It’s hard damned work trying to make this process look so bad

Ecma TC45 and the project editor are making great progress in the resolution of comments we received from the national bodies a few months ago. The result will be an even better spec than we had last year, and this really helps highlight the benefits of the standardization process. Of course there are a few folks out there who don’t care about actually seeing the spec improve, they just want to see it go away (it’s bad for their business plans). So we now have Rob Weir from IBM hard at work trying to find a way of turning the positive into a negative.

Well it looks like he’s taken his latest shot at the process, so let’s see if it sticks. I’ve had a few folks ask me lately about the comment resolution process in ISO, and specifically why the comments are currently password protected (meaning only the national bodies involved in the process can view them). I was pointed to this blog post by Rob Weir (IBM) on the recent batch of proposed resolutions to NB comments posted on Ecma’s website, and while I rarely have the energy to read an entire Rob Weir post, I gave it a shot. He opens by claiming the following: “First, Microsoft has managed to get JTC1 to clamp down on information“.

Theater at its finest… <g/> Of course Rob knows that Microsoft has no say in this process at all. It’s up to JTC1 to set the rules, and JTC1’s ways of functioning are fully documented (and no waiver has been granted to Ecma or Microsoft). IBM should be objecting to ISO/IEC, not to Ecma or Microsoft. Why isn’t it doing this?

It’s an unfounded allegation that shows the typical Fox News style of sensationalism Rob is practicing (sorry if there are any Fox News fans out there). As I said in the past, I would love to have the comments and responses public. There is a lot of great work right now coming out of Ecma TC45 as we help the project editor pull together the proposed responses for each issue, and I think those of you following along will be very impressed with the analysis that has gone in to each comment. Coincidentally, I’m in Kyoto right now at Ecma TC45 meetings and folks on the TC are definitely excited about the work being done (and look forward to the feedback we’ll get from the national bodies).

For the layman, finding your way around ISO/IEC resources is not always easy, but if you try hard enough, you can find most answers you’re looking for. If you want to build up your own opinion on this particular matter, you should definitely go through the simple guide I’ll included at the end of the post. Another great source of information you should check out before jumping to any conclusions is this comment posted by Jan van den Beld, former Secretary General of Ecma International in Geneva, who explains how things work within JTC1 (Ecma doesn’t make the rule here). Also, check out the “Ballot Resolution Meeting FAQ”, which gives the full picture of that part of the process.

If folks want to compare the OASIS Technical Process to the JTC 1 Directives, go for it. Just don’t try to put Ecma at fault. The reports on ODF 1.0 were made public by the SC34 Secretariat – so if there are issues over making DIS29500 reports public, ask the SC34 Secretariat. Or is IBM claiming that Ecma (or Microsoft) has now got the SC34 Secretariat captive? <g/>

The rules of JTC1 may be questioned, but that should be a debate, not an accusation.

The next claim Rob Weir (IBM) asserts is that a “few NB’s have requested the password” [NB= National Bodies]. He provides with this link to the National Body Membership list.

From what I understand, the password was distributed to all NBs having voted; they didn’t need to request it. Nevertheless, National Bodies who are part of the process have access to the proposed resolutions and can start working on analyzing them, thus giving them more time before the actual discussion meeting in February. There are no roadblocks in the way to prevent them from getting access.

This step of publicly documenting the responses ahead of the January 14th deadline is something that the editor and Ecma TC45 proposed early on to ISO and asked ISO if it was ok. We were told it was a good thing to do, but since the national body comments are considered to be internal documents, they wanted us to put a password on the site so they could control who got access. Remember, that if the editor wanted to he could have just kept everything to himself until the 14th of January. It’s a bit annoying to see Rob trying to stir up more controversy in something that is actually a pretty big positive thing (but those are just my personal feelings).

National Bodies will make their decisions in March based on the final Disposition of Comments report approved at the BRM – that is the way the process works, and IBM knows this. Again, if IBM wants to change the way it works, they should be in dialogue with JTC 1 about it!

Ecma TC45 has been hard at work helping to come up with good resolutions based on all the feedback. It’s a lot of work and it’s really progressing well. There were some really good suggestions, and I think we’ll see that this round of review will result in an even better spec than we had at the end of last year.

But for now we still have a ton of work to do!

Background: A simple guide through ISO/IEC JTC1 “Procedural document”
(I’ve highlighted the key areas):

  1. Look into the ISO/IEC JTC1 “Procedural document” (,
    Here are the rules for documents that must be kept private:

  2. This part lists the documents and how they should be accessed


  3. Then here is the key to read the table:


Comments (52)

  1. D.C. Parris says:

    I’m curious as to whether ECMA TC45 faces different rules than the ODF folks did, and if so, why?  Why did the ODF process manage to be (or at least appear) more open than the current ECMA process?  It just seems like they would have had to abide by the same rules as ECMA TC45, and add password protection as well.

  2. carlos says:

    >The result will be an even better spec

    >than we had last year

    what are you taking about?

    you are not "improving" a specification,

    just for "make the things better"

    you are now doing the home-work that you must have done

    *before* submitting this +6000 page beast

    to this frenetic rushed, money-driven, standardization  party

    please, skip the PPRR preset-phrases like this

    and do just one thing:


    the "better" format will come as a consequence



  3. David Lane says:

    Gotta say, Brian, this process looks totally shambollic.   I don’t think anybody is dumb enough to believe that you have honourable objectives.  Neither Microsoft nor ISO have provided anything like a reasonable answer to the basic question:  why do we need 2 standards in the exact same space.  We’re quite tired of all the drivel about Microsoft’s need to innovate, because we all know how ridiculous a fiction that is.  Why not just say it like it is – Microsoft can see its gravy train (*ahem* criminal monopoly *ahem*) slipping through its fingers and is fighting tooth and nail to slow the descent.  Forgive me for not being sympathetic.  

    Why can’t Microsoft work within the community  to improve the current Office Document standard, ODF?  After all, isn’t Microsoft on the OASIS committee that designed ODF?  Did they always plan to mount a rear-guard operation to undermine the ODF standard?  Too open for you?  By that I mean, actually open, as opposed to the name-only openness of OOXML…  Pathetic.

    Spin only works if people don’t have access to the truth.  Sadly, Microsoft, although the desktop’s in your control (well, the part that’s not being bot-herded), the Internet’s hostile territory for you, and your flock of chickens is coming home to roost.

    Good luck,

    Dave Lane

  4. says:


    I don’t believe ODF ever pulled together responses to the comments. They never had a BRM, and I believe most of the comments were left unanswered as they recieved enough "yes" votes to pass initially. So this specific rule never applied to them.



    I understand your view here, but it’s pretty naive. Any standard will have errata (usually a large amount). That’s why it’s important to have a good maintenance agreement in place so you can continue to improve. Nothing is perfect (if you disagree, please point me to something that proves me wrong).


  5. says:


    I don’t think you have access to all the facts. Just look at the history. ODF and OpenXML were developed in parallel. ODF had one set of goals and Open XML had another. These are the facts.

    If we used ODF and didn’t create Open XML, 99% of our customers who don’t give a lick about file formats would have to suffer. They would lose things from their files, and they would then lose faith in the product. They would go back to using the binary formats, and that would do absolutely no good.

    If the ODF guys decide to change their design goals, then they are welcome to reference the Ecma standard and change ODF to also be compatible with the existing base of Office binary documents. It would require a good re-working of their existing format though, and I doubt they will want to do that.


  6. David Lane says:

    Thanks for the response, Brian.

    I can’t figure out why you keep doggedly trying to make it look like Microsoft’s unilateral attempt at an alternative to the official ODF standard is "open" when it’s not (I’m referring to your references to "OpenXML").  I believe some of the 662 comments point that out…  

    Regarding "the facts," perhaps you can point me to a source (preferably non-Microsoft) to support your claim of the parallel development of ODF and OOXML?

