

  • Mike Brown

    4/7/2010 10:46:18 AM |

    @MSSUX,

    I think I made my point quite clearly: if you can't be civil then "shut up" (your words).

  • MSSUX

    4/7/2010 4:17:39 PM |

    @ Gay Mike - the Cheerful one.

    Well yeah you could have a point, but no one but you seem to understand it.

    I mean Linus Torvalds, who actually quoted me, after I had taught him how to create the Linux Kernal, - that statement that I made and illustrated to him, only makes sense in the context of what and how it was taught.

    Microsoft is a corporation and playing corporate kiss arse, is much the same as playing swallow the leader in forum group think.

    You know the "Two Legs Bad, Four Legs Good" bleat;

    Well for the uninitiated here, the "Hate Microsoft" epidemic tend to arise after much time after the average Windoze user starts to wake up to what an idiotic scam and imposition Microsoft is - and after the Windoze user becomes so insane from using Microsoft products that the just have to face a second learning curve and jump ship to Linux or even the inferior Mac OS, and discover how great it is just to not have to have to use any Microsoft products and the insipid bullshit that comes with it.

    It's at this point that the awareness arises at what a shitty imposition Microsoft is and the out pouring of wrath continues until anything other than a jolly pot shot at Microsoft is just a waste of time.

    I mean as an EX microsoft user of some 20 years... and a transitional into Linux for about 6 or 7 years., and a full time Linux user for 2 years...

    Well after trying to use MS XP to back up loads of data, from drive A to back up drive B, and constantly getting HI IQ interrupts like This Folder already contains MSTOSSA.jpg do you want to over write it?

    Every week I back up the same bunch of files the same way, with incremental changes to them and every week it asks the same few range of questions, but for different files....

    And YES TO EVERYTHING, isn't.

    The last 49 back ups, the file that can't be copied, was not too long.

    And this shit just goes on and on and on and on....

    I call Microsoft and say, "How do you switch these fucking idiot interupts off?" - They say, "I don't know". I ask, "Well can you put me through to the person who does?"

    They say, "You can't switch them off". I say "It's a program, someone wrote the code for it - that person can and will have the records on how to switch it off"

    Then they just bullshit, and stone wall and buck pass their way out of providing the real answer from the actual person who knows how to switch these idiot interrupts off.

    And Microsoft has closed down their "customer service" (?) in Australia and hired a foreign call center for the lowest possible price...

    Now the call center is staffed by totally clueless script monkeys...

    I mean shit begets shit begets shit - Bad OS's beget bad customer relationships, bad customer service begets bad customer relationships...

    Then toss in the fact that MS charges people in other countries twice as much for Win 07 for what they charge in the USA., and Microsoft rigged the shit out of the ISO certification, they have been nailed big time for ripping so many people off, and rigging the game for themselves.... and now  their ISO certification  program for their bogus standard - that is dead in the water.

    Microsoft is NOW in serious decline.......

    The fall of empires isn't like a big two story mansion crashing into the ground in one go, it's a slow process of being slowly being eaten away structurally by termites....

    Why do I hate Microsoft - for the fucking stupidity of the corporate mind think and their impositions.....

    I have tried to use their "MS Backup" program for all my important files and documentation" and that was just a total fuck up....

    It was like asking a cage full of rats to drive a car...

    So the ingrained safe habit of simply copying files as is from drive to drive was born.

    Using that ONE for instance of backing up a stack of data from one drive to the next - using XP and suffering a never ending stream of IDIOT interrupts and blocks half way through etc...

    Well when I copy the exact same files from drive A to drive B using Ubuntu Linux...

    The only thing that happens is that all the files get copied across.

    No interrupts, no stupid blocks saying the file name that I have copied across 49 times previously that year, is today too long....

    Then after years of enduring this shit from those idiots in Microsoft, then to find out that the scams and lies and games are just continuing right along...

    The software along with the common sense functionality, and the customer service is just getting worse and worse and worse...

    And the management of Microsoft - after rigging the ISO certification process to produce a bogus par baked standard is now just doing nothing to complete anything...

    That is just more people getting fucked around for nothing...

    More corporate moron bullshit....

    Microsoft is only going to get worse.....

    But I do enjoy getting stuck into Microsoft...

    And I do enjoy NOT having to use their software.

    I also feel so much better about exercising my decision to not be subjected to Propriety Product Lock-In....

    Good bye Microsoft and good riddance.

  • Jesper Lund Stocholm

    4/7/2010 4:25:49 PM |

    Hi Mike,

    Which is the big joke, of course.  There aren't going to be any other implementations!

    There is no incentive for "other implementers" to produce docs in the Strict version of OOXML until Microsoft Office does the same.


    I beg to differ. You need to consider the enormous ecosystem around Microsoft Office that are not Office suites but instead creates documents. For these applications it is essential that Microsoft Office can read the document but not necessarily important that Microsoft Office can save it again (as Strict). The same goes for OOo in its ecosystem - if OOo can't read your file, you're screwed.

    We (CIBER) are generating documents for both ODF and OOXML, and I can tell you for sure, that the minute Office 2007 and above can read Strict files, we are going to rewrite our code to - once and for all - get rid of all the legacy-crap we have had to create until now. We see this as a good thing.

    So prioritizing Read over Write seems like the only sane choice to me.

  • A. Rebentisch

    4/7/2010 6:06:01 PM |

    The sane thing to do would be Postel's law:
    Be conservative in what you do; be liberal in what you accept from others.

    In other words, don't generate new documents in transitional anymore, or at least generate "lesser" transitional documents to make the transition towards "strict" happen.

    Still no one ought to make "strict" the fetish as it is not great either, more a proposed file format as sanitized by ISO/IEC. Notably, the availability of all these documents mostly started only with Office07. As long as transitional documents are generated the compatibility fallacy would be invoked over and over again. We observed what pain it was to fade out IE6.

    The next big thing would be a filter which converts transitional and legacy documents to strict where it is possible without fidelity loss. The messier ISO/IEC Open XML gets, the brighter shines the star of ODF.

  • Mike Brown

    4/7/2010 6:36:41 PM |

    @Jesper,

    You're going to be generating Strict OOXML docs for MS Office to read but not edit and save?  Hmmm... wouldn't such documents be better off as PDFs?

  • Jesper Lund Stocholm

    4/7/2010 7:04:18 PM |

    Hi Mike,

    You're going to be generating Strict OOXML docs for MS Office to read but not edit and save?  Hmmm... wouldn't such documents be better off as PDFs?

    The files will be perfectly editable by the user opening them - that is the whole point. We'll generate documents to plug into the existing flow of work for our customers. These include text-documents with change-tracking, drawings etc and also spreadsheets with business-driven data. These data should be processable by the consumer (sorting, filtering, adding charts etc) so therefore PDF is not an option.

    But having Microsoft Office read Strict files enables us to create cheaper solutions for our customers since we don't have to deal with all the legacy stupidity of the past when creating our documents.

    I often come across solutions where "Export to spreadsheet"-functions are essentially a CSV-file targeted e.g. Calc or Excel. To the user it opens "a spreadsheet" with all functionality of the application available -  but in reality the data itself is much simpler.

    Creating Strict files for Microsoft Office is essentially doing the same thing.

  • Downunder

    4/7/2010 7:27:44 PM |

    @Alex, in reference to your second comment in the discussion, I see that Microsoft has made a specific commitment now to support 29500 strict.

    '[Microsoft] will support Strict no later than Office “15.”'

    blogs.msdn.com/.../...or-iso-iec-29500-strict.aspx

    I'm just highlighting this in case @Doug Mahugh's terse note about his blog post got lost in the fascinating debate above.

  • Mike Brown

    4/7/2010 7:41:32 PM |

    @Jesper,

    >> The files will be perfectly editable by the user opening them

    .. and, presumably, saving them afterwards.  At which point they'll get an Office dialog, warning them about a potential loss of features/fidelity (or however Microsoft phrases it) because Office can't save Strict OOXML files.  Or am I missing something?



  • Jesper Lund Stocholm

    4/7/2010 7:55:49 PM |

    Hi Mike,

    .. and, presumably, saving them afterwards.  At which point they'll get an Office dialog, warning them about a potential loss of features/fidelity (or however Microsoft phrases it) because Office can't save Strict OOXML files.  Or am I missing something?


    I have absolutely no idea - I have not tested their read-functionality. However, since T is currently a superset of S, there should be no reason do display a warning since all functionality of S is contained/covered in T as well.

    Smile

  • A. Rebentisch

    4/7/2010 10:42:33 PM |

    However, since T is currently a superset of S, there should be no reason do display a warning since all functionality of S is contained/covered in T as well.

