Over halfway there… including some positive changes to the Open XML standard


I had mentioned that the members of Ecma TC45 all met last week in Kyoto to continue working through the national body comments we received as a result of the ISO ballot for Open XML. We’ve been making a lot of progress over the phone the past couple months (we have 3 hour calls twice a week), but it’s always productive to meet face to face and really get through some of the more difficult pieces.

Today Ecma has updated the proposed dispositions’ portal for the comments that national bodies can go and reference. We’re a little over 50% of the way through the 3500 comments received. If you read the status report released by TC45 you’ll see that there have been some pretty significant changes proposed so far, and you’ll see more are on the way. Here’s a link to the TC45 report: http://www.ecma-international.org/news/TC45_current_work/New%20set%20of%20proposed%20dispositions%20posted.htm

Here’s a list of some of the changes that were proposed in this batch:

1 Allowing for ISO-8601 Dates

ECMA-376, the original Open XML standard adopted by Ecma, assigned a unique numeric value to each date in a spreadsheet, in order to improve the speed of date calculations. Based on the comments received from some National Bodies on this issue, DIS 29500 will be updated to allow date values to be stored using the format defined by the ISO 8601 standard.

2 Internationalized handling of weekdays and weekends

ECMA-376 allowed for a week that begins on Sunday or Monday, but not a week that begins on any other day, such as Saturday. Ecma is proposing a comprehensive range of options for what is defined as the first day of the week, and what is defined as the weekend.

3 Language tags

ECMA-376 used a set of integer values to identify the language applied to regions of a document. Ecma is proposing that the language tags specified in the DIS should instead leverage an internationally recognized practice for representing languages, IETF BCP 47. IETF BCP 47 is a Best Current Practices document that incorporates use of the ISO 639 standard for languages, ISO 15924 for scripts, and ISO 3166 for regions. This proposal directly follows recommendations from National Bodies in several countries.

4 Page Borders

ECMA-376 included support for a variety of graphical elements that could be used as page borders. Several National Bodies noted that this closed list of graphical elements was not sufficiently diverse and global in its contents. Based on that feedback, Ecma is proposing to change the Open XML standard to allow for custom page borders. This will enable implementers to determine the best option for including borders relevant to their applications.

5 Usage of ISO standards for grammars

ECMA-376 used its own notation for defining the grammar for some of the more advanced functionality, such as spreadsheet formulas and word processing fields. Several National Bodies noted that the existing grammars in ECMA-376 are non-standard and were not fully described within the DIS. In response to this concern, Ecma proposes to revise the notation for spreadsheet formulas and fields to use an existing ISO standard. Formula notation will now use ISO/IEC 14977:1996 – Syntactic metalanguage – Extended BNF. This proposal improves the ability for implementers to test and validate conformance to the specification.

I was surprised to hear from a number of people that they were skeptical TC45 would accept any changes to Open XML. TC45 is excited to work closely with the national bodies and investigate what the best solution would be to these issues (including of course changing the specification). While its tough news in certain ways for me as an Office developer; it’s great news for me as a TC45 member. We had made some design decisions within TC45 that many of the national bodies disagreed with, so we took that feedback into account and came up with proposals we think will address the national bodies concerns.

Jan van den Beld, former head of Ecma international, had a post this morning talking about his thoughts on the progress we’ve made so far: http://janvandenbeld.blogspot.com/2007/12/number-of-proposed-dispositions-for-nbs.html. Jan is impressed by the progress made, and also shares his thought on how many comments per page you would typically expect to get.

We still have a ways to go, and those of you following along know there are a number of other contentious issues we haven’t finalized on yet (but we’re getting very close). I’ll provide more detailed explanations of many of the responses over the coming weeks, and I’m sure you guys will find those interesting.

-Brian

Comments (64)

  1. hAl says:

    Allthough it has not really hampered spreadheet uptake in the last 35 years another comment duplicated by several national bodies was that the decimal dates used within the spreadsheets were limited to 1-1-1900 in spreadsheet cells.

    Is the added ISO date 8601 functionality for spreadsheet cells still limited to that date ?

    Is the added ISO 8601 date support of the same subset used by the XML standard?

  2. jones206@hotmail.com says:

    We’re seperately working through the issues of date limits (ie what’s the range of dates); as well as the leap year bug.

    I think we have good solutions for both of those as well within TC45, but we are still in dicussions on the final design.

    -Brian

  3. hAl says:

    What happened with the existing list of page borders.

    Did you scrap it from the spec, make it an annex or is these new custom borders a feature of extending on them ?

  4. luke says:

    Does this mean all those OOXML documents I have made so far will be incompatible with the eventual ISO standard?

  5. hAl says:

    [quote]Does this mean all those OOXML documents I have made so far will be incompatible with the eventual ISO standard?[/quote]

    Office Open XML should supports versioning.

    So you are using Office Open XML 1.0.

    The currently being standardized version might lead to a Office Open XML 1.1 version

    Generally a new version is fully compatible and it should be easy to convert to the new version if you want. Even if some features are dropped they still could be supported by extension or embedding mechanism so your data would not be lost.

  6. Keep up the good work.  Clearly, this is a huge job, but it is encouraging to know you have handled so many comments.  I’m looking forward to seeing the revisions that get hammered out if this makes it through the BRM.  As you know, I am one who has had serious reservations about the spec, but this level of focus and effort is good to see.

  7. Alex says:

    How will Office be updated to support any changes in the file format, and what will happen with regards interoperability with old files and the compatibility packs for older Office?