    As for your assertion of 90% of MS Office users suffering – they don’t seem to be suffering that badly when they make the switch to, which a) reads and writes legacy MS Office binary file formats as well as or better than MS Office, and b) seems to save everything in ODF format without suffering from "data losses" or insufficient expressiveness… or am I missing something.  Is there some extraordinary font of additional functionality in Microsoft’s latest Office which ODF is simply too inexpressive to represent?  Perhaps you could identify it for me.  Speaking of the facts, if I’m not mistaken, you’re just spouting unsupported assertions.  

    As for "the ODF guys" – aren’t you and your Microsoft cobbers part of OASIS?  Why didn’t you contribute to it?  Surely – I’m assuming you’re a smart guy here – you can see the fundamentally wrongness of trying to come along with an overly complex, product-specific, dependency ridden, patent-encumbered suspect mess and try to pass that off as an "open" standard (the gall) in opposition to a perfectly good pre-existing open standard written by a group of which *your company is a member*!!  

    It would be like America coming along after hundreds of years of a world (barring 3 or 4 rogue countries) contented using the Metric System and saying "wow, that Metric System… yeah, it’s ok, but we have this great idea for allowing people to more expressively measure new and old things alike – introducing Imperial Units!"  Mate, that dog won’t hunt.

    All the best,


  7. Dave S. says:

    OK – Let’s ponder this a bit. Hmmm. Data is kept in magenetically encoded domains or is physically dented into solid material or is kept in a vast array of tiny, highly insulated semiconductor capacitors – or some similar medium.

    It takes some know-how to put the date there and then get the data back and it’s -all- binary**. Then along come the XML folks and point out that, with a limited character set and some rules, this binary data can be read by simple programs, not requiring large applications to process.

    Brian posits that without using MSO-XML encodings (a binary format subset), Microsoft customers will have to suffer. "They will lose things from their files and they would then lose faith in the product." Like the big Red X? Like the workflow to publish mathematical and scientific papers containing mathematical formulae? I digress.

    What will cause this suffering? Am I suffering right now for not having gotten Office 2007? Perhaps I’m in the 1% that won’t suffer.

    I believe the crux of the matter lies in Brian’s statement – "They would go back to using the binary formats, and that would do absolutely no good."

    (Umm no. They would not go back. They would simply continue as they have, and as they do with XP. And they would still interoperate with most everyone else who had an app designed to do so within the last 10 years.)

    This begets the question – What was the good  sought in the use of XML as a basis for the MS Office 2007 suite that is independent of changing the format? Let’s skip the trite answers re XML is good, etcetera. The meat is – What functions are being added that could -not- have been handled any other way?

    Obviously the binary formats could have been documented, including a usage and concept explanation guide. They must have been, to ensure the new formats were capable of representing every masterstroke (and pimple.)

    The XML interface to the outside world could have been added as a separate file zipped with the binary file, allowing any other content to be zipped up with it – one very good reason for using zip on its own.

    No more intelligence can be added to the billions of documents with their mis-fit conglomerations of styles – documents, many of which no doubt depend on the file on machines long ago crushed and discarded, leaving the original formating in tatters. There is little other meta-data in them. What there is could be extracted by MS Office components and zipped as separate XML.

    Nor is intelligence discerned in the variety of crippled documents. For example, many are mangled by auto-numbering madness. Most every document I see where auto-numbering has been used, there are additional numbers carefully placed in line with the list by hand with space characters, dozens sometimes, used to carefully align the new digit with the old.

    For some reason, though, the auto-numbering is invisible to copy and paste, so a numbered list in MS word can’t be lifted and pasted into a text only e-mail. Nope. The numbers need to be re-added. Maybe that suffering is cured in 2007, but, still not XML related.

    Mayhap units were added to Excel so that people will be warned when they add $25 to 35 miles. A useful change, instead of partly shading cells to give visualization to the number already there.

    There must have been a mission statement once – Microsoft will use XML as the basis for the Office suite because…

    I’m pretty sure it did not end with …it will mummify all the old ways as a brand new way.

    Yet, isn’t mummification of the old MS binary formats the primary goal presented as -the- reason for ISO approval of MSO-XML?

    ** Everything is analog until you get really small scales, where it looks binary again, but at tinier scales yet, physicists are trying to make it infinitely analog. Go figure. For comp-sci folks, binary just simplifies it.

  8. I absolutely knew we were going to see a posting like IBM’s Rob Weir’s latest . There was simply no question

  9. I absolutely knew we were going to see a posting like IBM&#39;s Rob Weir&#39;s latest . There was simply

  10. nksingh says:

    @David Lane:

    Brian helpfully posted a history of the two formats from conception to about a year ago towards the beginning of this whole standardization debacle.  I suggest you read it first before accusing anyone of making unfounded assertions.  It’s common sense that ODF and OOXML were produced in parallel when you consider that they were both released about a year apart and it takes more than a year to develop a full document format.  It should also be noted that Microsoft has had XML formats to hold their documents since Office 2k3, which was probably mostly coded in 2001 and 2002.  

    There’s some benefit to thinking about things from both sides, rather than just the reflexively anti-Microsoft one.  You’re likely to learn more that way.

  11. Rick Jelliffe says:

    Thanks for posting these, Brian. Rob is making a sterling effort to maintain his mischief-making with only limited material, but he is largely preaching to the converted, as far as I can see. I hope he is paid by the word! A nice gig 🙂

    Here is a comment on the ODF standardization, compared to OOXML. ODF did not have a substantial number of comments for several reasons:

    First, ODF 1.0 had a substantially more limited scope and its reviewers accepted that there was an ongoing process of development and improvement (this pragmatism and goodwill being noticeably lacking in review of DIS 29500),

    Second, ODF bypassed the normal review channels  in several countries, I have been told by several experts; it would be voted on by the JTC1-equivalent committee at the National Bodies not the SC34-equivalent committees; this means a vote by people not necessarily interested in the technology,

    Third, one of the ODF editors was Patrick Durusau, a long-serving and respected member of SC34 who made sure the draft had ISO-isms (these are not necessary in a fast-tracked draft, however for such matters as conformance language, it is the discipline of doing things consistently that is the win from adopting ISO-isms even when not strictly necessary)

    Fourth, experienced reviews approach standards quite systematically, and fairly top-down: so conformance language issues and systematic errors are high on their priority; inexperienced (and parroting) reviewers tend to find only low-level problems–the devil may be in the details and experienced reviewers don’t ignore detailed errors of course.

    Fifth, ODF re-uses the vocabularies of external standards quite a bit (which is not the same as adopting the standard: look at the grabbag use of SVG or XSLFO element names, for example) which have already had long-term detailed review.

    Sixth, many nations failed in their obligation to independently review OOXML, but merely regurgitated the same material. The review process was subverted to the extent that it became a form letter campaign.

    Seventh, many people are ignorant of what a voluntary technical standard is, and think that somehow it automatically becomes a regulation. Consequently they had a deal of passion to motivate them.

    Eighth, because of the misleading information spread about time periods (e.g. "only 1 month for review") and misleading or inept information about procedures (e.g. "potentially anything that could confuse someone may be a contradiction") reviewers adopted a scatter gun approach where even the flimsiest of issues was reported.

    Ninth, the echo chamber effect meant the same issues were circulated, giving them spurious credibility. One of the most ludicrous examples is that accessibility review, where the lack of documentation about a decade old bug in one version of Word concerning Japanese "Fullwidth" spaces (we don’t have them in the West, was given as a reason why the standard could not be implemented: accessibility people concerned with fullwidth spacing minutae, give me a break!

    Now, all that being said, there are several different issues with making things available.