    Exactly. In other words, OOXML-S "read support" is a two-liner.


  • Peter

    4/7/2010 11:38:37 PM |

    If MS Office can (eventually) save in Strict, does this mean building 100% interoperable read/write applications will be possible?

    When I start up OpenOffice 3.1 Writer and I do a Save As... and choose Microsoft Word 2007 XML, to what standard (if any) am I using?

  • Jesper Lund Stocholm

    4/8/2010 1:43:45 AM |

    Hi Mike,

    To put it another way, do you know anybody that can write but can't also read?  The former implies the latter, does it not?

    That might be true in a "helicopter-view" or on management level, but when it comes down to implementing, "read" is not inferred by "write" in any way. The code doesn't write itself, you know.

    Smile

    So on that basis, yes, you should have prioritised writing Strict documents.

    Are you seriously suggesting that you'd prefer a situation where Microsoft Office can save a document in a format that it cannot open?

    The good thing about supporting read mode for strict documents is also that the good guys from IBM or ORACLE can start implementing strict in their applications. Wouldn't it be fun if they could taunt Microsoft with allowing read- as well as write of the document format in Microsoft Office - before Microsoft could?

  • Mike Brown

    4/8/2010 9:43:01 AM |

    @Jesper,

    >> Are you seriously suggesting that you'd prefer a situation
    >> where Microsoft Office can save a document in a format that
    >> it cannot open?

    Read my post again, Jesper.  The "on that basis" part that you quoted refers to my assumption that being able to write certain a file type implies that a system can read it too.  You've even quoted my assumption further up in the same post!

    You're free to disagree with that assumption, of course - in fact, you already have... I think.


    >> since T[ransitional] is currently a superset of S[trict],
    >> there should be no reason do display a warning since all
    >> functionality of S[trict] is contained/covered in
    >> T[transitional] as well.

    So, as an Office 2010 user, I receive a Strict OOXML document from somebody.  I open it, edit it and then save it.

    It's now a Transitional OOXML document, and yet the user is to have no warning of this change?  Even if, as you say, no actual data is lost because Transitional is a superset of Strict, I'd still argue that there's been a significant change here.

    That change would become rapidly apparent when I send it back to the originator, and their chosen software read and writes Strict OOXML but not Transitional.  Even if their software does deal with both kinds, they may not be happy to be sending Strict docs out and getting Transitional ones back from me.

    It's all a bit messy.









  • W^L+

    4/8/2010 12:18:25 PM |

    @Alex The complaint of this blog posting is not about OOXML itself, but of Microsoft's decision to implement T (only) in a new Office product. T should be a dead language, like Latin. The International Standards community did not want it be to used for new documents.

    What might be useful for T is a phased deprecation plan, to make its purpose more obvious to everyone. But right now I think the correct approach (for us standardizers anyway) is to introduce conformance labelling provisions in the standard so that implementations had to make it very plain exactly what kind of OOXML they supported. People procuring systems, armed with the information, will be empowered to make more a informed choice.


    Clearly stated (and labeled) conformance is a good first step. I'd also like to see some interop testing / certification. I appreciate what Doug is saying about prioritizing read functionality for OOXML Strict. At that point, it would be interesting to see just how compatible implementing applications really are.

    @MSSUX You aren't helping here. People like me are working from inside big, tradition-bound organizations to try and open the door to ODF-supporting applications, and all it takes is some policy-writer to read a site like this full of your foul-mouthed rants and Ubuntu fanboyism and just like that, years of effort are out the window. This isn't Slashdot. These are grown ups who have jobs.

    @Doug is there any plan to improve your ODF interoperability, so MSOffice's ODF files work similarly with OpenOffice / StarOffice / Lotus Symphony / KOffice / AbiWord? That's really where I'm interested.

    And (please) make it easier to select file formats on the fly when saving. Even support staff in my agency find it difficult to do with Office 2007.


  • Rick Jelliffe

    4/8/2010 12:20:41 PM |

    Mike: The industrial publishing industry has a different set of use cases and processing workflows than desktop application users.

    For example, in Australia Department of Health and Agin generates the Legislative Instruments for various schedules of the Pharmaceutical Benefits Scheme by pulling in data from various databases, text databases and spreadsheets. The information is published in various selections, sorts and formattings to XSL-FO->PDF, DOCX->Office&OpenOffice, XHTML (3 different websites) and .mobi and .RDF.  

    While XSL-FO is the preferred form, since it gives better access to professional typographic features, notably for complex tables and print publication, the reason DOCX is used is that it does give a chance for last-minute editorial/production changes at the DoHA side: however, this is only an emergency (that has never been used AFAIK.) It would be trivial to add ODF support if that became a requirement too.

    So the idea that you have applications that write but don't read the same format, or read but don't write the same format, is in fact the *typical* situation in industrial publishing with markup. It is good if there are desktop application stakeholders at the table, but their use-cases are not the only ones, nor *more* important than industrial publishing industry's uses-cases. We want to preserve information through a flow, be able to augment information in a flow at any time, and we want to reduce the deadening effects of Word on any workflow system, since our customers often have adopted it and are seeking to extract value from their investment, wise or not.

  • Jesper Lund Stocholm

    4/8/2010 12:26:41 PM |

    Hi Mike,

    The "on that basis" part that you quoted refers to my assumption that being able to write certain a file type implies that a system can read it too.  You've even quoted my assumption further up in the same post!

    Yes - yesterday it just suddenly ocurred to me what "write over read" would actually mean - hence my reply to you.

    Smile

    It's now a Transitional OOXML document, and yet the user is to have no warning of this change?  Even if, as you say, no actual data is lost because Transitional is a superset of Strict, I'd still argue that there's been a significant change here.

    Yes, I agree.

    That change would become rapidly apparent when I send it back to the originator, and their chosen software read and writes Strict OOXML but not Transitional.

    I don't think there will be many applications out there trying to compete with Microsoft Office that won't be able to read T documents. So for competitors like IBM, Google, ORACLE, Apple etc - read-support of strict files doesn't change anything. But the applications of those vendors already support the featureset of T in their support for the binary document formats.

    Even if their software does deal with both kinds, they may not be happy to be sending Strict docs out and getting Transitional ones back from me.

    It's all a bit messy.


    Yes - I agree. All I am saying is that given Microsoft had to make a prioritization of read/write, "read" is the only one that makes sense to me.

    Smile

  • Segedunum

    4/9/2010 2:39:52 AM |

    @Gareth Horton: "working well with them on Excel compatibility since 1991"

    Just curious Gareth. Compatibility with what, exactly?

  • Marcello Romani

    4/9/2010 5:39:52 PM |

    Mr Brown,
        Microsoft has always acted against international and open standards to maintain its de-facto and not-100% fully interoperable formats. It has done so in the Office world, where each new version was made intentionally slightly incompatible with the previous one to force users to upgrade (read pay). It has employed a similar strategy with Active Directory, taking the kerberos specs and "tuning" them in such a way that running a multi-vendor AD environment is virtually impossible. It even tried to "microsoftize" the Internet with IE6, which is still causing headaches to web developers (read loss of time and money) because of the quirks in its CSS implementation.
    The examples could go on and on.

    How could anyone believe Microsoft would abandon its well established, de-facto (non-)standard "office" format in favour of an international standard that anyone could implement ? It simply doesn't make sense.
    The entire ISO 29500 story demonstrates that Microsoft has once more tried to disrupt international established, open, good-will procedures just to have an economic advantage.

    It's so sad that ISO hasn't been able to protect itself (and its credibility) from such attacks.
    I think 29500 should be declared invalid and deprecated since no one (not even its main sponsor) is able or willing to implement it. Moreover there's already a good, international, accepted and currently implemented standard, ODF, that does well today what OOXML was supposed to do.

  • Carlos González

    4/10/2010 4:37:23 AM |

    So... beside all the justifications, excuses, and half-baked attempts to bend the true....

    Is OOXML still an ISO standard without a reference implementation and it will remain so for a long time? Because frankly, all the rest is irrelevant, tasty yes, but irrelevant nonetheless.

  • Stephen Walli

    4/10/2010 6:45:25 AM |

    @Carlos:  Reference implementations are tricky things.  If there's a reference implementation, do you trust the reference implementation or the specification when you encounter a difference?  The correct answer is of course it depends.  But generally you need to trust the spec and then fix it (and possibly correct the reference implementation). This takes time.  Think of the IETF process (a good one) where there needs to be two genetically independent implementations -- not necessarily open source -- that communicate before the DIS becomes a full Internet Standard.  