  8. omz says:

    >I think we have good solutions for both of those

    >as well within TC45, but we are still in

    >dicussions on the final design.

    mmm, nice way to build "standards" … discusing and rushing it in a couple of months

    keep up the good work Microsoft and ECMA

    Thanks god you don’t make food and chemical "standards".

  9. Hi Brian,

    first of all congratulations on your progress. That’s something I wanted to hear about for a long, long time and I’m glad that Microsoft and Ecma are finally starting to fix serious issues in the spec. Yes, sometimes it’s painful, but that’s part of the job.

    I do agree with omz that it’s a bit sad that this is happening so late in the game, and as I’ve said before if OOXML had not been fast-tracked a lot of the current conflicts could have been avoided. Still, better late than never, right? Hope to read more from you soon, keep it up!

  10. Karellen says:

    "DIS 29500 will be updated to allow date values to be stored using the format defined by the ISO 8601 standard."

    I take it from the language used that ISO 8601 dates are not required, and that the old integer format can still be used. Which means that all applications that implement OOXML still have to implement 20-year old bugs.

    This does not ease creating independent implementations in the slightest, and in fact complicates it now that there are two different ways to do the same thing, one of which is still broken in subtle and surprising ways.

    Do you guys miss the point on purpose, or are you really this stupid? What’s that saying … sufficiently advanced incompetence is indistinguishable from malice.

    Fail.

  11. David Lane says:

    Karellen, may I congratulate you on the best corollary of a Clarkism (or any-ism) I’ve seen in quite some time.  Malice indeed.

    Thanks for that.

    Dave

    P.S. Brian – it’s "separately" not "seperately".  Trap for young players.  Just remember "a rat".

    Oh yes, it just occurred to me that perhaps your proprietary browser doesn’t indicate spelling errors like my open source one does.

  12. hAl says:

    Of course we would wish now that ODF had the same kind of scrutiny.

    It might have explained what ODF means by a glyph or by an ideograph-alpha value because the specification seems to completely lack info on what those mean (making such items completly impossible to implement purely based on the ODF spec)

  13. hAl says:

    @karellen

    All current regular spreadsheet implementations already support decimal date values and have been doing so for more than twenty years. Supporting decimal values is much easier for spreadsheets than supporting ISO dates.

    It is actually the addition of ISO dates that might make more work for independant spreadsheet implementations support as not all of them already can handle the ISO format subset. (of course not a single spreadsheet implementation is able to handle the full ISO 8601 standard implementation because of it’s complexity)

  14. Mike Brown says:

    Some common sense on date formats, at last!  Congratulations; you are now where you should have been a year ago (and that’s being generous to you).

    These kind of changes are so *not* what ISO Fast Track is all about.  This should have been sorted way before you ever got to Fast Track submission, not two months before the final BRM.

    But then, it doesn’t sound like it’s even sorted now, does it?  We have some kind of agreement to implement some kind of ISO 8601 date compatibility, but the how is not yet specified.  Will it be the default?  Will it be a requirement, or as Karellan asks, will implementers still be free to use the old, buggy integer format?

    You really gonna have all this ready for February, with reference implementations to demonstrate?  Methinks not.

    In truth, you’re going to be asking the delegates to vote on a load of Microsoft promises, are you not?  And we all know what they’ve been worth down the years.

    Cheers,

    – Mike "man of the people"!

  15. Reggie says:

    Mike Brown,

    ODF 1.0 wasn’t ready for prime time in the least, yet ISO rubberstamped it based on promises that it would improve in future versions.  Promises that are still unfulfilled, btw (ODF 1.1 was finalized by OASIS almost a year ago, yet hasn’t been submitted to ISO yet (i.e. ISO ODF is still utter trash)).  But I didn’t see you or your kind raising any issues regarding it.

    BTW, why do you anti-OOXML people even post here?  You hate OOXML, we get it.  Now leave.

  16. David Lane says:

    Hi Reggie,

    I think you’re referring to those of us who have reason to believe that Microsoft is a disingenuous corporation which regularly screws its customers by abusing its monopoly status to maximise revenue, and feel the need to expose the injustices perpetrated by Microsoft and its general lack of ethics in a public forum.  If it’s owned by Microsoft, so much the better.

    We also find it useful to enumerate those injustices and highlight breaches of ethics and anti-community spirit Microsoft and its sycophants demonstrate in places where Google happily indexes it for other interested parties to find in future.  

    Cheers,

    Dave

  17. David Lane says:

    Now, I’d greatly appreciate it if someone at Microsoft could respond to my oft-asked but never answered question: why not simply use OOXML (without pissing everyone off by abusing ISO’s systems) as the default file format for MS Office 2007 while simultaneously supporting ODF as an integrated (not "plug-in") file format that government and other open standards-aware individuals and organisations could set to "load and save by default"?

    Those users would realise, of course, that not all of the vast richness of MS Office 2007’s innovative capabilities can be properly represented in ODF… but that’d be OK, because pretty much nobody would care about that stuff anyway.

    Seems like a win-win for Microsoft.

    1. no need to risk the PR disaster of another failed attempt at ISO standardisation (and all the promises that would require) or negative publicity of getting caught rigging the vote like last time.

    2. everyone who cared could exchange documents via a single, open ISO standard format for Office documents that could be read and written by multiple independent software implementations.

    3. anybody who cared about the difference between what could be expressed in ODF (on whose further development Microsoft would be most welcome contribute!) and OOXML could save in that format.  And they could collaborate with anyone who happens to have MS Office 2007 (but nothing else, of course, as would be the case regardless).