    For a start, the ISO Copyright Officer needs to be approached to make sure that DIS 29500 is available from the Publicly Available Specifications site at ISO. (Not to be confused  with the PAS type of ISO standard, nor the PAS process for fast-tracking IS documents through JTC1.) This then removes one of the necessities for drafts to be kept private ("secret" is entirely the wrong word here.) Under that basis, it may be reasonable for ECMA to request that any editor’s draft be made available to the public too.

    ISO usually does not like near-final documents making their way into the world, because people relying on them can implement things wrong against the final standard. But this is a case where there is substantial reason to expect that people wouldn’t be too confused. (Of course, there is a good counter-argument, that it is counter-productive to provide fodder for  the kind of manipulative antics we have seen.)

    But as for the specific ballot comments, that should be a matter for each National Body to decide on their publication. As for specific dispositions of comments by the editor, they are targetted at the NBs not the general public, and the editors can keep their focus on the actual comments at this stage, not subsequent ones. It is up to the National Bodies to get their national public involved, and certainly anyone interested should have long ago tried to join their NB’s committees and get up to speed.

    In May of 2006 I was at the ISO meeting in Seoul when people were discussing the upcoming ODF vote. (This was out-of-session; I am not giving up any confidences.) The question was asked, "What dirty tricks could Microsoft play to get ODF halted?" and one of the suggestions was to get a kind of denial-of-service attack on a Ballot Resolution Meeting: have so many issues that they could not be dealt with in time. As it turned out, Microsoft did not oppose ODF getting passed as an ISO standard however, I am completely sure that long-standing SC34 members, even those who do not care for either OOXML or ODF, will take a very dim view of any attempts at those kinds of dirty tricks.

    I don’t care if Rob blackens Microsoft’s name, or puts the most elaborate spins on the technical decisions in Office Open XML. But his willingness to blacken anyone who dares to disagrees with him, in particular the willingness to link to sites that claim corruption on no evidence and the eagerness to blacken ISO’s reputation, are very reckless.

  12. Alan Bell says:

    I would really like to be able to at least highlight the NB comments for which Microsoft/ECMA have a response.

  13. David Lane says:


    Thanks for your input, but a link or two to some source material to support your facts would’ve been more helpful.  As for reflexive anti-Microsoftness…  well, their dis-ingenuity is difficult to verify (which is why they use it), but palpable to those of us who have been watching them since the early 90s…   They’d rather that we argue the technical minutiae of the proposed standard rather than focus on the basic question of whether it should be considered for standardisation at all…  

    No one has addressed the question of why we would want their unwieldy Imperial Unit standard (OOXML) vs. the existing, far cleaner Metric System (ODF) at all.  Seems to defeat the whole purpose of standardisation if you have mutually exclusive competing ones…  

    Microsoft are clearly taking the gamble that they can leverage their (criminal) monopoly to squeeze out ODF and retain their file format lock-in.  Anybody who thinks otherwise hasn’t been paying attention.

    Regarding Microsoft’s use of XML in older versions of MS Office… so what.  

    Abiword, Gnumeric, Star/Open Office, and various other proprietary and open source projects have used XML file formats longer.  Last year in NZ, a few alert people foiled an underhanded attempt by Microsoft to patent the use of XML as a means for storing generic Office document data…  Thanks to bountiful prior art, it became clear to the NZ patent office that Microsoft had no claim, and Microsoft (rather than suffer loss of face accompanying a total withdrawl) revised the patent application to the point of toothlessness.  

    Microsoft tried and failed to push through similar patent applications in South Africa and other jurisdictions – like the US (damn that prior art).  Microsoft subsequently tried it in places in hopes that it would pass through other national patent system unnoticed despite being rather easily overturned if anyone was paying attention.  What does that say about Microsoft?  

    Microsoft is a deceitful organisation.  We have learned over the past 20 or so years that you trust it at your own peril.  Thus my reflexive anti-Microsofticity.



  14. Fiery Spirited says:

    Well…microsoft and ECMA did not choose the JTC1 rules…but I truly doubt that Microsoft did not review the options before JTC1 was chosen. If Microsoft really wanted an open format they could open their own binary format and grant full rights to use that information by the open source world without fear for patens or other intellectual monoploy methods.

    Microsoft would end with translators to ODF and could drop all the work wasted on OOXML. Why not do it, you did drop the 2003 XML format already so your customers is used to the office threadmill, right?

    "If the ODF guys decide to change their design goals, then they are welcome to reference the Ecma standard and change ODF to also be compatible with the existing base of Office binary documents."

    If ODF can be improved to better support Microsoft only features, Microsoft would only need to tell in what way, and the ODF movement would have every reason to comply. Such support would allow ODF products to convert Microsoft documents better.

    Facts are that Microsoft are not participating in ODF process and until we see the ODF standard process reject proposals from Microsoft there are no grounds what so ever to say that ODF can not be a perfect format for Microsoft also.

    As for the many Microsoft rethorics about compability with legacy microsoft binary formats. I am still looking for anyone that understand how partial representational compability with old formats would make ODF a better standard. You can not load the OOXML document to old versions of Office, can you? So why not develop a ODF patch for old versions instead of the still missing OOXML one?

    We never see any microsoft people claiming that ODF is incapable of describing old microsoft formats since that would open the path for demands of actual proofs about the exact compability problems. Instead we see general FUD about ODF lacking binary compability with old Microsoft formats as one of its design goals. That claim is safe to tell the ignorant since it is vague enough that you can never be called to give evidence about why this ever would be a problem in the real world.

  15. Sam McCall says:

    <blockquote>Of course there are a few folks out there who don’t care about actually seeing the spec improve, they just want to see it go away (it’s bad for their business plans).</blockquote>

    You sound a bit like Fox yourself! Two objections:

    1) Microsoft wouldn’t like ISO to shoot down OOXML, it would be bad for their business plans. That’s not an argument.

    2) Saying everyone who doesn’t want this process to succeed has a vested interest is plainly untrue. I think there’s a perfectly valid case to be made that docx is a major improvement over doc (mainly the fact that it’s documented), but that it’s not suitable as an ISO document format standard (*way* too much cruft).

    My perception is that there’s a demand to save in open standardised formats, because those standards are meant to be neutral, clean, and easy to implement. Pushing an XML version of your internal file format through a process doesn’t give it any of these qualities, and the changes needed aren’t the incremental ones you can do during the ISO process.

  16. Ian Easson says:


    Reading totally uninformed statements and phrases like:

    "Microsoft’s unilateral attempt at an alternative to the official ODF standard.." (Dave),  or

    "overly complex, product-specific, dependency ridden, patent-encumbered suspect mess.." (Dave again), or

    "You can not load the OOXML document to old versions of Office, can you?" (Fiery)

    makes it all the more important that Brian get his FAQ finished, so that he can just point to the correct information.

  17. A says:

    I just want to throw in a little comment that most of you are really missing the point, at least when you talk about ease of implementation.  

    Have any of you ever tried to write a full text processor?  The problem isn’t reading the information from the file, its trying to figure out what is supposed to happen when you’re splitting a table across a page, with the paragraph property "keep with next" turned on, a couple of images with tabs wrapping around them and the cell is merged..oh yes, and maybe you’ve got document grid turned on too.

    Why do you think OpenOffice doesn’t have perfect support for older versions of Word after all these years?  Can you point me to a application that does?  Can you point me to two applications that implement ODF in exactly the same way?  If ODF is so easy and clean, everyone should get it perfect, no?  

    (Just want to add a note that I know nothing about ODF other than what I read here, nor do I have anything against it.  I am very familiar with OpenXML, and well aware of many of its annoying quirks, as well as the fact that Office itself doesn’t implement the format right all the time either!)