    Good standards that came from a wide body of implementation experience prior to standardization tend to need more negotiation during the standards process, and the implementations [possibly] need to change to conform to the standard.  (Think UNIX systems and the POSIX stds process.) A situation like ODF where the spec came from an open source OO.o implementation, the standardization itself changed the spec, and the reference implementation then changed to meet it, is another situation.  

    But I'm sure Msft would love to claim that MS Office is the reference spec.  You don't want to be in the situation where another conforming implementation (if there ever is such a thing) needs to change to suit a particular closed product, or worse yet an old revision of a closed product.  Bad karma.

  • Carlos González

    4/12/2010 2:14:58 AM |

    @Stephen: I agree with you, every standard that is born from an implementation, which I think are all but OOXML, go trough painful phases of adjust well from the implementations or from the standard itself trough the standardization process.

    But really, I could not think of any other "standard" that goes trough a fast-tracked process of standardization before having at least ONE implementation, and probably none has gone 2 years without implementation having an ISO stamp on it.

    But hey, I guess this is normal, like Alex said a lot of times.

  • Rick Jelliffe

    4/12/2010 2:32:27 PM |

    Carlos: Sorry, you are talking through your hat.

    Many and perhaps most standards do not have one 100% complete implementation at the time of standardization. Would you exclude ODF on those grounds?

    Most standards bodies do not require or allow reference implementations, to prevent ambiguity. Would you exclude OASIS on this ground?  ISO and JTC1 have no procedures in their Directives to allow reference implementations as conventionally thought of, and certainly SC34 has no history of supporting them.

    However, many standards are based on documenting existing technologies, warts and all, but with rough edges cleaned up. Take the Linux ABI as an example. Or PDF.

    Take W3C for example: their "call for implementations" is not a call for multiple complete implementations, or even a single complete implementation: it is at the *feature* level. If a feature has absolutely no implementation, it will be removed. The changes in ISO OOXML are almost entirely syntactical changes or schema-level changes: there is *no* change in the feature set is there? So OOXML clearly has the implementation experience that would satisfy W3C's guidelines at least.   See
       http://www.w3.org/2005/10/Process-20051014/tr

    Where a standard does not have a conforming implementation merely due to trivial-to-fix syntactical differences, which is largely the obvious approach is to tweak the syntax where the difference is not important, and to help get more active agreement where the difference is in some way important. If your answer to every problem is "Off with their head" then ISO is not the organization for you.  

  • Segedunum

    4/12/2010 8:06:29 PM |

    @Rick: Sorry, you are talking through your hat. Many and perhaps most standards do not have one 100% complete implementation at the time of standardization.

    I think we've been through this over many years Rick. If you don't have a standard and a set of specifications that proves itself to be implementable then you have a problem. Either the standard is totally useless or you need to go back to the drawing board, find out why and do something about it.

    Most standards bodies do not require or allow reference implementations, to prevent ambiguity.

    Totally and utterly irrelevant, but then, you know that already don't you? What we have here is a standard that after several years of bashing away has no verifiable implementations whatsoever - reference or otherwise.

    Whether the ISO or any other standards body requires a reference implementation or not is utterely irrelevant. Standards exist to be implemented, used and be useful. If they can't then they're useless. That's exactly why Alex has ended up posting this and why he's ended up talking about what it is that Microsoft Office is actually implementing, or not implementing and talking about verification checkers.

    If you don't know the importance of validation checking when it comes to standards and specifications, regardless of whether a body requires them or not, then..........you're away with the fairies.

    Would you exclude OASIS on this ground?

    What? Stop talking about OASIS by desperately trying to create parallels between it and OOXML. It's utterly pointless and proves nothing. This has nothing to do with OASIS, which has several implementations incidentally.

    The rest of your comment is non-sensical twaddle that seeks to desperately hunt around for examples, or rather, excuses, as to why OOXML is completely, totally and utterly practically useless as a standard. Alas, all of the 'examples' you give have practically working examples, and that has clearly pushed along development of various standards.

    I congratulate you Rick. The mental gymnastics you've forced yourself through with this would have driven most people insane by now. I hope it's been worth it.

  • Alex

    4/12/2010 10:10:37 PM |

    @Segedunum

    > Standards exist to be implemented, used and be
    > useful. If they can't then they're useless.

    That's not quite right. Quite a few "aspirational" standards have never been properly or fully implemented/used (OSI, SGML , ODA and HyTime all fail these tests in some way). Yet the work embodied in them was incredibly useful in spawning future standards. I think there is a place for such standards.

    However OOXML is not an such an aspirational standard and the Transitional variant has (setting aside defects) been implemented in many applications -- even OO.o will read OOXML files! Most implementations of OOXML are not office suites, but business systems processing MS Office documents in some way. There is an enormous and valuable ecosystem there.

  • Stephen Walli

    4/12/2010 11:48:22 PM |

    Gents, I would argue that standards exist to encourage and enable multiple implementations that interoperate in some fashion. This applies to things as complex as software and operating systems behaviour, and as simple as hardware and a threaded nut and bolt.   Conformance to the standard's normative specification is everything.  Standards define their own conformance criteria to allow people that care to judge the quality of an implementation claiming conformance.  

    This then allows one to judge the quality of a standard:

    First, is the specification good enough to enable a conforming implementation to be built against the entire conformance criteria?  [Saying that the standard's specification came from an existing implementation therefore there's an existence proof is NOT the same thing.  Saying lots of people "support" the standard, without anyone fully conforming is a useless exercise.]

    Then, are there multiple implementations.  

    Lastly, do outputs from multiple implementations behave correctly, i.e. work on other conforming implementations as required by the standard specification and conformance criteria (and hopefully according to expectations by users of the standard).

    @Alex: I appreciate the work of OSI et al were useful educational experiences for standards developers.  However, they were never widely implemented and participants (and the implementers) probably still feel the disappoint and sting of them as the failures they were.  A lot of hard work went into those lessons.  

  • Rick Jelliffe

    4/13/2010 12:11:22 AM |

    Segedunum: I was responding to Carlos, who raised the issue of reference implementations, who wrote that without a reference implementation all the rest is irrelevant. It was an issue that had been previously brought up, and which Alex had dealt with. If you don't like the issue being brought up, then rant at him not me.

    As to the supposed lack of implementation, in Alex's recent smoke test, testing a 60mb OOXML file using the Transitional schema, Alex got 84 validation messages, all of the same type, where there was an element with an attribute using "on" rather than "true". Do you really want to back yourself into the corner where you are claiming that this is something that can never be fixed, and fixed trivially, and that it demonstrates the impossibility of implementing OOXML?  

    (Actually, Alex' test was of files that Word generated. It would be more to the point to see whether Word accepted a version of the file with those values corrected.)

  • Segedunum

    4/13/2010 12:27:28 AM |

    @Alex: That's not quite right. Quite a few "aspirational" standards have never been properly or fully implemented/used (OSI, SGML , ODA and HyTime all fail these tests in some way). Yet the work embodied in them was incredibly useful in spawning future standards. I think there is a place for such standards.

    The ultimate end goal in either case is to a get usable, practical standard that can be usefully implemented - however far down the line that might be. The closer the better. If you don't get to that point with the first iteration then you need to sit down, take stock and ask why that didn't occur. We're at the stage now where OOXML has failed as a standard that is practical and we need to take stock and ask whether further iterations are possible to get it into that state.

    I'm afraid covering up by saying that it doesn't matter if a standard cannot be used won't save you.

  • Rob Weir

    4/13/2010 12:45:06 AM |

    Alex, no doubt there was a large and valuable market around Microsoft Office well before ISO approved OOXML, and it would have existed even without ISO's approval. Remember, OpenOffice and others also read Microsoft binary formats, though these are not standards.

    So the question is: what tangible net benefit as accrued for anyone other than Microsoft for OOXML's approval as an ISO standard, over the alternative of Microsoft simply publishing their file format documentation privately, as they did 10 years ago with their binary formats?

    Arguing that the Office market is large or that the same apps that implemented the binary formats also implement the XML formats is rather weak, as I'm sure you can see.  If you followed that argument you would be starting a new initiative for standardizing the binary formats, since they are much more widely used today that OOXML.

    JTC1 Directives says, "The social and economic long-term benefits of an IS should justify the total cost of preparing, adopting and maintaining the standard."

    So what are the benefits of OOXML as an IS over a set of EC-mandated technical disclosures on Microsoft's web site? I'd like to see that argument that the benefits justify the costs, especially if, as you suggest, the standard is not being adequately maintained, and no one implements the Strict variant.  Or is the argument that as the investment in maintenance shrinks to zero, then the threshold for social or economic benefit also shrinks to zero?  That would be an interesting interpretation.