    The only rationale I can see for Microsoft refusing this avenue is the fact that it intends to wield OOXML like Rob Weir and Groklaw suggest: as Microsoft’s own pet ISO standard that they can alter and tweak as they choose (making other non-Microsoft implementations "slightly less compatible") and using it to perpetuate lock-in and force MS Office upgrades while simultaneously being able to tout their software as being "open standards compliant"…  

    Cheers,

    Dave

  18. Mike Brown says:

    @Reggie,

    >> ODF 1.0 wasn’t ready for prime time in the least, yet ISO rubberstamped

    >> it based on promises that it would improve in future

    Says you.

    But that’s by the by now, isn’t it?  ODF *did* pass ISO standardisation, and without its supporters having to pack the vote with their stooges either.  The question now is whether OOXML should do the same.  I don’t believe that it should.

    >> Now leave.

    Not a chance, matey!

    Cheers,

    – Mike

  19. Doug Mahugh says:

    As Ecma announced today, the project editor (Rex Jaeschke) now has proposed dispositions for over 50%

  20. Karellen,

    As hAl said earlier, it is more the rule than the exception that spreadsheets support dates in decimal values – so ODF is really the odd one in the class. Also, we (programmars etc) are actually quite used to deal with dates in either ISO-form or as numeric values. Most of the programming languages I know support either method, and it really is not that hard. It may look really, really hard – but it’s really not.

    It seems that the criteria for acceptance of OOXML is for it to be so simple, that you could take the most stupid programmar at IBM, blindfold him, stuff a penguin up his ass, tie one arm around his back, tickle him with a feather … and he would still be able to make a error-free implementation of OOXML in 21 days.

    Well, if that is the notion … I must say I beg to differ.

    :o)

    /Jesper-san

  21. As Ecma announced today, the project editor (Rex Jaeschke) now has proposed dispositions for over 50

  22. David,

    Let me give you a possible reason for Microsoft wanting ISO-approval for OOXML:

    OOXML and ODF were developed more or less in parallel and has been that for more than 10 years. Until ODF became an ISO-standard, no-one really cared about it – and I really don’t think anyone at Microsoft thought much of it (as with the Internet in the mid-ninetees). However, when ODF became ISO-approved, the ODF-supporters started to use ISO-approval itself as a "badge-of-honor" and started arguing to various governments that it was essential for a standard to be ISO-approved. To be honest, I don’t think Microsoft saw this one coming.

    You are correct that ISO-denial does not mean that OOXML will go away – but the OOXML vs. ODF battle is not about the formats; it’s about market control. The anti-OOXML-crowd has been really, really succesful in persuading politicians that the ISO-badge is what’s it all about and Microsoft cannot afford not to be a choice. On top if this, the European Commission asked Microsoft to submit their file formats for standardization which I believe is a good thing.

    You say that it is not really that important that something dissapears when converting from OOXML to ODF, but this might apply to you. In the government sector, where documents might have a legal status, this is certainly not the case. Here it is a legal demand (at least in some countries), that all information in all documents is preserved, and that’s why ODF is not good enough. It might be good enough for the desktop user – but there are other users of office applications out there, whose requirements differ.

    :o)

    /Jesper

  23. Mike,

    @Reggie,

    >>> ODF 1.0 wasn’t ready for prime time in the least, yet ISO rubberstamped

    >>> it based on promises that it would improve in future

    >

    >Says you.

    Well, I do too. There are huge holes in the specification where the most prominent one is missing specification for spreadsheet formulas. Other examples could be mis-use of existing standards where it has been either extended, modified or delimited.

    How can you with any reason say that you have a format for spreadsheets – and then exclude specification for formulas?

    > But that’s by the by now, isn’t it?  ODF *did* pass ISO standardisation

    Yes – but what I and others are simply politely asking is to have the same standards applied to approval of OOXML as was the case with ODF in a sense as "If it was ok with ODF, surely it must be ok for OOXML as well". Sadly, to me this doesn’t seem to be the case. ODF flew through ISO because no-one cared about it except those supporting it. I attended the ISO/JTC1/SC34-meeting in Kyoto, and I was really surprised to see how easy it was to be able to put something through if noone cares.

    If only ODF had been through the same scrutiny as OOXML, maybe the list of 100 defects published by a single man Dr. MURATA Makoto (isn’t he now convenor of SC34/WG1?) at http://www.jtc1sc34.org/repository/0942.htm would have been fixed before acceptance … along with the comments from the final vote, which have all been ignored. Some of the defects are trivial and spelling errors, but some are actually quite serious since they allow interpretation of how to implement ODF and even induce errors.

    :o)

    /Jesper

  24. Gavin Beckett says:

    @Jesper

    I think there’s another possible interpretation of the increased scrutiny of OOXML in comparison to ODF.

    People voted Yes with comments for ODF, and then realised that this wasn’t a strong enough option to get the comments addressed. They have learned their lesson, and now choose to vote No with comments, in the hope that this ensures OOXML gets fixed before approval as a standard.

    The "people" I refer to here are neither ODF nor OOXML supporters/opposers, they’re just professionals who want to see standards that are fit for purpose.

    You are right that ODF 1.0 has flaws. ODF 1.1 fixes lots of them, and ODF 1.2 fixes even more. OASIS need to submit back to ISO of course, as these versions are not ISO approved. This is not a broken process though, it’s working. ODF 1.1 was implemented into OpenOffice over time, and I have every reason to think that 1.2 will be as well. The comments you reference were posted this December – there’s no reason to think they won’t be addressed is there?