    I don’t care how complex the format itself is, it’s putting them together that’s tough.  Excel can store its dates in all kinds of weird and wacky ways, but trust me, it wasn’t that hard to support.  Now trying to draw the cell corners right so that the double border style joins perfectly, now that took time.

    So I will agree that a format can be "neutral and clean", I doubt that any useful format will be "easy to implement".  Plain ASCII text is really easy to implement, but it doesn’t really do all that much either.

    I won’t comment on whether OpenXML should be made a standard or not, doesn’t really matter to me one way or the other.  But it frustrates me sometimes when people pick on some little tiny thing in the format and claim that it will make the format hard to implement fully.  Trust me, I’ve been doing this for over 6 years, I know which half of the problem is the hard part.

  18. Daca v-ati bucurat de un pic de liniste in ceea ce priveste Open XML, ei bine, IBM declara prin Rob ca

  19. Rob Weir says:


    If you read my post in full you will see that I link to the ODF (DIS 26300) Disposition of Comments report.  We had to respond to all ballot comments, just as OOXML will.  In the case of ODF this report was and is publicly available to download.

    Also, you don’t explain why the Ecma process lacks transparency.  Why are the TC45 mailing list archives not public, for example.

  20. Sam McCall says:

    A: you’re right, and I was overstating and simplifying. Apologies.

    I think the scale and the desire to ’round-trip’ features of documents can make it a significant issue though, e.g. implementing graphics using SVG with the help of a library, vs implementing VML + DrawingML to fully interop with MS office.

  21. BigAl says:

    Can Brian or any MS employee please answer one simple question: will it be possible for me, an independent developer, to write software that can FULLY read/write OOXML formatted files? If the answer is no, then it isn’t an open standard. Full stop.

  22. David Lane says:


    Good question.  You might want to further qualify that, however: "without being obliged now or ever to pay licensing royalties or fight patent infringement claims in any jurisdiction."

  23. BigAl says:

    @David: true. However, I want to know if anybody can actually implement the proposal as it stands, from a purely technical p.o.v. So even before I worry about the legal issues, is it technically feasible? (And don’t say "Yes, Office 2007 does it." It does not.)

  24. says:

    Thanks for all the comments folks


    That’s a great bit of background. Thank you



    I think what you’re seeing is that there are some comments from folks with geniune intentions, and then there are those with a malicious agenda. People that want to just get this format killed tried to come up with examples that would resonate with the public. As a result they picked some rather obscure things that don’t really have much impact on the file, but that spun right sounded like a big deal.



    Thanks for pointing that out. It’s surpising how many of the comments you guys decided to blow off (especially the ones from China). They raised some great points, like the weakness in the current field design.

    Also, have you guys sorted out the international issues Israel and Egypt pointed out? From what I can see, OpenOffice still has to break away from the spec and do numbering in their own proprietary format.


    David and BigAl,

    Absolutely you can implement everything. If there is anything that you need to implement but cannot, please let me know.

    Dave, the OSP is there for that purpose. It ensures that we will not ever enforce the IP used for implementing the file formats.


  25. David Lane says:


    Thanks for the response.  Are you sure we can implement everything?  Even the parts of the OOXML spec that refer to legacy Microsoft Office binary file format implementations, for which documentation isn’t publicly available?  

    As for the OSP – perhaps you can provide a link for that, as I’ve seen the TLA, and assume it has something to do with a "promise", but I’m not sure how to find out what it actually means.  As for promises from Microsoft… the road is spotted with the flattened carcasses of companies who put their trust in Microsoft (and its willingness to stick to things like contractual terms, anti-trust legislation, etc.).  How many legal proceedings is Microsoft currently involved in worldwide?  How many will it settle to keep them low-key in the media?  How many will it obliterate by employing more lawyers and delay tactics?  Microsoft’s very good at that sort of thing.

    Let’s make no mistake – as a public company, Microsoft does what it needs to do to maximise profit.   As a hugely powerful monopolist with massive margins and more money (and therefore influence) than most countries, its strategies include compromising ethics where convenient, and breaking (or "lobbying" to  change) the law if the profit is greater than the potential penalty.  Some people call that good business.  I call it a disgrace.

  26. says:


    Yes you can implement everything. Please tell me what you are blocked on right now, and I’ll investigate.

    Here is the link to the OSP.

    IBM just recently copied this approach as well (although they call it the ISP). It is not just a "promise" but is an actual legally binding document. Doesn’t matter if you trust us or not, we can’t take it back.


  27. David Lane says:


    What I’m blocked on:

    1. there has been no good explanation for why doesn’t Microsoft attempt to work with ODF rather than insisting on inventing its own standard based on what’s easiest for it rather than what’s best for the industry and the user.  Your "90% of users" argument (see above) doesn’t fly, sorry.  Why not simply say what everyone knows: Microsoft cannot afford to let its monopoly go, and will do whatever is required to maintain it.

    2. Microsoft used various illegal and unethical means to try to influence the first ISO vote on the OOXML fast track process – bribing voters, influencing rules of order, etc.  These infractions are well documented.  Why should anyone offer Microsoft anything other than contempt now?

    3. Your colleague, Steve Ballmer, publicly claims that Linux infringes on 250-odd Microsoft patents, and yet refuses to name any of them.  Is a corporation that publicly states that Linux and other open source software users owe them royalties for undisclosed patent infringements worthy of anything other than ridicule, much less trust?  

    Those are the first blockers that spring to mind – with time, I’m sure I could enumerate others.  The fact that people are worried about the myriad of technical issues which make OOXML unpalatable is totally missing the point.


  28. says:


    1. Sun created ODF. Microsoft created Open XML. Sun submitted ODF to OASIS. Microsoft submitted Open XML to Ecma.

    ODF was not some well established standard that Microsoft just chose to ignore. They were both developed in parallel. The first versions of Open XML were started almost 10 years ago.

    2. This is just rediculous. Show me proof. There was only one incident I’ve heard of and that was just a new field guy not understanding what was appropriate or not. Meanwhile you had an IBM employee from Germany writing the "official" ISO document for another country. You need to open your eyes a bit Dave.

    3. He’s not exactly a colleague, more like a bosses bosses boss. I don’t believe Microsoft ever claimed that Linux and OSS users owned them royalties, just that there were patent infringements. This is so far removed from Open XML though, that I’m not really the right guy to comment on it. If that’s why you’re against Open XML though, well… not sure what to say.


  29. hAl says:


    I don’t think you can implement all of ODF just using the specification. It leaves a lot more things unexplained than the OOXML spec does.

  30. RobertLilly says:

    Hi Brian

    That fact that you are arguing to the "Open Community" about how OOXML is so open and tranparent should make you sit back from your keyboard and wonder! If MS wants to step into the "Open" world then they should first understand it is about collabaration.

    The hearsay comments about OOXML developed parallel to ODF is irrelavent. MS is a member of Oasis and said we will "wait and see" and then went on this rampage alone, it was formated in a closed enviroment with no collabaration from others and is simply a deliberate departure that is unneeded and harmful to "Open Standards"

    One would understand that a company indoctrinated with the culture by a huge contingency of IP lawyers and lawfirms can take time to change. But maybe these lawyers (MS) do  not care to change or about ISO and open standards, because it gives them less work and hence, MS complete disgregard for the process.


  31. Andy says:

    Armenia, Azerbaijan, Bangladesh, Barbados, Belarus, Bosnia and Herzegovina, Congo, Costa Rica, Côte-d’Ivoire, Croatia, Cyprus, Egypt, Fiji, Jamaica, Jordan, Kazakhstan, Lebanon, Morocco, Kuwait, Nigeria, Pakistan, Panama, Qatar, Romania, Russia, Saudi Arabia, Serbia, Sri Lanka, Syria, Tanzania, Ukraine, United, Arab Emirates, Uzbekistan

    — What an Alliance for the Open XML Standard!