  • Segedunum

    4/13/2010 1:11:45 AM |

    @Rick: ...who wrote that without a reference implementation all the rest is irrelevant. It was an issue that had been previously brought up, and which Alex had dealt with. If you don't like the issue being brought up, then rant at him not me.

    I had a rant at your totally inadequate and rather pointless answer. Carlos was talking about why a standard would be fast tracked when it doesn't even have one provable implementation to work from to improve said standard and garner meaningful feedback. You keep trying to compare this to ODF and there is no comparison whatsoever there.

    As to the supposed lack of implementation, in Alex's recent smoke test, testing a 60mb OOXML file using the Transitional schema, Alex got 84 validation messages...

    I have no idea why we'd be interested in the transitional schema. So not even the transitional schema, compatibility with which should have been sorted over three years ago (and was roundly rejected), actually works in practice? Brilliant. Grab yourself a cookie.

    Do you really want to back yourself into the corner where you are claiming that this is something that can never be fixed, and fixed trivially, and that it demonstrates the impossibility of implementing OOXML?

    Do you have a firm date when this will be trivially fixed?

    Then fix it already. Talk, and words, are cheap. Either fix it, get someone else to fix it or get the ever so helpful Microsoft to come in and do the Right Thing(tm). Everyone has been waiting for well over three years for stuff like this to be fixed. Nothing has happened. That's what has ultimately culminated in this blog post.

    Painted in that light, the notion that this will be trivially fixed (and that's just what you say has been found in one file validated against transitional) is utterly laughable, and there's only one group of people who have painted themselves into a corner there.

    (Actually, Alex' test was of files that Word generated. It would be more to the point to see whether Word accepted a version of the file with those values corrected.)

    Errr, no. The support needs to work both ways. It needs to accept files formatted correctly and also generate correctly, otherwise all that will happen is that a 'correct' file opened and re-saved in Office is likely to be incorrectly formatted.

    Anyway, it'll be nice when a day comes some time in the next ice age where we have an alternative and well used implementation that will generate such files to provide us with that cross-validation that other formats like ODF have had for some time. However, we don't have that, do we?

  • Alex

    4/13/2010 1:36:18 AM |

    @Rob

    I'm wasn't arguing that the size of the Office market justified the existence of OOXML as a Standard (though that case could be made), I was simply countering the assertion that it was unimplemented.

    It remains to be seen what the cost/benefit ratio is for the OOXML project is. Certainly many of the heavy-weight policy influencers must have reckoned it would come good: in your own country (for example) see how at JTC 1 level you've got the Departments of Defense and Homeland security lining up with many big corporates to approve it.

    My own past experience has been that allowing Microsoft to control the Office formats has been damaging, as unexpected changes to the formats have bad impacts on systems that rely on them. Thus having them maintained with greater transparency and broader stakeholder input using the International Standards framework makes sense. I certainly don't think it would be wise to start cultivating another European-level class of quasi-standards instead.

    We'll have to wait and see what the new few years bring. One interesting touchstone issue is whether the current initiative to bring better CJK typographic support to OOXML (and ODF too of course) will bear fruit. If future software appears that both implements the Standard and supports this kind of Internationally-sponsored initiative then things will have moved in a very positive direction. But right now that's still a big if, of course.

  • Rob Weir

    4/13/2010 8:26:28 PM |

    Alex, I think with a little reflection you will agree that the US Department of Homeland Security and the Department of Defense participated neither in the development nor the maintenance of OOXML.  They, like the NBs of Kazakhstan, Syria and Malta, voted Yes on the Fast Track ballot and then were never heard of again concerning this project.

    So I'd put the question to you again, as someone who actually is participating in the maintenance of OOXML in WG4.  What is the return on invested time here over what you would get if OOXML was just 8,000 of EU-mandated technical disclosures released on Microsoft's web site?  Based on what you are seeing, why is this worth your time?  I'd be interested in knowing, since you are one of the few people in WG4 who is not a Microsoft employee.

    And do you seriously believe that Microsoft is not controlling OOXML presently?  To be clear, there is ISO OOXML and then there is MS OOXML.  They match in part, but differ in detail and appear to be diverging.  You observed that "allowing Microsoft to control the Office formats has been damaging, as unexpected changes to the formats have bad impacts on systems that rely on them".  OK.  True.  But isn't this most relevant to MS OOXML?  And isn't MS OOXML still controlled by Microsoft, who is free to subset and extend ISO OOXML freely? And I don't think we can truthfully say that MS OOXML is maintained "with greater transparency and broader stakeholder input".

    It looks very much like with ISO OOXML you have the shadow of a tiger in a cage and think you've tamed the tiger itself.  But the tiger is still free to move and will do so when it is hungry.

  • Rick Jelliffe

    4/13/2010 8:29:25 PM |

    Segadunum: Why make blanket statements, then complain that I am quibbling when I point out that they are too much? For example, you make statements about lack of implementations, then it turns out you are tacetly excluding Transitional.  

    Suppose I point out that your argument -- that there should have been an implementation of Strict (or are you including Transitional again?) at the time the standard was fast-tracked -- is odd, because it would involve requiring some kind of telepathy for MS to know ahead what the BRM would come up with. I suppose you will then claim that you really only meant in the period since the BRM.

    If you stuck with less fabulous comments, not re-hashing the failed tirades of 2007, you would find a lot more agreement.

  • Istvan Sebestyen

    4/14/2010 6:27:44 PM |

    Dear Alex,

    In your March 31, 2010 blog on OOXML you wrote the following:

    “Most worrying of all, it appears than Ecma have ceased any proactive attempt to improve the text, leaving just a handful of national experts wrestling with this activity. It seems to me that Microsoft/Ecma believe 95% of the work has been done to ensure the standard is “useful and relevant”. Looking at the text, I reckon it is more like 95% that remains to be done, as it is still lousy with defects.”

    Then…..

    “Moving forward:
    …….
    •  Ecma need to commit adequate resources to standards maintenance and pro-actively seek to improve the text, working together with SC 34, if there is any appetite to improve the Standard to the point where it can be a trouble-free, or even good, basis for interoperable office applications. “


    I am confused by the above statements.

    As you know - under your chairmanship -  we have had on July 21/22 2008 at the British Library Conference Centre in London, a meeting of ISO/IEC JTC 1/SC 34/Ad hoc 1 on ISO/IEC 29500 Maintenance procedures. There the JTC 1 SC34 part declared that they wanted to have full control over the maintenance of ISO/IEC 29500. Ecma agreed to this and Ecma has committed itself to have an active delegation of experts in the SC34 process, and take over its final result for synchronization with the Ecma version of the standard (as all the subsequent versions of ISO/IEC 29500 and ECMA-376 subsequent editions have to be in synch). As a result of all this, in December 2008 we have approved ECMA-376 Edition 2, besides the original ECMA-376 Edition 1 standard (in Ecma both Editions are in force currently for market reasons). We intend to synchronize with JTC 1 any subsequent future improvements of the standard.

    Now I am surprised to hear that according to your views the Ecma delegation has not  done its duties properly?

    As far as I know we constantly have had in the SC34 WG4 meetings one of the largest and still most active delegations and Ecma has been submitting Defect reports into the SC34 WG4 process on a regular basis. I have not heard any official complaints about the quality of co-operation of our delegation with JTC 1 SC34.

    I am surprised on the tone of your blog because ECMA-376 is on the Ecma side one of the most popular standards that we currently have. In terms of standards download it is constantly on the position 3 or 4 in the list of most downloaded Ecma standards. Several thousand the various parts of the ECMA-376 Edition 1 and Edition 2 were downloaded in 2009, and the figures for 2010 so far are very similar. Both Edition 1 and Edition 2 are about equally popular in the downloads.

    Regarding implementation of an Ecma standard par. 9.3 of the Ecma Bylaws say the following:
    “It is not mandatory for Ecma members to implement any Ecma standard”.

    Well, this type of rule is true for any other SDOs as well. The reason is that rather often standards approved are not being implemented in practice at all for various reasons. However, in this particular case this does not appear to be the case: ECMA-376 Edition 1 has been implemented many-many times and  by many implementers, we see a clear roadmap for the implementation of ECMA-376 Edition 2, but when and how that is implemented in not an Ecma decision. That is clearly out of scope for standardization.

    Kind regards,
    Istvan Sebestyen
    Secretary General of Ecma International

  • Alex

    4/14/2010 8:55:58 PM |

    @Istvan

    What I wrote was that it appears Ecma have "ceased any proactive attempt to improve the text, leaving just a handful of national experts wrestling with this activity".