    Anyway, my key point is that your reasoning is "if it was ok with ODF, surely it must be ok for OOXML…" but that another reasonable point of view is that it wasn’t right for ODF, and it isn’t right for OOXML. We can see that ODF is being improved appropriately, so that mitigates the "mistake" but doesn’t make it a good idea to repeat it.

    BTW, I’m a customer, not a developer or a vendor, although I can appreciate those points of view. When I speak of these things, it’s with the point of view of someone who is trying to actively manage their suppliers to deliver what my organisation needs. I give Sun, IBM, MS, Oracle, Novell, whoever, a hard time when necessary, and thank them when appropriate too. We work in a mult-vendor environment, so interoperable standards are hugely important.

  25. Gavin,

    "People voted Yes with comments for ODF, and then realised that this wasn’t a strong enough option to get the comments addressed."

    Yes – they were basically screwed at the (non-existing) BRM, and even though I want OOXML to pass ISO, I fully agree that e.g. Denmark voted the way it did (No, with comments) for OOXML.

    About broken process:

    I do not think the process regarding ODF is/was broken – I think the process is basically what you’d expect for a real-world standardization process. Standards are proposed and are approved. Most of them have flaws in them and they are dealt with in subsequent revisions. This how it works with ODF and this is how it should work with OOXML. Even the directives for ISO/IECD/JTC1 clearly state, that standards proposed through Fast-track and possibly also PAS, are not required to adhere to the normal ISO-way of writing things. It is perfectly ok to "normalize" the proposed standard in later revisions.

    If you haven’t noticed, ISO moves really, really slow and things do not change overnight – and nor do criterias for approval of a standard.

    You say that is was a mistake to approve ODF, but I do not necessarily agree – at least not from an ISO-perspective. Both PAS and Fast-track procedures are made to enable the work of other organizations to be ISO-approved in a jif – and ISO knows very well that mistakes sometimes go through. This is in fact written in the JTC1-directives and this is why they have terms like amendments, revisions and "maintenance" itself.

    /Jesper

    :o)

  26. Mike,

    Another thing:

    You say:

    "But that’s by the by now, isn’t it?  ODF *did* pass ISO standardisation"

    Ok – let’s make a deal, then. We all accept the state of the world we live in, and I’ll stop talking about how ODF was rushed through ISO, how vendor-controlled it is by Sun and IBM and how flawed it was at the time of submission. I will then expect you (and others) to stop whining about why Microsoft didn’t use ODF and why Microsoft chose to submit OOXML to ISO.

    We are where we are – and by leaving these things out of the discussion, we could possibly spend our time discussing things that matter … like technical merit of the disposition of comments.

    Deal?

  27. Francis says:

    Wow–what great news. I am really impressed the sample changes above. They go to show that ECMA and its partners genuinely do care about the process and the standard that emerges.

    Incidentally, please make sure that when you allow for ISO dates, that traditional-style dates are extended backwards to include pre-1900 dates. This would simplify historical research and reconstruction (e.g., with old stock market values, geological records, etc.)

    It’s also highly gratifying to see formulas and fields move to a standard notation. I hope part of this change will also be a provision for the use of portable and relative URIs (instead of absolute MS-DOS paths) in Word fields.

    Keep up the good work!

  28. Mike Brown says:

    @Jesper

    You can talk about what you want, mate.  It’s a free world.  If you want to talk about "how ODF was rushed" then you go right ahead and do so, if you think that was the case.

    I meant "that’s by the by" in the sense that it’s OOXML that’s up for standardisation now, not ODF; nothing more, nothing less.

    But I’ll humour you.  Let’s say, hypothetically, that you’re correct, and that ODF was "rushed" and "flawed".  Does that then mean that another rushed and flawed spec should be get a guaranteed pass, just to balance the books?  Two wrongs make a right is it?  Is *that* your argument?

    If you want to talk about a "the state of the world we live in", why don’t we talk about the state of an ISO committee stacked full of Microsoft’s stooges, eh?  Let’s talk about the standards that can’t get approved now because of all the new "P" members that haven’t voted for anything apart from OOXML since they joined, despite all the cajoling that ISO officials have done.

  29. Mike,

    My point, which you are obviously missing, is that I do not think the process was flawed and I do not think that we should allow OOXML to pass just to "balance the score". My point is that ODF was submitted through a process in ISO and so is OOXML. Talking about what is flawed and not does not make any sense, since ISO has already accepted that OOXML could be fast-tracked – just as ISO accepted to allow ODF through PAS.

    Stooges: Could you please explain what you want to gain from starting to throw dirt? Should we also open the subject of the Kenyan response to ISO in spring 2007, that was clearly written by an IBM employee? OR the Malaysian one? Should we also open the subject on the intense lobbying of IBM in the NBs – resulting in comments directly copied from the IBM-list (with preservation of grammatical errors?) Should we also open the subject of IBM not giving a crap about OOXML until submission to ISO … and then starting their DOS-attack on the process?

    If you ask me – each corporate side of this has behaved "slightly non-gentleman-like" in this process and none of them are pure enough to cast the first stone.

    About SC34 not functioning: Please do a cross-reference with the list of NBs voting yes to ODF and the list of NBs not voting in SC34. Also – it is not fair to pin down this on the coorporations taking part in the ISO-process. The members of ISO are nations and not coorporations. It is their responsibility to simply vote "abstain" if the subject is not clear to them. It is actually pretty easy.

  30. Mike Brown says:

    @Jesper

    >>ISO has already accepted that OOXML could be fast-tracked

    It did so because ECMA has a special relationship with ISO that allows direct access to the Fast Track procedure for an standards that ECMA itself has "passed".  I don’t believe that ISO does any pre-vetting in this case; it assumes that ECMA has already done that before Fast Track submission.  It will be interesting to see if this relationship survives the fall out of this whole sorry business.  And fall out there *will* be, a-plenty.  