  32. Alan Bell says:

    Martin Bryan was the convenor of WG1, he has just left at the end of a three year fixed term, his end of term report sounds like he is quite happy to be shot of it all, especially the vote stuffing by the new P members.

  33. David Lane says:


    Thanks for the non-clarifications to address my "blockers".  

    Regarding blocker #1: Sun might have created ODF, but it was turned into a standard by OASIS, in which Microsoft is a participant.  If OASIS was able to get ISO to recognise ODF as a standard then bully for them.  Sorry to say it, Microsoft, you *missed the boat*.  Your OOXML is reactionary, and too late.  Why not just get on with doing the right thing and supporting *the open standard*, ODF.  Don’t you see how ridiculous you look trying to make OOXML look like an honest effort to provide value for your customers?  

    If a ragtag team of volunteer (and IBM and Sun) and KOffice developers (among others) can not only reverse engineer undocumented binary Microsoft formats (to the point where my company’s customers use OpenOffice to, among other things, recover corrupted MS Office documents) but also implement ODF, DocBook, and various other open standard formats, then surely all the brilliant coders at Microsoft could do it before breakfast, right?   Oh hang on – Microsoft put out a brand new browser to stem the threat of Firefox, and despite their crack team of developers, they still can’t make it as W3C standards compliant as any other browser out there…  I guess I shouldn’t be too harsh – at least PNG transparency is supported now, something like 5 years after every other browser supported it…

    Blocker #2:  Andy, above, gets the gist across rather succinctly.  The known voting "irregularities" collated on <a href="">No OOXML</a> indicate that Microsoft has no standards for ethics, much less software.   The big question is:  was Microsoft more successful in its deceit elsewhere and managed to get away with it?

    Blocker #3:  Steve Ballmer, like it or not, represents your corporation.  He’s obviously an embarrassment to the corporation (Linux is a cancer!), but he also shapes public opinion on Microsoft.  If he says that Microsoft believes Linux and open source infringe on 250+ Microsoft patents, and then refuses to reveal which ones… well, I think I could introduce you to quite a few primary school children who could explain how immature that is.  How can an organisation run by a guy like that claim to be worthy of trust?

    Good luck, Brian.  


  34. David Lane says:

    To those following along at home – most (but not all?) of the web links I included in my last post were stripped or at least mis-formatted.  I’ll put a proper rendering of my comment up on my own site.  You can find it by clicking on my name above.


    On a slightly different tack – I was hoping you could give me some Microsoft "insider" insight into this little question.  Despite the amazing amount of <a href="">effort Microsoft went to in New Zealand</a> to sway public opinion (some might interpret it as desperation) in favour of OOXML (!!), when presented by the facts – largely by some incredibly well informed <a href="">NZ Open Source Society</a> members, Standards NZ still voted against giving it the "fast track" tick with quite a few objections (45?).  

    Helen Robinson, MD of Microsoft New Zealand, <a href=",helen,robinson">resigned</a&gt; the following week.  Coincidence?  Hmmm.  The question is: was she canned for having failed to "take sufficient measures" to ensure a victory, or did she suddenly find a conscience and leave due the persistent nausea she felt at being associated with the sort of behaviour Microsoft exhibited in the course of the voting process?

    By the way, Brian, as you get flustered, your spelling suffers.

  35. Sam Hiser says:


    What happened to your blog? It looks like absolute cat-vomit.

    It was so nice before.

    Missed you at XML 2007 in Boston this week.

  36. David Lane says:

    Apologies, Sam, if I had anything to do with the feline spew.  This comment form doesn’t indicate what sort of markup is valid…  having seen links in previous comments, I assumed that I could just include some… Apparently the comment parser isn’t good at dealing with HTML.



  37. says:


    You are very entertaining, and I thank you for that. It’s funny that you think I get flustered having a discussion burried in the comments section of my own blog. I suppose it would be possible to get me flustered, but I doubt it. I always have spelling mistakes in the comment section, as I can’t write those in Word (unlike the original blog posts).

    You have a strong religious zeal, and that is what makes the discussion fun in certain ways. It’s like arguing politics. You rarely make any progress, but it still is entertaining.

    I have no clue why Helen resigned. I’ve never met her.

    I don’t know how this brainwashing happened, but it’s obvious you’re convinced that ODF was some great thing, and Microsoft chose to go a different way. Sure, we are members of OASIS, just like IBM is a member of Ecma. ODF was created by Sun, and then it underwent some minor revisions in OASIS. The group working on it was very small, there wasn’t much interest from outside. It wasn’t until it was standardized that people started to pay attention (and by people I mean IBM). At the point it was standardized, there were only three people who had attended more than 75% of the meetings (and one of those guys, Gary Edwards, now seems to be under attack by folks like Rob Wier). It was by no means some great big community effort.

    Either way though, we (Microsoft) do indeed support ODF. We’ve helped fund an open source project that translates between Open XML and ODF and we continue to work on ways to better integrate that into our product.

    There is absolutely no way we could have used it as a default format though. It’s qualified to be an alternative format, but most certainly not the default.



    I would have loved to have attended the XML conference. I’m in Kyoto right now for Ecma meetings though. I talked to Vijay about it a bit though and it sounds like it went well.

    You don’t like the new look? I don’t remember why I changed it exactly. Some folks were complaining about the layout a bit. I think it looks ok.


  38. Elektro says:

    "There is absolutely no way we could have used it as a default format though. It’s qualified to be an alternative format, but most certainly not the default."

    So? Because you are unable to implement an international standard you want to declare your own format the international standard and screw up the international standardization process?

    Why is OOXML incompatible with ISO 26300 even in parts where it could be? Why doesnt your company embrace the French proposal (which seems very reasonable to me) to bridge the two format? Why cannot your company effort any changes to the format it submitted to ECMA.

    Who cares about YOUR lock-in in YOUR platform.

    The whole process stinks.

  39. says:


    We are indeed changing the format. It underwent changes during the Ecma process, and it will undergo more changes now as part of the ISO process. It isn’t our format any more.

    Open XML and ODF had different design goals. ODF was not designed with the legacy base of documents in mind, and for us that’s something that is needed by our customers. We already had a format that was designed with this goal in mind, and based on feedback from various governments we decided we shoudl take that work and submit it to a standards body for long term maintenence.


  40. David Lane says:

    Well defused, Brian.  Good work.  Actually, I’m intrigued by the fact that you continue to engage, despite the fact that only sycophants will find your reasoning… reasonable.  I was considering asking you about brainwashing earlier but decided that would be too patronising, even for me.  

    Religious zeal implies "blind faith", something which I don’t have.  I do, however, have confidence (based on evidence) that my observations of the world are every bit as relevant as those you have made.  I also believe the my perspective of Microsoft’s position in the world, being where I am (in the IT "trenches" outside of the US of A), is quite different from – and no less valid then – yours.

    It’s fairly clear to me that, though you’re probably a pretty nice guy, you’re too closely aligned to your paycheck to recognise how Microsoft’s actions ripple through the IT world like a bull in a china shop (with an equally oblivious body awareness, I might add).  

    As I did during the NZ-based OOXML debate I had with your recently emigrated Redmond-based colleague (another casualty of the NZ debacle?) Sean McBreen in the days leading up to SNZ’s unfavourable (to Microsoft) ruling, I’ll take this opportunity to refer to Upton Sinclair’s sage words: "It’s difficult to get a man to understand something if his salary depends on him not understanding it."

    With OOXML, Microsoft is try to prove to itself that it’s influential.  Don’t worry, it is.  But for all the wrong reasons.  Eventually, it’ll all come out in the wash, and sadly, despite the bulk of Microsoft’s empire, it won’t be pretty…

    Just remember, Microsofties – non-Microsoft IT professionals out in the world out number you substantially.  It would behoove the team there in Redmond to refresh their understanding of the concept of "goodwill".