    I'm not sure whether I'd characterize that as Ecma not doing its "duty" - under-resourcing is almost a natural hazard in standardization. And I have no quarrel whatsoever with the Ecma delegations to WG 4 (the "delightful, talented and diligent people" I referred to). But I would certainly repeat that we do not have "healthy maintenance" of this Standard - the size of text we have and the poor state it is in require a much greater level of resourcing than we currently have. Ecma TC 45 is a ghost of its former self, and I see Ecma as being the conduit through which corporate representation needs to take place. (So, to be strictly fair, this is not so much Ecma's problem as an SDO, as a problem of its corporate members in not committing sufficient resources for Ecma to perform adequately in its critical role as liaison).

    As you may be aware, Ecma 376 as published deviates in many respects from purported implementations of it (most notably, Microsoft Office). This is just one of the many categories of problem that WG 4 experts are having to understand and resolve, and why the resourcing problem is critical.

  • Alex

    4/16/2010 12:36:10 AM |

    @Rob

    I didn't say the DoD (etc) are participating in the maintenance of OOXML; yet they approved it. That's just how it works *shrug* -- techies and standards people on the ground to some degree have to deal with what the policy-influencers decide.

    If Microsoft ends up implementing some kind of non-standard parallel version of OOXML (what you refer to as "MS OOXML") then that is their business, and not something standards bodies can (or should) control. In this case the ROI of the SC 34 effort is definitely questionable. Whether it matters or not is up to the users who consume, and care about, document formats: broadly, the market. Maybe we'll find ourselves in a situation (e.g. as we found with C++) where an implementation in the ball-park was good enough for users not to care. But I personally think not: the whole document format debate is moving towards an emphasis on conformance testing and interoperability -- so my suspicion is that some key procurers will care about standards conformance.

    I don't think it's quite accurate to say that I am "one of the few people in WG4 who is not a Microsoft employee" - but even if I was, so what? I have said it before and I'm happy to repeat: it's a good thing that Microsoft send people to participate in a standards committee. Technical and standards experts are particularly valuable. More please! As you know, decisions in International Standardization are taken at a National level: MS and Ecma never vote on anything. And anybody who thinks a high proportion of MS people on WG4 would somehow make it answerable to Redmond hasn't been attending very closely to the business of WG 4 as it happens ...

    However I don't think anybody involved in under the illusion that Microsoft has been "tamed" by the standardization of OOXML, or that standards bodies have the power (or even the desire) to "tame" any large and powerful corporation. I think a much better way of seeing it was articulated by Rick earlier on here:

    "It is because of standardization that we can hold OOXML and ODF implementers up to objective scrutiny about their actual support for open formats: in particular using the objective metric of validation." [my emphasis]

  • Rob Weir

    4/16/2010 1:42:50 AM |

    Alex, you have not answered my question.  What is the value of standardizing OOXML over Microsoft simply publishing their own technical disclosure on their format?

    As you know, SC34 does not do conformity assessment, so Rick's statement about holding up implementers "to objective scrutiny about their actual support for open formats" is bogus.  SC34 can't do that.  Maybe you or Rick can play the scourge on your blogs.  But that is not an ISO/IEC standardization activity.

    Finally, I wonder what would be satisfactory evidence to you that Microsoft was controlling OOXML?  I wonder if you are clinging to a non-falsifiable view of the matter?  Can you even admit the theoretical possibility that a corporation, by joining a large proportion of NBs and attending all NB meetings, and sending a large delegation to all SC34 meetings, and in many cases being the sole Head of Delegation or representing the majority of the meeting, while at the same time dominating debate within the NB via their partners, that doing these coordinated activities can give that corporation control over a standard?  Is your judgment so clouded that you cannot see that possibility?  Or do you just believe that these things have not taken place?  Same question to Rick.

  • Jesper Lund Stocholm

    4/16/2010 2:02:42 AM |

    Hi Rob,

    Alex, you have not answered my question.  What is the value of standardizing OOXML over Microsoft simply publishing their own technical disclosure on their format?

    If we for a minute pretend that Microsoft had never submitted OOXML to neither ECMA nor ISO ... if you really think that the documentation would have been as good (ehm) as it is today, that other people, as their competitors, than Microsoft would have the slightest amount of control of how the document format would evolve and that the process of bug-fixing, correcting and amending the text would be done in a way even remotely as transparent as it is done today - then you clearly are a bigger idiot than you seem to think I am for participating in WG4.

  • Alex

    4/16/2010 2:29:54 AM |

    @Rob

    > Alex, you have not answered my question.  What is the value of
    > standardizing OOXML over Microsoft simply publishing their own
    > technical disclosure on their format?

    I thought I answered it twice. I repeat: if the Standard is adhered to then the advantage of standardizing it is in both the improving quality of the written spec and the admission of new features which would probably not otherwise appear (e.g. CJK typographic support); if the format is not followed then the value is questionable.

    On testing, you are correct that JTC 1 does not itself perform conformance testing; nobody said it did, and I acknowledged this earlier on in these comments. Yet SC 34 can produce conformance testing suites -- and I fully expect conformance testing initiatives to appear over the next few years which have rather more weight even than my or Rick's blog!

    > Finally, I wonder what would be satisfactory evidence to you that
    > Microsoft was controlling OOXML?  I wonder if you are clinging to
    > a non-falsifiable view of the matter?  Can you even admit the
    > theoretical possibility

    Sure I can "admit the theoretical possibility" - conspiracist though it may be. What I am saying is, that in actual fact, this view is wrong.

  • Rob Weir

    4/16/2010 9:47:35 PM |

    @Jesper, yes, actually I think that any concessions Microsoft has made in terms of technical documentation was done in response to the EC's antitrust investigation.  ISO was irrelevant and remains irrelevant in terms of how Microsoft publishes their formats and interfaces.  Two points of evidence for this.  1) Microsoft reversed direction and decided not to submit their XPS standard to ISO, as they had earlier planned.  It simply is not worth it.  2) Their relative disengagement in WG4 is coincident with the settlement of their EC antitrust investigation.

    @Alex, adding additional features, like CJK or ISO 8601 dates, or Strict conformance classes, Microsoft does not implement is irrelevant. I don't see any value there, certainly none corresponding to the multi-year effort consumed in creating this standard.   Maybe if you think of a standard as a work of literature, to be read and appreciated for its fine language, then you may be satisfied by the "improvements" in OOXML.  But in the real world it is seen as a colossal failure.  

    And you still haven't answered my question as to the value of standardizing OOXML.  All you've done is speculate that a possible amendment, sometime in 201-12, might be relevant to a possible version of MS Office sometime in 2015-16.  That does not quite make a great story for the value of SC34.

  • Rick Jelliffe

    4/17/2010 12:36:11 AM |

    The value to my employer is that we now have three standard languages in the industrial publishing output space that we can (and do) use: XSL-FO, OOXML and ODF. In the SGML days, when we would have to choose systems like FrameMaker MIF because it was easiest to convert to, desktop technology was blocked off from us.

    Now, because of standard XSL-FO we can do quite complex typesetting (not as free as in the 1990s though), because of standard ODF we can access a certain range of mid-range features, and because of standard OOXML there is no part of Word that is blocked off from us, coming or going.

    One thing that the JTC1 process gives is it allows broader participation and broader review. (In the case of XSL-FO it has come back-to-front, since XSL-FO is mainly the ISO DSSSL team grafting what they learned in DSSSL and its review into the WWW world!) When you look at the review comments for ODF 1.2, it is striking the number of them that come from the ISO community.

    To say that ISO standardization achieves nothing would be as ludicrous as claiming that it is a game-changer. But in fact, that is what Rob has managed to claim both at the same time: standardization achieves nothing, he says on one hand, and yet if that is so why did he spend so much of his time a couple of years ago touring around the world saying it was really important that it should not be an ISO standard?  

    The answer of course is that Rob is against broad participation, unless it is primarily by large companies and he is not keen on review by outsiders: consumers should be passive and content with what our betters give us: we should all march for Document Freedom Day but not actually participate on technical issues; the vendors know what is good for us, that is their job.

    The thing is that OOXML can be a success even if it benefits any large vendor. They are not the kings of the world. Small companies like my employer want Microsoft and other vendors to be channeled into making systems that are more useful for us, by whatever means. Standards are just one way.