    >> Stooges: Could you please explain what you want to gain from starting to throw dirt?

    Don’t even start on this one.  I will leave outoging ISO WG1 convener, Martin Bryan, to "throw dirt":

    "The influx of P members whose only interest is the fast-tracking of ECMA 376 as ISO 29500 has led to the failure of a number of key ballots."

    http://www.jtc1sc34.org/repository/0940.htm

    ECMA 376 = OOXML, for those that didn’t know.

    >> on the intense lobbying of IBM in the NBs

    IBM lobbied against OOXML?  I don’t think there’s any argument there.  But so what if it did?  In any vote of any kind, candidates are allowed to argue their case, aren’t they?  And does that make any of what IBM said in that lobbying factually *wrong* (spelling mistakes aside)?

    In any election that I’ve ever voted in, there’s always been party functionaries outside the polling station, urging you to vote for their candidate.  Last I heard, this is perfectly acceptable behaviour.  I mean, you are free to ignore them, aren’t you?  (And let’s face it, most people do!)

    I’d say that there’s a big difference between that on the one side, and on the other side, discovering that your vote doesn’t actually matter, because your opponent has drafted in a huge number of poor, foreign nationals and given them passports to your country (on condition that they vote for who they’re damned well told to).

  31. Mike,

    " I don’t believe that ISO does any pre-vetting in this case; it assumes that ECMA has already done that before Fast Track submission."

    Thai is not correct – SC34 voted on allowing EOOXML as a fast tracked proposal.

    ""The influx of P members whose only interest is the fast-tracking of ECMA 376 as ISO 29500 has led to the failure of a number of key ballots.""

    Yes – and if you pay attention, he talks about P-members as do I. The responsibility of voting lies with the individual members of SC34 – the national bodies. Even if all members of all national bodies decided to quit – each national body can vote "abstain" if they choose so. Please again do the cross check on the yes-voters for DIS26300 and non-voting countries in SC34 now.

    I’ll ask you again:

    Could you please explain what you want to gain from starting to throw dirt?

    What is the purpose?

    "And does that make any of what IBM said in that lobbying factually *wrong* (spelling mistakes aside)?"

    Well – some of it was wrong and with those issues popping up in various national bodies, you get the idea of the effort IBM paid to this.

    "In any election that I’ve ever voted in, there’s always been party functionaries outside the polling station, urging you to vote for their candidate.  Last I heard, this is perfectly acceptable behaviour.  I mean, you are free to ignore them, aren’t you?  (And let’s face it, most people do!)"

    So when IBM does it, it’s behaviour as expected … but when Microsoft does the same – the result is Microsoft stooges?

    "I’d say that there’s a big difference between that on the one side, and on the other side, discovering that your vote doesn’t actually matter, because your opponent has drafted in a huge number of poor, foreign nationals and given them passports to your country (on condition that they vote for who they’re damned well told to)."

    So this will more or less be the same situation where Kenya – a relatively poor African country – sent in by far the largest list of objections of OOXML being allowed access to fast track? I wonder how many passports exchanged hands here.

    Please – stop throwing dirt – it’s really counter-productive for any intelligent discussion. The task at hand is dealing with the dispostion of comments and to decide if the fixes are good enough for final approval.

    :o)

    /Elf

  32. Robertlilly says:

    @Jesper

    You said is best and clarified Microsoft position: "but the OOXML vs. ODF battle is not about the formats; it’s about market control."

    The transition from  Microsoft’s defacto .doc format to a Microsoft defacto XML .docx format is just about that, nothing more and nothing less. It is only about lock-in and force up-grades for the self-proclaimed Overlord of office documents.

    Statements on why OOXML is an "Open Standard" continue to be preposterous. It simply is not on every level. Please refer to David Lane comments as the only reasonable way for Microsoft to be welcomed into the world of "open standards". Meaning open to all and welcomed by all.

  33. Suite aux réunions qui se sont déroulées la semaine dernière , l’ECMA a annoncé quelques modifications

  34. Wu MingShi says:

    Dear Brian,

    Do you think ECMA can publish a working draft. I know about the problem with NBs comments should be kept confidential and it is not ECMA’s right to unilaterally waive this confidentiality clause.

    However, there are several ways around it. The low tech way will be to "black out" the confidential part. Another will be to publish only the proposed changes to the already public OOXML proposal.

    Things like using ISO dates alongside integer values raise anxiety level. On one hand, it is great that we can use ISO dates. On another, does it means conforming implementation must implement both integer and dates? There are other questions that needs answering as well, such as is the ISO dates restricted to the same range as the integer dates format etc.

    The sooner we can discuss these issues the faster we can work out issues like this and this can only be to the benefit of OOXML ISO process.

  35. marc says:

    congratulations Microsoft

    you have managed to use the fast-track mechanism

    to actually develop a "standard" in real-time

    ( those who are luckly enough can track the real-time progress at http://www.ecma-international.org/dis29500_brm )

    another Microsoft "abuse of the system"

    keep the rushing Microsoft! party on!

    What standard do you want to develop today(tm)?

      marc

  36. Rob Weir says:

    Jesper, you are incorrect in saying that SC34 voted to accept OOXML as a fast track submission.  I’d be interested in hearing where you got this false impression from.