    Again, best of luck – I hope, when the dust settles, that you’re still standing.  If so, pop down to NZ, and I’ll buy you a beer just to show there’re no hard feelings.



  41. says:


    I was hoping to get a trip down there at some point, and if so I’ll definitely look you up.

    Why don’t we take a more pragmatic approach for now.

    Based on where we are now (we can’t go back in time), what specifically do you want to see changed in the spec?

    (I know you don’t believe this, but please assume for now that Open XML and ODF were designed in parallel with different design goals. Using ODF as the default format would not work for Office, so we do need another format.)

    Based on that assumption, what do you want to see us do differently with the format we currently have?


  42. David Lane says:

    Wow, that’s a very magnanimous gesture, Brian – and I must say, I wasn’t expecting it.  Thanks, don’t mind if I do suggest some changes to Microsoft’s approach:


    1.  Support ODF in all Microsoft products as an integrated file format, on par with legacy binary MS Office formats – not a lame aftermarket plug-in.  Users should be able to select "always save in ODF format rather than OOXML".  By all means offer OOXML, too, if you want.

    2.  Stop trying to push for OOXML ISO standardisation.  You don’t need to – by implementing it in MS Office, it’s going to have widespread use anyway by (and effectively lock-in) those who don’t know any better (i.e. those who use Outlook Express, for example – sorry, couldn’t resist :)) and if your software fully and seamlessly supports ODF as an alternative (non-native) format, you wouldn’t need OOXML standardised to bid for those meaty contracts in places like Massachusetts, Paris, Vienna, etc.

    3.  I want to see Microsoft *actually* cooperate with the global IT community and do something *right*, rather than something designed to give Microsoft’s monopoly further anti-competitive leverage; for Microsoft to actually *be* trustworthy, rather than paying hordes of PR staff to make it *appear* trustworthy; to use words like "open", "genuine", "myth", "shared", and "trustworthy computing" in the way that most people think of them in the context of IT rather than as ‘SoftSpeak[TM] – where black is white and white black.  

    4.  I’d love to see Microsoft lobby the US government to repeal the DMCA, discontinue (and invalidate existing) Software Patents, and perhaps, while they’re at it, scrap the "Patriot" Act.

    5.  I want Microsoft to accept the fact that monopolies are ultimately bad for everyone.  In the worst case, monopoly of freedom leads to painful revolution, in which the monopolists tend to suffer.  In fact, I believe that the US Constitution makes a few references to that possibility…

    6.  To gain goodwill, Microsoft can make the following gestures:

    – gift all of its software patents (and I encourage IBM and other software patent holders to do this, too) to, regardless of the eventual fate of Software Patents,

    – publish, under the Creative Commons license of its choice, all relevant documentation for all MS  Office/Exchange/AD/SharePoint-related file formats, data structures, and communication protocols to allow anyone to build truly interoperable software without resorting to reverse engineering,

    – release the Linux versions of all the MS applications that are no doubt running in MS Research facilities as I type – fill in the others as time allows.  I reckon many people would even be willing to pay some sort of fee for using them.  I’ll leave it to you guys to figure out the details…

    – stop referring to "Shared Source" as if it was somehow equivalent to actual open source software.

    – make "Microsoft Genuine Advantage"… advantageous to someone besides Microsoft

    – bloody well ditch the Trident rendering engine for IE and adopt a working, reasonably standards aware one like Webkit or Gecko.  The GPL lets you do it, you know…  I mean, you’re giving IE away anyhow.

    f) take the Microsoft PR and Legal departments out back, give ’em a beer, a slap on the back, and then sack the lot of them.  Permanently.

    Microsoft – rather than trying to retain its ability to manipulate the market by force – could actually compete by producing *better technology*…  the profit margins might go down in the short term (with 90%+ profit on each copy of MS Office, I’d think you’d have a bit of margin to spare, eh), but I’m confident that Microsoft could actually do much better in the long run, and have a more sustainable business, if it had a bit more focus.  And didn’t have to spend so much time engineering software to break if it wasn’t used in accordance to its arcane licensing terms…  After all, at the current rate of "innovation" it’s that self-protecting overzealous complexity that will eventually make Microsoft a non-starter in the marketplace.

    The suggestions above would be a start down the road to reformation for Microsoft – in my humble opinion.  Over time, people might allow themselves to be convinced that you weren’t trying to lock them in, to fleece them, to "upsell" them or simply get them over a barrel to maximise per-customer revenue.  And – get this – you wouldn’t be.  But that would be a perfect world.  


    Thanks for that, I feel cleansed.


  43. says:

    OK, that’s a bit of a broader list than I was expecting, but let me give it a shot (note my question was specifically around the spec, not Microsoft as a whole):

    1) I think there probably are some improvements that could be made in the UI, but it would require some significant changes to Office. Currently the converter infrastructure doesn’t go through Open XML, and that would need to be updated. So I think this is a reasonable request (not saying it will happen, but it’s a reasonable request). Note that there are already a few options for folks that want to use ODF, and the experience is close (but not quite) what you describe.

    2) We submitted to ISO based on feedback from governments. Believe me, it would have made our jobs at Microsoft easier if we hadn’t gone through standardization, as it means we no longer have control over our file format. We were told though that we needed to standardize our format to ensure that everyone else could use it, and we couldn’t change it. So I don’t really think taking the format away from ISO is even an option for us. If it turns out that ISO doesn’t want to maintain the format, I guess that’s a different story. I’m hoping they will take it (as it’s better for the communities confidence in the format).

    3. I think we actually do compete on innovation. Sorry if you don’t.

    4. That’s not really my call Dave (although I agree the Patriot act is pretty scary).

    5. I don’t think Microsoft in any way wants to see a Monopoly on freedom (I think bush already lays claim to that 🙂 )

    6. As I said, Balmer is a number of levels above me, and all these things are really more at his level, not mine. (I love how you assume we have versions of all the products running on Linux in our research departments… you may be giving us too much credit there)

    7 (f). Again, that’s not my call (although I have a number of friends in those groups who I have a lot of respect for, so I think I may share a beer with them, but pass on the sacking).

    Dave, I would like to know seriously though what your complaints with the Open XML spec are. Why are you so opposed to it (technically). I can’t change your opinion of Microsoft, but I can look at proposing changes to the format via TC45 if there are things you don’t like.


  44. juan says:

    brian, you really are not ashamed of having thrown to ISO fast-tracking this +6000 page "specification" ?

    you really think that any Microsoft-independent developer can implement 98% of it in a usable timeframe (say, in less than 25 years? )

    if your answer to this questions is : yes

    i think it’s useless to try to argue with you … and i will say the same words that dave said:

    "good luck brian"


  45. BigAl says:

    Brian, this is what I want Microsoft to do:

    1. Stop pushing the "legacy document" argument. You know as well as I do do that that is simply a red herring to confuse the PHB’s. A NEW format standard is… um… NEW. "Legacy" means OLD. MS Office can already handle OLD formats. What we want is NEW.

    2. Stop pushing OOXML to become an ISO standard. Work with everybody else to make ODF better. By all means use OOXML in MS Office — you already have a monopoly on the desktop, so OOXML will become the de-facto standard in any case.

    3. Provide ODF capabilities in MS Office. That way governments around the world can tick the "ISO Compliant" box and you can keep on milking that cow. (And it will be VERY easy for you to do so; don’t even try to argue that it is not technically possible.)