    Given the unwillingness of the courts to split Microsoft up, the inability of anti-trust officials to deal with bundling, the insincerity of Microsoft's competitors who are hooked into their own monopolies and patent-mania, the inability of corporate-lead open source efforts to make compelling desktop applications (Firefox excepting!), the reliance of many vendors on SOE ideas that promotes concentration, and the slow death of some alternative technologies (i.e. Java) under bogus and irregular consortia, the ISO effort looks like one of the few areas where positive steps for more openness has actually paid off--to a small extent with much better documentation (2000 unedited pages to 8000 reviewed pages! come on: is that really noting?) And in the medium term to have Strict *if* Microsoft implements Strict for write as well as read.  

    One of the most *successful* standards has been ISO SGML. After 2 years of its development (by SC34's predecessor with championing by IBMers) and release, in 1988 is was barely used anywhere. After 7 years it had found certain niches in certain industrial publishing industries, but compared to WYSWYG it looked a complete failure. Then along came HTML. And the SGML profile called XML. And now have a standard generalized markup language with rigorous markup, all the technologies or methodologies pioneered by SGML are ubiquitous, from the remote control to the ATM.

    You tell a standard's success from after its direct influence has run its course (as has ISO SGML's).  

    Or take ISO Schematron as another example. It was made in 2006. By 2008 it was pretty hard to find projects that had used it. In the last few weeks I have been nosing around and it is used far more than I expected. We had a farewell party today for a guy in our office, and he said that he had been interviewed by a company I have never heard of who use Schematron.

    I realize that having a calm long-term view goes against megalomanic ideas of winners and losers, and the requirement to maintain a hysterical narrative.

    There is an issue where I agree with Rob, however. I think it would best for the CJK people to be creating use-cases and education materials and doing outreach to Westerners first, to get vendors and other parties on side for the need. If they don't, then merely producing extensions and sticking them in a standard will be pretty useless. (By the way, creating use cases and educational material and doing outreach is indeed what the CJK people are doing now, first.) It will be interesting to see whether the ODF and OOXML groups can walk the walk of harmonization and convergence when it comes to new features at least. And whether the vendors, after talking about openness of their standards, can include openness to East Asian requirements, against a record of setting these aside.

  • Rick Jelliffe

    4/17/2010 1:56:46 AM |

    Correction:  I meant  "The thing is that OOXML can be a success even if it doesn't benefit any large vendor."

  • Rob Weir

    4/18/2010 4:06:13 PM |

    Rick, your argument is clearly circular, essentially, "the standardization of OOXML is valuable because now I have a standard".  Is that really the most we can say?

    I think you are confusing the value of a specification with the value of standardization.  I can see how OOXML as an ISO standard was valuable to Microsoft.  It allows them to claim a standardized format in a world that increasingly is asking for standards.  But JTC1 Directives does not ask for NBs to look at whether it was valuable for Microsoft.  It says, "The social and economic long-term benefits of an IS should justify the total cost of preparing, adopting and maintaining the standard."  So again, what was the value of standardizing OOXML over Microsoft merely publish technical disclosures, like they have done for 10,000's of pages of other materials?  If the best you can say is "Standardization is good because you end up with a standard" then I think you make my point for me.

  • Jesper Lund Stocholm

    4/18/2010 5:53:50 PM |

    Hi Rob,

    Thank you so much for proving my point that I am not the biggest idiot.

    Smile

    On thought, though ... if ISO is so irrelevant wrt OOXML (where work is actually taking place) then ISO must be even more irrelevant wrt ODF, where barely any work has taken place since approval in 2006. Can we expect you to drop any plans for ISO submittal of ODF 1.2 in the future then, or are you being as big a hipocrite as always?

    Oh - and about your claim that two national bodies approached you with concerns about MS-stacking (which, I personally think is one of your usual big fat lies) - did you tell them as you did to Alex, that they either speak up at the plenary or shut up and stop "whining about it"? Or file an official complaint to the secretariat, chairman or even JTC1 about it instead of complaining to you while having a smoke outside? It would be much more beneficial for everyone to have it out in the open since, you know, nobody listens to you and your easily digestable retoric anymore.

  • Rob Weir

    4/18/2010 8:56:21 PM |

    Jesper, thanks as always for your deeply considered thoughts on this.  But I think you are conflating two different things:  whether a specification is _suitable_ for standardization in ISO, and whether there is _sufficient value_ relative to the considerable time and effort required to standardize something in ISO.  These are two different things, and one does not imply the other.

    For example, take the W3C's XML standard.  I think it is perfectly _suitable_ for an International Standard, in terms of technical quality, linguistic and cultural adaptability, interoperability, etc.  However, it is already universally adopted, so there is almost zero value in having the W3C make a PAS submission to JTC1.  Once you have 100% adoption, what is the point?  Where is the value in doing that extra work?

    As for OOXML, my primary objections have always been that it was not suitable for standard, being as it is an XML representation of a single vendor's application, with no consideration given to any other vendor's requirements.

    As for the value of standardizing OOXML in JTC1, no one seems to be willing to answer that.  But I think we all know the answer.

    As for ODF, as you know I argued against sending it to JTC1.  I said it was a waste of time and resources and there was little benefit to making this effort. But one SC34 WG Convenor told me that it was pointless to argue against someone else wasting their own time.  So I relinquished my objection.  So long as it does not waste my time, and does not distract the ODF TC, I'm fine with others working on an ODF 1.1 amendment.    I certainly think the standard is _suitable_ for an amendment.  But the value of that effort is lost on me. Several SC34 NBs have asked for us to send ODF 1.1 along.  Maybe you or Alex can explain it to me? Remember, it was never my priority to send ODF 1.1 to ISO.  But SC34 NBs did repeatedly ask for it.

    One perspective that might help you think about this is that there has been, in the last 15 years or so, a rush away from standardizing in ISO and toward doing this work in more agile consortia.  In response to this, ISO/IEC created the PAS process in order to bring more relevant to work to ISO/IEC.  Without a strong working relationship with standards consortia, ISO/IEC will fall into further obscurity.  Two days ago I saw a presentation analysis the ICT standards relevant to Spanish public administrations, with a pie chart showing what organization has standardized them.  Only 10% came from ISO.  The other 90% came from consortia, with the W3C, OASIS and IETF leading the list.

    So I think ISO/IEC needs to be a better job at explaining in general what their value is.  Why should anyone bother with it?  Aside from being a a cloistered shelter for those who are temperamentally incapable of dealing with the rigors of multi-vendor consortia work, what value does it bring? What can it do that could not be done better in the OASIS of the W3C?  The only distinction ISO/IEC has is its quaint one-country/one-vote decision making process.  But to the extent this system is dominated by Microsoft, WG4's decision making process is indistinguishable from Ecma TC45's.  So not much of an advantage there.  And much to its detriment, ISO/IEC lacks a coherent IPR policy (look at the i4i patent and JTC1's impotent lack of action on it), lacks the means enforce contribution controls needed to ensure an unencumbered open standard (in other words, they've learned nothing from Rambus)and has a 20th century business model based on selling their standards.  So as far as ISO/IEC goes I am not a fan.  It is not my preferred venue for standardization and I would not recommend it to anyone unless it was of a fundamental international technical nature, like language codes, country codes, Unicode, etc.

  • Jesper Lund Stocholm

    4/19/2010 3:23:18 AM |

    Hi Rob,

    Jesper, thanks as always for your deeply considered thoughts on this.

    Yeah, well ... you somehow managed to push a trigger I did not know I had. Sorry if I came out too harsh on you.

    Smile

    s for the value of standardizing OOXML in JTC1, no one seems to be willing to answer that.  But I think we all know the answer.

    When it comes to the value of standardizing OOXML, I think we need to look further than Microsoft, because it is - clearly - in Microsoft's value to have OOXML maintained in ISO.

    Take the Danish mirror committee to SC34 as an example. We deal exclusively with document formats and even more exclusively with OOXML. In the 3 years since creation of the committee, we have talked about ODF maybe once or twice. But present at every meeting is IBM along with the "Danish Unix User Group" (ORACLE/Sun participated diligently until this year) - all massive opponents of OOXML. One could argue that they'd just sit quietly in the back and say nothing and wait for anything related to ODF to appear, but they are not. They come fully prepared to each meeting with qualified oponions of the topics at hand and we have some really good technical discussions about OOXML and its maintenance and future.

    So even though your usual line is being broadcast that noone but Microsoft has any value of having OOXML in ISO, the factual situation in Denmark is, that IBM Denmark an other opponents of OOXML seems to see enough value in it to provide the resources to participate in the work.

    One of the reasons for me pushing OOXML into the hands of ISO was to provide a forum where competitors of Microsoft could have some sort of influence on the document format of (primarily) Microsoft Office ... a forum they could use if they wanted to. I have no idea of what the situation is outside Denmark, but as long as competitors of Microsoft like your company IBM actively participates in OOXML maintenance in ISO, I have absolutely no doubt that it provides value to them as well - and that having OOXML in ISO is "the right thing (tm)".