    First, SC34 does not vote anything on Fast Track submissions.  JTC1, the parent committee, does that.  Second, even JTC1 has no input on whether a Fast track is accepted or not.  The JTC1 Directives 13.1 clearly say that "The criteria for proposing an existing standard for the fast-track procedure is a matter for each proposer to decide."  There is nothing in the process where JTC1 or SC34 NB’s are asked whether they accept the submission.  The only opportunity they have is to vote ‘No’ after the 5 month ballot.

    This lack of a "gag reflex" in JTC1 procedures is a defect in the system, IMHO.  I think it would have been better for everyone if OOXML had been quickly examined by experts back in 2006, the presence of serious flaws noted, and the submission immediately rejected at that point.  Ecma could have then used all of 2007 to make fixes, rather than trying to do it all in a few short months.  

    In any system, human or computer, the ability to have a fast failure mode is essential.  If it is impossible to fail any Fast Track in JTC1 until the proposal has been in the process for 15 months, then the system has a serious problem.

  37. Mike Brown says:

    @Jesper

    >> Kenya – a relatively poor African country – sent in by far the largest

    >> list of objections of OOXML

    My analogy of rigging an election by "drafting a huge number of poor, foreign nationals" was just that: an analogy.  I did not mean to imply any direct correlation between "poor, foreign nationals" and the financial status of Kenya or any other ISO member nations.  My apologies to those countries for any unintended offence.

    >> So when IBM does it, it’s behaviour as expected … but when Microsoft

    >> does the same – the result is Microsoft stooges?

    You see no difference between arguing your case, and packing committees.  Between lobbying and vote rigging?

  38. gareth_horton@datawatch.com says:

    @Rob: I think it would have been better for everyone if ODF had been examined by experts, (even a 100th of the number involved in reviewing OOXML) the presence of serious flaws noted, and the submission immediately rejected at that point.  OASIS could then have made fixes, rather than trying to do it all in a few short months.

  39. Reggie says:

    Hypocrisy, thy name is Rob Weir.

    Rob, you have some nerve saying that OOXML should’ve been examined by "experts" before being submitted for ISO fast track when the ODF format that was submitted to ISO is full of holes itself.  Have you no shame at all?  Has spinning FUD for the past two years totally desensitized you to your own hypocrisy?

    And here’s another point.  You and your employer, in trying to convince governments to mandate exculsuve use of ISO ODF 1.0, have been telling those governments that IDO ODF 1.0 is sufficient for their needs when you know damn well that it isn’t so.  You are *lying* to these governments.  Brian Jones is too polite to call you out for what you truly are, a two-faced liar.  

    And that’s what IBM pays you to do.  There is no "Rob Weir" on the OOXML side, a minister of propaganda whose sole purpose in life is to spread FUD, half-truths, and outright lies.  Your mother must be so proud.

  40. Rob Weir says:

    @Reggie,  Microsoft seems to whine alot whenever they have competition. Is this for lack of practice?

    Rather than changing the subject, in the "hit and run" style of the numerous Microsoft fanboys here, why don’t you address the question of the quality of OOXML as it was submitted to JTC1.  Do you believe it was a prudent thing for the  proposal to be made in this condition?  Was it respectful of the time and resources of JTC1 and JTC1 NB’s who had to review this seriously flawed 6,000 page specification?  

    You imply that you were not pleased with the quality of ODF.  Do you believe that this excuses or justifies the submission of OOXML? Should the submission of OOXML then justify or excuse the submission of further standards by Microsoft or others of over lower quality?  Where does this logic take you in the end?  Is the goal to be the lowest quality standard ever?  Is that your aspiration?  Is that the best you can come up with?  Is that your argument for ISO approval?

  41. Rob,

    I appologize for not being clear (or actually wrong about this), because as you said SC34 did not vote for allowing OOXML access to Fast-Track – the JTC1 secretariat did. I am sure you already knows it by heart, but otherwise look at section 13.3 and 13.4 in JTC1 Directives ( http://www.jtc1sc34.org/repository/0856rev.pdf ). So you are correct – SC34 did not vote on accepting OOXML on fast-track, but the fast-track DIS procedure consists of a initial 30-day review period where objections from NBs could – and was – submitted. JTC1 then decided based in these that OOXML could actually continue with the 5-moonth review. You conviniently left this part out.

    But, you know, this actually clearly hightlights the problem with what you are saying. It is not as much /what/ you say – it’s what you do /not/ say. An example of this was you using the contact list of SC34 as basis of your argument, that very few NBs have asked for the password+username of the protected website. For anyone outside of the process they might look at the list and think: "Gee, that’s a short list" … but the list has nothing to do with what you are saying. Based on the list, Denmark should not have asked for access – but we have.

    :o)

    /Jesper

  42. skc says:

    Rob, sounds to me like you want the world to hold OOXML to a higher standard than they did with ODF. Thats oddly both noble and sinister.

  43. hAl says:

    [quote]@Reggie,  Microsoft seems to whine alot whenever they have competition.[/quote]

    It does seem IBM has now moved into thinking of blocking standardization as a tool for competition rather than having their software doing the competition.

    You can basically summerize the IBM viewpoint as. We are against OOXML because we chose ODF for our Office suite productline.

    Even if 99% of comments submitted to the national bodies by IBM were solved in this standardization proces their position would still be against standardization.

  44. Rob Weir says:

    @Jesper,

    The 30 day period was purely for raising the issue of contradictions with existing ISO or IEC standards.  That was the only topic that could be discussed.  That was not a period where objections could be considered that the standard was too large or inappropriate for Fast Track processing, or that the quality was too low. The Directives have no "quick failure mode" for proposals that are clearly inappropriate.