    4. On a more general note: a culture change at Microsoft is needed. Please stop the patronizing attitude of "Microsoft makes a gazillion dollars a day, therefore we know everything". You don’t. Instead of spending how much time and money on silly campaigns (like this one) why does Microsoft not rather spend that time and money improving their products? During fiscal 2005 Microsoft spent or earmarked nearly 25% ($8 billion) of their income ($36 billion) for litigation. What?! Does that not tell you people that there is something seriously wrong here? And then you cannot understand why people are suspicious of Microsoft?

  46. Andrew says:


    1. OOXML is not a legacy format.  In addition, customers do not want two versions of Office.  Implementers do not want two specs (that look very similar).  

    2. If OOXML is not an ISO standard the alternative is worse.  Multiple interoperable standards = choice.

    3. Possible, but ODF’s feature set is not comparable (ie is limited).  See point 2.

    4. MS has a big R&D spend.  Is not Office 2007 a good example of an improved product?  Open Office looks shambolic in comparison.

  47. David Lane says:

    Hi again Brian,

    Apologies for the delayed response!  I’ve been away for the weekend in the Alps north of Christchurch…   Gorgeous summer weather. 🙂

    Thanks again for giving me the opportunity to indulge my fantasies about what – broadly speaking – Microsoft could do to earn my respect.  Yes, I know your question was specifically about the spec, but I am allowed to dream, aren’t I?   I’m just sorry that you didn’t comment on some of the more realistic options I threw on the table.  

    As for your responses…  hmmm, they don’t seem to have that fresh crinkle of veracity about them…

    1.  I could be mistaken, but I think that you’ve got a precedent for natively supported non-native file formats in MS Word.  If the list of MS file formats OpenOffice im/exports is any indication (I can’t check a recent version of MS Word haven’t used MS Office since I wrote my thesis in 4.3 back when I lived and worked across the lake from Redmond – and based on the miserable experience swore never to use a Microsoft product again – around 1994) then you’re already supporting at least a dozen outdated (formerly native) file formats and non-native integrated (from the UI perspective) file formats.  

    From memory, I recall that Plain text, Rich Text (RTF?), HTML, and a few (now obsolete) MS Office legacy binary file formats are supported for both read and write, no?  What makes ODF so different?  

    Aren’t there libraries in MS Office’s well structured object oriented framework, with clear functional boundaries, to make adding this sort of thing fairly trivial?   Surely all those billions of R&D dollars that have gone into MS Office would have resulted in a clean, well documented code base, right?  Nobody’d spend all that money just on hiring a huge army of people to test all the features in an office suite by brute force trial and error?  And surely it didn’t take all that R&D funding to adopt the "Ribbon"?  

    If that’s not the case, you might want to consider quietly borrowing  the file parsing/writing code from OpenOffice or KOffice because those teams seem to be able to add pretty comprehensive support for new formats (e.g. ODF) in one or two open source development cycles…  (nb. from experience, the ratio is 3 to 10 open source development cycles : one Microsoft development cycle)

    2.  I’m very skeptical that any government said "Please provide us with an ISO standard file format for Microsoft Office – that’s got both the words ‘Office’ and ‘Open’ in its name –  and which is

    – completely new and unproven

    – requires us to purchase another wave of software upgrades from Microsoft (who has a 60+% profit margin, assuming substantial ‘discounts’),

    – which is completely incompatible to the already ISO approved ODF file format (which is what we specifically requested in public), and

    – which no one, not even Microsoft, has completely implemented.

    Oh, and make sure that through your arrangement with the ECMA, you retain the right to alter the spec unilaterally in future, thus potentially breaking any other implementations."  Isn’t that what this:


    Frankly, Brian, I appreciate your willingness to engage in heady banter, but please I don’t insult my intelligence.

    I suspect what governments around the world (I’ve listed a couple examples in previous comments – more are listed elsewhere, e.g. on, although you might find it hard to find them since searching on Open Document Format and Open Office seems to return a lot of hits for Office Open XML Document Format…  Coincidence?  I think not.  Namespace pollution?  I think so.  Mildly clever, but in drunk frat-boy sort of way) said something like:

    "There’s something we didn’t consider when we first started using Microsoft Office – every document we create digs us deeper into the proprietary format-induced hole some refer to as the impending ‘Digital Dark Ages.’  We have been entrusted with the responsibility of retaining – in an accessible form – the peoples’ information for posterity.  Eventually, someone’s going to figure out that you guys have us totally over a barrel, and we’re going to get voted out of office (or at least have to scramble to find some scapegoats).  Please, oh please stop this madness, and make your software conform to open document standards so that we can at least *start* to turn this dire situation around."  

    What I’m saying is that they didn’t tell you make OOXML – I wouldn’t believe that for a second.  They told you to solve a serious problem created – let’s face it – by your corporate employer’s past business practices.  

    Let’s be clear about this – Microsoft chose to use OOXML rather than ODF because it was seen as a way to retain its monopoly hold on the "office franchise".  If OOXML got a few testimonials  of support from a few government officials after some fancy dinners and a game or two of golf, I wouldn’t be too surprised.

    3.  Don’t get me wrong on this one!  I don’t disagree that Microsoft innovates!  Not at all – on the contrary:  Microsoft certainly innovates in

    – PR where, for example, Vista is a huge success "beating all previous sales records", where Zune is "tops in its category" on Amazon (what else is in that category?),

    – legal hackery (e.g. brilliantly liability-free End User License Agreements)

    – and "unofficial" influence (e.g. almost undetectable influence of government agencies in Nigeria, the US, Norway, Switzerland, and just about every other country, and only occasionally getting caught…).

    I don’t believe that either of Microsoft’s only 2 real profit centres demonstrate significant *technical* innovation.   Vista is an unmitigated disaster – even Microsoft is starting to hint in the press that this is the case, in hopes of cushioning the stock market reaction when it becomes undeniable.  

    MS Office 2007 is probably the best nag in the Microsoft stable, but I gotta say the "Ribbon Interface" which is the only new feature I’ve really seen touted, isn’t all that exciting.  Nor is it necessarily all that innovative.  See, for example: (see section on Prior Art) (see "Controversy" section)

    Wouldn’t it bunch Microsoft’s knickers if, rather than simply copying Apple and Xerox, the major innovation in their flagship product was in fact copied from a open source applications developed by rag-tag bunch of uncouth, self-funded, idealistic "computer enthusiasts"?

    4.  Come on, Brian, I’m sure as a thinker and principled person you must have a view on software patents!  Seems the haloed bossman emeritus had a position on the issue:

    "If people had understood how patents would be granted when most of today’s ideas were invented, and had taken out patents, the industry would be at a complete standstill today."  William Gates III, 1991

    5.   Microsoft does, however, want to maintain a monopoly on the desktop, and leverage that to achieve a monopoly of the earbud, server room, cellphone, and everything in between, no?   You know what something that grows without regard to its environment is normally called?  Yep, a cancerous tumor (unless, of course, you’re asking Steve Ballmer, in which case it’s Linux).  

    6.   Don’t go all coy and bashful on me now, Brian.  I’d be somewhat embarrassed for Microsoft if they weren’t hedging their bets.  Surely they can see the writing on the wall.  Vista’s a dud, Leopard is clawin’ their asses, and Linux (with no handy sales figures to provide real evidence of uptake) with gPC, OLPC, Ubuntu, etc. are attacking on every other front…  Needless to say, if MS Office can run on OS X, then it’s not a huge stretch to run it on a Linux desktop.  

    Quick story: I once asked a PR manager at the Jade Corporation (a Microsoft partner, I believe, based here in Christchurch, NZ) when they would have a Linux version of their software system available, and he said "never – there’s been absolutely no demand from our customers whatsoever" with a smug smile.  Two weeks later, they bombarded the media with the release of Jade for Linux, version 1.0…  That was 5 years ago.  And still nobody’s heard of them…

    Oops – apologies about the "f)" – it was just supposed to be another bullet point.  

    Right – you requested my specific complaints about OOXML, so here they are.