    PS: Was the beer as bad as I feared?

  • Rob Weir

    4/19/2010 4:46:55 AM |

    Jesper, by that logic, in any country where IBM is a member of the JTC1 mirror committee you could say that we support every standard that ever is approved by ISO/IEC.  But that would be an incorrect statement, right?  We obviously vote against some standards, and vote in favor of others.  But that does not mean we walk out of the room when OOXML is discussed.  Don't mistake protocol and civility for approval.  In any case, I can safely say that IBM has received no value from OOXML as a standard.

    But you still have not stated what the value of standardizing OOXML is in ISO, beyond the benefit you would have from a technical disclosure that Microsoft puts on their web site.  For example, the main "influence" SC34 had on OOXML was in the creation of the Strict conformance clause, which Office will not write out until 2014 or so.  Compare that to OASIS ODF 1.1, which is not an ISO standard, but which Microsoft has implemented since 2009.  Compare those two dates carefully and please tell me again what the value of ISO is?  

    Look at Microsoft's "Proposed Undertaking" to the EC, where they describe their interoperability commitments.  Which format do they promise to support for the next 10 years?  Hint:  It isn't OOXML.  And for their file format choice ballot screen, which standardized formats will they allow?  Hint:  OOXML Strict isn't on the list.  So tell me again, which standard has had more influence in determining what Microsoft Office writes out? And what standards organization has more influence in determining what Microsoft Office writes out?

  • Wolfgang

    4/19/2010 9:10:44 AM |

    @Rob Weir
    "The interesting thing, I think, is how the situation is very different in the browser world, where it appears the Microsoft is taking standards seriously now."

    I find this very easy to explain:

    In contrast to the office desktop market, MS has long lost the leadership (technologically and spread-wise) concerning the web and its standards (ironically mainly due to their abysmal support of standards in the first place, think IE6) and their share and influence in the browser market simply isn't big enough any more for any funny business regarding HTML5 - they are simply *forced* to behave.

    I think if MS's market share still was over 90% in the browser market (as it still is in the office desktop area) they wouldn't give a toss about HTML-standards...

  • Rick Jelliffe

    4/19/2010 10:00:45 AM |

    Rob: What is the point of answering you if you won't read the answer?

    Perhaps in bullet form would be easier:

    1)Documentation: "much better documentation (2000 unedited pages to 8000 reviewed pages! come on: is that really noting?)"

    2)Syntax improvements: "broader participation and broader review." Supposedly Word reads the BRM Strict: for industrial publishing where OOXML is mainly a terminal format, this is a significant chunk of our use case. (We want write as well, because it opens up new use cases, or at least does not block them as much.)

    3)Rich range of standards for different uses: "The value to my employer is that we now have three standard languages in the industrial publishing output space that we can (and do) use: XSL-FO, OOXML and ODF."

    4) Feature complete: "because of standard OOXML there is no part of Word that is blocked off from us, coming or going."

    5)Progressive improvement:"And in the medium term to have Strict *if* Microsoft implements Strict for write as well as read." But certainly not fast enough, hence this blog.

    6) Get vendors at table to engage them for issues important for NBs, eg CJK typesetting:  "And whether the vendors, after talking about openness of their standards, can include openness to East Asian requirements, against a record of setting these aside."

    7) Have a forum for discussing harmonization and convergence of ODF/OOXML: "t will be interesting to see whether the ODF and OOXML groups can walk the walk of harmonization and convergence when it comes to new features at least. And whether the vendors, after talking about openness of their standards, can include openness to East Asian requirements, against a record of setting these aside."

    8) But not forgetting the main point, which is that judging a standard's success after 2 years is ridiculous. The standards timeline is 5 years, 10 years, 15 years. Standards cross-pollenate and evolve.

    So all these are substantive points, not related to "standardization for standardization's sake". And there are others that were not in my response, too. But of course, you are just trolling.

    All these points are served *better* by having an international standard for OOXML than they would be by having a consortium standard, a technical note or just some pages on the website. (That the original drafts at ECMA were just those web pages, and ECMA had to expand then from about 2000 to 6000 pages shows how inadequate they were.)

    How much better? Time will tell. I certainly agree that the benefit is not so absolute or necessary that a reasonable person could not reasonably decide that it would be better not to have the International Standard the OOXML. (Unfortunately, reasonable people with that view don't seem to get much air time.)  And I think Transitional has now delivered everything needed of it as far as documentation etc and is now counter-productive.

    But the benefit does not need to absolute, necessary and immediate to justify the documentation for some technology becoming an international standard: it just needs to better with it that without it.

  • Rob Weir

    4/19/2010 12:50:24 PM |

    Rick, let's take your items:

    3,4 and 7 are tautological, just saying "standards are good because standards are good"

    4 is not true as of Office 2010 where they include dozens of extensions which although syntactically valid OOXML are not documented by the ISO standard.  They are documented by implementation notes on Microsoft's web site, which again proves my point that ISO/IEC is not really adding value here.

    2,5,6 and 8 have not occurred yet.  And your 8 contradicts your 4.  If the main point of OOXML was to be Fast Tracked as a full-description of Microsoft Office's formats, then it is inconsistent to now treat it is if it were an emerging standard whose value might be seen only 8 years from now.  And if it was intended to be a forward-thinking standard to advance the state of the art in this area, then the hasty preparation in Ecma, the Microsoft-dominated maintenance in SC34/WG4, the lack of implementation of any of the new features and the fact that Microsoft is now extending the standard outside of the ISO/IEC process, these factors combined give a very poor foundation for forward-thinking work.  8000 pages of mainly legacy crap is not really a good starting point for creating the future. In any case, to suggest that the trajectory of OOXML would be similar to the trajectory of a Schematron is rather odd.  You should know better.

    That leaves #1, that you got additional pages.  But how many thousands of new pages was just pasting a redundant copy of the schemas into the text of the specification?  Was that really a substantial value-add by SC34?  Maybe next time paste in two redundant copies of the schema for double the benefit?  Yes, there was the Excel spreadsheet formula documentation, but that happened in Ecma, not in SC34.  And Microsoft already had that specification available, so when it was requested they just handed it over.  It wasn't like they sat down and wrote 2,000 pages of new documentation for anyone, or that SC34 added any value to it.  In fact the BRM approved it dispite the fact that almost every financial and statistical function was defined incorrectly.  I think that makes it have negative value.  Finally, note that Microsoft would have given the same specification to the EC as well, if asked for it, and published it as implementation notes on their web site.  So I would not claim this as an benefit of the ISO/IEC process.

  • Jesper Lund Stocholm

    4/19/2010 6:22:13 PM |

    Hi Rob,

    by that logic, in any country where IBM is a member of the JTC1 mirror committee you could say that we support every standard that ever is approved by ISO/IEC.  But that would be an incorrect statement, right?

    Indeed it would - that is why I am not saying this. I am saying that IBM is a very active member of the "OOXML committee" in Denmark. Who knows if you are a member of other JTC1 mirror committees as well. I am not in any way saying that since you are a member of e.g. a mirror committee to SC22 then you endorse OOXML.

    But IBM is indeed very active in the OOXML maintenance in Denmark. First of all you paid the annual fee of €3000/yr to participate in the OOXML committee. Secondly you do your homework for each meeting - which we have around 8 or so annually. IBM is not just another of the sheep we have historically met at meetings that were only there to raise their hand and say "three letters good, five letters bad". And thank you for this! Thirdly, the first (I think) defect report from Denmark was actually initiated by IBM (removal of the relyOnVml attribute) and IBM fought vigirously to have it submitted.

    So no matter how you try to spin it, a value of having OOXML in ISO (as opposed to having it solely in the dungeons of Redmond) is that competitors like IBM gets a chance to participate in the maintenance of it.

    But who knows - maybe it is just a scam or a scherade from IBM the work you contribute or maybe it is just a chance for you to mess with MSFT - but that is not important. Either way IBM seems to get enough value from it to participate ... regardless of reason.

  • Jesper Lund Stocholm

    4/19/2010 7:36:14 PM |

    Hi Rob,

    (just thought of this)

    Compare that to OASIS ODF 1.1, which is not an ISO standard, but which Microsoft has implemented since 2009.  Compare those two dates carefully and please tell me again what the value of ISO is?


    Wait a minute - I seem to remember a series of articles you wrote consistantly hitting on Microsoft's ODF implementation for not being good enough. I even seem to remember you saying that "it's not hard".

    But now you are using Microsoft's (to paraphrase on your words bck then) less-than mediocre ODF-implementation to justify that the OASIS process is better than the ISO-process?