    The JTC1 Secretariat merely decided that the numerous allegations of contradictions could not be resolved, and that the process should proceed to the 5 month ballot. this is allowed  by the Directives.  

    You earlier said, "SC34 voted on allowing EOOXML as a fast tracked proposal".  I think it is clear that this statement was not accurate, or even close.

  45. Dear Rob,

    "You earlier said, "SC34 voted on allowing EOOXML as a fast tracked proposal".  I think it is clear that this statement was not accurate, or even close."

    Yes … and I said:

    "I appologize for not being clear (or actually wrong about this), because as you said SC34 did not vote for allowing OOXML access to Fast-Track – the JTC1 secretariat did."

    What part of this is not clear to you?

    :o)

    /Jesper

  46. Rob Weir says:

    @Jesper,

    Just pointing out that your statement that the JTC1 Secretariat decided that OOXML was suitable for Fast Track processing was also inaccurate.   No one other than Ecma has a say in whether a proposal is suitable for Fast Track processing.

    Note that this process has been far from clear or easy-to-understand. JTC1 Directives are poorly written, and parts of the process have been improvised.  So I don’t blame anyone for getting confused.  Not everyone has had the opportunity to study this in depth.  

    I will continue to point out errors where I see them.   And you, I’m sure, and others, will continue to challenge me to remain accurate.

  47. Doug Mahugh says:

    The last week has seen some interesting discussions and useful how-to posts on Open XML blogs … Three

  48. Brian Jones blogged recently about the latest round of proposed changes to the Open XML standard. If

  49. Brian Jones blogged recently about the latest round of proposed changes to the Open XML standard. If

  50. Robertlilly,

    "The transition from  Microsoft’s defacto .doc format to a Microsoft defacto XML .docx format is just about that, nothing more and nothing less."

    Seriously – has this first dawned upon you now? If you think that the fine folks at IBM, Sun, Novell etc have thought for one second about “the little guy”, I am afraid that you are sadly mistaken. ODF has one single purpose: to be able to tap into the Microsoft Office install base. For years the competing Office application vendors have tried to break the dominant position of Microsoft on desktop office applications – but with little luck. They finally did the only reasonable thing and worked together to create a competing format. Microsoft and their opponents have just chosen different way to preserve their revenue streams – and naturally you might like one more than the other. This reason for IBM releasing their Symphony suite as free is not to make you happy. It is to spread the usage of ODF in the market place – and by that create a market for their IBM-hub focused primarily on ODF. However – if you are so, pardon my choice of words, naïve to think that ODF has been created for “the community”, I’ve got some emails from some guys in Nigeria, that I’d like to send you :o).

    Now, what the ODF-guys have brilliantly done is to round up the entire OSS-world in support of ODF, but seriously, guys, it’s not about you … it’s about them. Whenever I hear someone talking about “interoperability-problems with OOXML”, “lack of transparency” and “platform-dependant issues”, for my inner eye I always picture me Rob sitting behind the curtain in his puppet-theatre, diabolically laughing: “Dance, puppets … dance!!!”.

    All the best,

    :o)

    /Jesper

  51. Well, speaking about respecting the national bodies, you should all remember, that the national bodies are actually pretty sovereign in choosing what to work with and what to pay attention to. If they don’t have an interest in a subject – they are under no obligation to deal with it. There is at the moment a huge interest about topic-maps – being lead in SC34 by primarily Steve Pepper from Norway. But I think it would be fair to say, that the enthusiasm is not “equally distributed” in SC34. Some countries are more interested in some areas than others – and this is really the corner stone of the work in ISO. This was actually most clearly seen when IS 26300 was passed through ISO. At the time, in Denmark, there was absolutely no interest for it at any level – so no mirror SC was set up in Denmark and Denmark voted “Abstain” at the final ballot … cos’ no-one really cared about it. This is clearly not the case now, where about 30 organizations are represented in Denmark in dealing with DIS 29500. This is, of course, their choice. It is not “mandatory homework”. You are perfectly right, that the work being done in SC34 with regards to DIS 29500 is many times bigger than the work done dealing with IS 26300, but I really think it is a good thing and in perfect spirit with the intentions of the JTC1-procedures.

    Maybe the large interest is caused by the national bodies realizing what happens, if you don’t give a crap about a given proposed standard; you risk ending up with a standard, that is under-specified, poorly documented, ambiguous and extremely difficult to implement … as was the case with ODF.

    :o)

    /Jesper

  52. Mike,

    "You see no difference between arguing your case, and packing committees.  Between lobbying and vote rigging?"

    Yes – and you are perfectly right. It is clear that there is no doubt that IBM has rallyed up their allies and made them join the NBs in order for them being able to carrying out their DOS-attack on the JTC1-process. It is clear that the IBM attack on DIS 29500 has had three fronts through the process:

    1. "OSS-community on steoroids"

    2. Flooding the NBs with an insane amount of irrellevant comments (and some good ones too, btw)

    3. Packing the NBs with their allies in order to reach deadlocks in the process with the sole purpose of prolonging any decisions.

    Sadly, I think they have succeeded on all three acounts.

    /Jesper

  53. omz says:

    >Maybe the large interest is caused by the national >bodies realizing what happens, if you don’t give a crap

    >about a given proposed standard; you risk ending up with

    >a standard, that is under-specified, poorly documented,

    >ambiguous and extremely difficult to implement … as

    >was the case with ODF.

    Jesper "Fud" Stocholm

    :-)

             omz

  54. Mark Baird says:

    We have built a wordML to HTML transform and while I appreciate the XMl out of Word I do not appreciate the need to understand the logic that Word uses to interpret the XML.

    A perfect example is the way Word exports it’s paragraph borders. To read and understand this XML you need to know Word’s logic when it imports wordML and draws the borders. The wordML is not what you think it is. How does this make it an open standard if the XML is not  as straight forward as possible?

    <wx:pBdrGroup>

    <wx:borders>

    <wx:top wx:val="solid" wx:bdrwidth="60" wx:space="10" wx:color="auto"/>

    <wx:left wx:val="solid" wx:bdrwidth="60" wx:space="10" wx:color="auto"/>

    <wx:bottom wx:val="solid" wx:bdrwidth="60" wx:space="10" wx:color="auto"/>

    <wx:right wx:val="solid" wx:bdrwidth="60" wx:space="10" wx:color="auto"/>

    </wx:borders>

    <wx:shd wx:bgcolor="FFCC99"/>

    <w:p>

    <w:pPr>

    <w:pBdr>

    <w:top w:val="single" w:sz="24" wx:bdrwidth="60" w:space="10" w:color="auto"/>

    <w:bottom w:val="single" w:sz="24" wx:bdrwidth="60" w:space="10" w:color="auto"/>

    <w:between w:val="single" w:sz="8" wx:bdrwidth="20" w:space="10" w:color="auto"/>

    </w:pBdr>

    <w:shd w:val="clear" w:color="auto" w:fill="FFCC99"/>

    </w:pPr>

    <w:r>

    <w:t>Multiple paragraphs with a top and bottom border</w:t>

    </w:r>

    </w:p>

    <w:p>

    <w:pPr>

    <w:pBdr>

    <w:top w:val="single" w:sz="24" wx:bdrwidth="60" w:space="10" w:color="auto"/>

    <w:bottom w:val="single" w:sz="24" wx:bdrwidth="60" w:space="10" w:color="auto"/>

    <w:between w:val="single" w:sz="8" wx:bdrwidth="20" w:space="10" w:color="auto"/>

    </w:pBdr>

    <w:shd w:val="clear" w:color="auto" w:fill="FFCC99"/>

    </w:pPr>

    <w:r>

    <w:t>Paragraph</w:t>

    </w:r>

    </w:p>

    <w:p>

    <w:pPr>

    <w:pBdr>

    <w:top w:val="single" w:sz="24" wx:bdrwidth="60" w:space="10" w:color="auto"/>

    <w:bottom w:val="single" w:sz="24" wx:bdrwidth="60" w:space="10" w:color="auto"/>

    <w:between w:val="single" w:sz="8" wx:bdrwidth="20" w:space="10" w:color="auto"/>

    </w:pBdr>

    <w:shd w:val="clear" w:color="auto" w:fill="FFCC99"/>

    </w:pPr>

    <w:r>

    <w:t>Paragraph</w:t>

    </w:r>

    </w:p>

    </wx:pBdrGroup>

    The other problem with Word is the inconsistent results in the wordML as the round of edit cycles increases (missing elements to help determine font size, missing font information, etc). I have plenty of examples. My guess is that this problem is with the core application and people working around problems in the code base instead of properly fixing the code base. The XML is only as good as the in memory DOM structure.

    The market will start to see these issues and decide which open format they will adopt.

  55. hAl says:

    @Mark Baird.

    You dropped some reformatted XML in your post but you have not actually explained what is actually wrong with it or the way Word interprestes it.

  56. Ian Easson says:

    Mark, you ask "How does this make it an open standard if the XML is not  as straight forward as possible?"

    There’s nothing wrong with simplicity, unless it interferes with the prime design intent of something.  So, don’t forget that the design intent of OOXML is to faithfully represent the contents of existing billions of Office documents out there.  Although I have not looked at your example, my comment is the usual answer to questions like "Why does OOXML do things a certain way?".

  57. Reggie says:

    Rob Weir continues to avoid the issues I raise to him.  He now seems to admit that ODF 1.0 is full of holes (but says that’s no reason to hold OOXML to that same low standard), while at the same time fails to address the fact that IBM is lobbying governments to standardize on ODF 1.0 by misleading them into thinking that that standard is sufficient for their needs.  This guy is unbelievable (in more ways than one).

  58. ben says:

    Who cares about the thoughts of a person that works for CompTIA’s lobbying arm?

  59. omz,

    >Jesper "Fud" Stocholm

    Well … I always like it when someone goes after me and not my words – it usually means that I’m doing something right.

    :o)

  60. Aidan Thornton says:

    Most of the issues in the ODF specification listed in the linked official report are typos or bad grammar where the meaning is still clear. Certainly worth fixing, but nothing critical. OOXML has worse issues, and leaving them to a later version means that we’ll be stuck with them forever.

    (Incidentally, the reason ODF doesn’t specify spreadsheet formulas is that doing it right is major task in itself due to widespread dodginess in existing implementations. The only reason Microsoft could include them in OOXML is that they basically just documented whatever Excel did, and this is wrong in various cases.

    I believe OASIS are currently working on a formula standard for ODF that balances backwards compatibility and mathematical correctness. In fact, Wikipedia reckons that they got their first draft out before Microsoft released the formula specifications part of the OOXML specs.)

  61. Ecma is posting a new status update today. We’re in progress updating 500+ new national body comments.

  62. Ecma is posting a new status update today. We’re in progress updating 500+ new national body comments

  63. First of all Merry Christmas to everyone who reads this blog and happy upcoming New 2008 from me and

  64. Owner Blog says:

    First of all Merry Christmas to everyone who reads this blog and happy upcoming New 2008 from me and