    Are you now arguing that Microsoft's ODF-implementation is satisfactory? That would seem to contradict your articles that clearly demonstrated that all the OASIS-process had provided for was a bad implementation of ODF.

    Which one do you chose?

  • Rick Jelliffe

    4/19/2010 10:37:52 PM |

    Rob Wier wrote:  "3,4 and 7 are tautological". No they are real positives as far as our business building publishing systems is concerned.

    "4 is not true as of Office 2010 where they include dozens of extensions which although syntactically valid OOXML are not documented by the ISO standard."  Yes, ISO standardization provide little benefits if you make a requirement for clairvoyancy. This is a colossal backflip: is Rob actually saying that OOXML can be faulted because it didn't have information relating to future software not developed at the time of the standard? I though it was important to have reference implementations, fidelity, etc. Oh well.    

    (And, on this issue of extensions, it is notable that OOXML's MCE provides a mechanism for keeping extensions out of the main standard, and to provide mechanisms for graceful degradation or lesser functionality.)

    "2,5,6 and 8 have not occurred yet." But Rob's original question was "What is the value" not "What has the value been so far." Pretending that you asked a different question is the sign of a troll, someone intent on making mischief and not participating in any dialog seriously.

    On contradiction, I leave it to readers to figure out Rob's mysterious point. I would have thought the concept that the future often confounds our expectations was common knowledge.

  • Rob Weir

    4/24/2010 2:42:35 AM |

    @Jesper, Yes, I have been critical of Microsoft's ODF support in Office 2007 SP2. But I do know that they have promised to support ODF 1.2 when it has been standardized, and at that time my main complaint (spreadsheet formula compatibility) will be addressed, if not sooner.  So I believe that ODF 1.2 will be supported (read and write) in MS Office before we ever see MS Office support OOXML Strict (read and write).

    To your other point, I think you confuse attendance with participation, and participation with support.  The list of standards IBM expresses their support for is listed here:  http://www-03.ibm.com/linux/ossstds/isplist.html  You do not see OOXML on that list.  If reporting defects on OOXML was an expression of support, then you would need to count me as OOXML's greatest supporter.  But that would clearly be absurd.

    General rule of thumb.  When a company supports a standard, they typically are not coy about it, secretly sending one person to attend a meeting in some far away corner of the world, to labor in secret.  If they support a standard, they issue press releases and put the info up on their web site.  As an example, look what Microsoft did when they initially added ODF support to Office.

    @Rick,

    Let me see if I have this right.  Microsoft is excused for not supporting OOXML Strict because of the long lead time in creating Office 2010.  When Office 2010 was initially being planned, they had no idea of what would be in the approved ISO/IEC 29500, so there is no way they could have planned for it.  But at the same time, Microsoft is excused for not submitting their Office 2010 enhancements to ISO, because...? Hmmm.. wait a minute.  I would have thought that this "long lead time in creating Office 2010" would have given them ample time to make this material available, via Ecma, or directly via an NB of their choice.  But they didn't do that, did they?  No.  In fact, at the Seattle SC34 Plenary, they had the opportunity to do that, but instead held it back and discussed it only at their private interoperability event the day after the SC34 meeting ended.  They've since posted the material as implementation notes on their web site.  So I don't think there is a lack of time issue.  It is more a lack of will, and as I said before, a lack of perceived value.

    Also, you put too much faith in MCE.  Not all OOXML extensions in Office 2010 are expressed in MCE. There are places where they just add undocumented new stuff to attribute values, where this is not documented in ISO/IEC 29500, and no MCE fall-back is given.  More on this in a future blog post.

  • Jesper Lund Stocholm

    4/26/2010 4:46:29 PM |

    Hi Rob,

    The most fun thing about debating with you is that you never know in which direction you are gonna go when you start losing the argument.

    @Jesper, Yes, I have been critical of Microsoft's ODF support in Office 2007 SP2. But I do know that they have promised to support ODF 1.2 when it has been standardized, and at that time my main complaint (spreadsheet formula compatibility) will be addressed, if not sooner.  So I believe that ODF 1.2 will be supported (read and write) in MS Office before we ever see MS Office support OOXML Strict (read and write).

    Who's talking about ODF 1.2? You specifically mentioned Microsoft Office 2007 SP2's implementation ODF 1.1 and not ODF 1.2 .

    You know - the implementation where you in your test of spreadsheet interop said:

    (...)But the degree of incompetence needed to explain SP2’s poor ODF support boggles the mind and leads me to further uncharitable thoughts.(...)

    But now you are using this specific implementation to salute the "OASIS-process"? That's rich, Rob!

    Also - that you are trying to somehow pay tribute to OASIS process by using Microsoft's future implementation of ODF 1.2 is laughable. At the very plugfest you attended recently, Microsoft mentioned that it would support ODF 1.2 nine months after ISO-approval.

    Please tell me again how this shows that the OASIS-process is better than the ISO-process?

    And finally about IBM's participation in OOXML maintenance:

    To your other point, I think you confuse attendance with participation, and participation with support.

    Again - who's talking about IBM supporting OOXML? I said IBM was actively participating (not only attending) in OOXML maintenance in Denmark. Who knows what you are doing in other countries. The public records of our meetings in Denmark are clearly evidence to the participation of IBM in our work ... not to mention the enormous amount of hours IBM poured into the national body processes during the DIS-processing of 29500.

    Please stay on the subject, Rob ... otherwise it's a waste of time talking to you.

  • Jesper Lund Stocholm

    4/26/2010 10:33:04 PM |

    Hi Rob,

    Also, you put too much faith in MCE.  Not all OOXML extensions in Office 2010 are expressed in MCE. There are places where they just add undocumented new stuff to attribute values, where this is not documented in ISO/IEC 29500, and no MCE fall-back is given.  More on this in a future blog post.

    Yup - "Graceful degredation" is only possible when using ACB's (Alternate Content Blocks). When using MCE ignorable namespaces and extension lists (pre-defined extension points in the document format) there is no (obvious) way to provide fall-back scenarios and functionality.

    ... looking forward to your no doubt very, very interesting post on this.

  • Rob Weir

    4/26/2010 11:36:57 PM |

    Jesper, I'm afraid that you've lost track of your own argument.  Let me point you back at it.  I asserted that SC34 has added no value to OOXML.  All they seem to have done is created a frankestein standard cobbled together at the BRM that no one, not even Microsoft has implemented.  I asked a reasonable question:  what is the value of ISO standardization for something like OOXML -- a single vendor's technology. You're best (or at least first  -- maybe you're holding a better argument in reserve) response was that IBM participates in some obscure committee in Denmark, therefor we must be receiving value from ISO standardization of OOXML.  But I pointed out that participation is not the same as support, and certainly not the same as receiving value.  And I think I lost you at that point.  So let me restate.  A fireman shows up at a house fire with hose and ax and ladder.  He is very busy.  Does that mean that he gets value from the fire?  I think that would be an odd view to take.  The fireman is very opposed to the fire.  He merely tries to contain damage from the fire and eliminate it if he can.   IBM is a member of Ecma.  Does that mean we support Ecma-376?  No.  We voted against it.  IBM employees submitted many technical defects on DIS 29500.  Does that mean we support OOXML and get value from its standardization?  No, we urged voting against it.  It is important to distinguish the firefighter from the pyromaniac.  Both are present at fires.  But one is trying to put the fire out.

    As for Microsoft's support of ODF 1.1, my point is that ODF 1.1, even without ISO approval, is far more implemented than ODF 1.0 was.  So if there is any liability in ODF 1.1 not being approved by ISO, then it seems to have passed unnoticed.  ODF 1.1 is far more broadly implemented than OOXML Strict,  SC34's Frankenstandard.  So I don't see the value of ISO standardization here either.  Maybe you can explain it to me.  What unique value has SC34 or ISO in general brought to ODF?

  • Rick Jelliffe

    4/29/2010 10:10:19 AM |

    Rob wrote "Let me see if I have this right.  Microsoft is excused for not supporting OOXML Strict because of the long lead time in creating Office 2010. "

    No.  It is obviously not what I said, as Alex' quote shows. I think that was Jesper's point.

    (Rob-appreciators will enjoy his touch here of putting his own phrase "long lead time" in double quotes as if it had been said by someone else, presumably me.  Hmmm, I thought IBM employees were supposed to be super careful about invented quotes... )

    P.S. It will be interesting to see if the EU's investigation of the TurboHercules complaint (about IBM's monopoly problem)  will be resolved in part by requiring IBM to put their protocols etc into an International Standard, a la OOXML  Smile  

Comments are closed