    I, personally,  have no specific *technical* complaint about OOXML.  I leave it to my colleagues in the NZ Open Source Society like Matthew Cruickshank (see to see why he has an interest) to read and digest 6000 specs.  

    I do, however have three fundamental problems with OOXML:

    1. the fact that you have claimed that its purpose is to somehow (and I’m not sure I understand how OOXML can possibly do this) support "legacy binary file formats" without explicitly including the full specification for those binary formats, too.  I think that’s a total misdirection (as does BigAl for one – any others?) and I resent it.

    2. the fact that Microsoft trying to make OOXML into an ISO standard.  

    3. the fact that Microsoft has actively tried to game (some might choose a more accurate and less charitable term) the ISO system.  

    ODF is *the* ISO standard for office documents.  OOXML is not, nor should it be.  Microsoft wants to use OOXML, and that’s great, go for it.  It doesn’t concern me as I’m not ever likely to run MS Office (particularly since it seems it’ll never run on Linux, and you won’t see me with Vista… ever).  But two different standards for the same thing is a senseless waste of resources.  ODF and OOXML server exactly the same functional domain.   Microsoft didn’t say "no can do" when it came to supporting Word Perfect’s file format back in the days when there was a competitor, did it?  Of course, now Microsoft more or less owns Corel, no?

    @Andrew – Sorry, I’m afraid you’re a bit mixed up.  In reference to your point #2:  two standards for the same thing != choice.  Two (or more) different *implementations* of the *same standard* = choice.  

    You’re confusing choice with "monopoly protection".  

    Two standards for the same thing means lost Mars Climate Orbiters due to one team inadvertently sending telemetry data in imperial units rather than the metric units the telemetry software was designed for…   That’s what we call a "humiliating disaster".

    Also, can you please be a bit more specific about the "improvements" in Microsoft Office 2007 over what’s available in 2.3?   I’d like to know, because perhaps – if I agree with your meaning of "shambolic" (I have my own – see my first comment above) –  I might be able to do something to help fix the shortcomings of OOo.  

    All the best to both of you,


  48. Jose_X says:

    On the Microsoft Promise:

    The pledge on this page is funny. It basically says Microsoft won’t sue for A, B, C, D, and E. It doesn’t say Microsoft won’t sue for V, W, X, Y, and Z. Since Ballmer and Microsoft terms elsewhere have made it clear that there are things for which Microsoft will sue, as (hypothetical) manager of MS Office, I would put all the lock-in secret sauces in V, W, X, Y, and Z and dare any open source implementation to try and provide a compatible product after a long period of reverse-engineering.

    Word is from some that have been experimenting that already current saved files in MSO2007 are taking advantage of secret sauces not spelled out in the OOXML candidate spec when producing new content. This not only shows the worthlessness of "open" OOXML in the hands of a company that wishes to hide their secret sauces, but it shows Microsoft representatives are misleading the public about their product whenever they state that the closed portions of OOXML are for legacy purposes.


    A note on OOXML and where it might thrive:

    Unfortunately, this thread died because of inattention on my part. Still the question posed was not answered by the other poster.

    Hearing others’ description of it, I would say that OOXML is a much better format than ODF to be used by a group with controlling market share to keep all attempts at compatible products at bay. These same qualities make the format brittle *even should* competing implementations *want* to interoperate.

    OOXML supposedly stores a file as a set of "assembly" instructions much more than does ODF. If there is any error in assembling the end product, everything could come out completely wrong. With ODF, you always have a general idea of what the final file is supposed to look like and contain so that if a particular office suite wanted to throw the others off it would have a difficult time since it would basically have to produce the end product (unless it was a really twisted implementation). With OOXML, any of the instructions along the way could say, "scratch everything done so far if a particular part of the file is not 1232312312 and doesn’t say ‘made by Microsoft’ ". In this way Microsoft office suites could much more easily hide the true contents of so-called "open" OOXML files as contrasted with ODF files. As Microsoft Office manager I would definitely prefer OOXML.


    On the worth of Microsoft unofficial marketing promises:

    The short of my gripe is that Microsoft is not going to implement any real openness in their products. Their stockholders would eat them up if they did. They are looking for marketing material (hence ISO approval) for bait and switch sales. Marketing will talk about an open format but sell the customer a product that will work differently in practice. [The story of the late Ed Curry gives a glimpse of Microsoft bait and switch. The many court lawsuits and criminal proceedings against Microsoft provide even more material of Microsoft execs and contracts saying one thing but the company doing another. So sue me, right?]


    On the importance of doing business with companies that have a record of trustworthiness and avoiding the others:

    I think that companies doing business with another company need to consider the trust issue. Well-defined computer specs are fairly precise. English is not. Lawyers make good money arguing any and both sides of an argument. Contracts are only as good as the lawyers writing up your side of it and willing to defend it in court and your willingness to pour funds into a lawsuit against a wealthier foe. Oh, and let’s not forget that Microsoft actually writes contracts where the customer effectively agrees not to hold Microsoft liable for virtually anything.


    Why going open source makes sense:

    Again, with open source, there is nothing to argue over because all the relevant information is openly available and computer languages are much more precise than the language used to write contracts. Why even try to determine trust and consider legal disputes or ambiguously worded contract terms when you can just look at the internal source code of the product you are being sold? Open source is the tech industry’s embodiment of the classic show me don’t tell me.

    The only way to ensure that you (as a customer) get what you think you are getting is to buy an open source product. The currently most advanced open source "office suite" uses ODF. Any open source product (including libraries naturally) would come with the source code which would define the file format *actually* implemented with the product instead of the file format *claimed* implemented as splashed in color on the outside of the box.


    Although I personally want Microsoft to remain on course trying to push OOXML through ISO, I feel I have to offer what might actually be useful advice to Microsoft given that Brian Jones bravely has allowed on his blog all sorts of openly critical comments against the company employing him. I agree very strongly with Dave Lane that Microsoft should back away from trying to make OOXML an ISO standard. Microsoft is playing the name game without regard to damage inflicted on ISO and others. This will not help Microsoft’s goals of earning trust from many in the community or who have in the past bought Microsoft products. We are not in the 1990s. There are real alternatives now and more people than ever have had or will have the opportunity to look over Microsoft’s dirty laundry.

  49. JimStaunton says:

    How can you claim that "It isn’t our format any more" when the proposed standard essentially states "do X like Office 2007 does X", without bothering to specify Office’s behaviour.

    That’s still a MSFT format, not a standard.

  50. Tim G says:

    Brian Jones said:

    "If we used ODF and didn’t create Open XML, 99% of our customers who don’t give a lick about file formats would have to suffer. They would lose things from their files, and they would then lose faith in the product. They would go back to using the binary formats, and that would do absolutely no good."

    But the MSOOXML spec. INCLUDES the binary formats like a sausage wrapper without listing the ingredients in the sausage.  What’s in the sausage?  Beef, Pork, Venison?  Tripe?  We dunno.

    So what I will run into as a developer trying to implement the MSOOXML standard will run into is a point where I the standard says "display this (legacy) office format properly" with no definition or explanation of how to display that format properly.

    Bah!  The only way to accurately claim that this standard is open is to say:  "The MSOOXML standard is completely open, except for the areas where it isn’t." and the ways in which it isn’t are significant.

    If MS wants an open standard, MS is going to have to really make one.  Right now MS is just trying to game a system in order to CLAIM that it supports on open format for PR/marketing use for those who lack the clue to cry "foul" and throw the BS flag.


  51. says:


    Are you really a developer? Do you work on office productivity software?

    Do you really think the handful of compatibility settings that your refering to ("layoutFootnotesLikeSomeOldVersion") are the secret sauce to the formats?

    In Office, we try our damndest to get people to turn those settings off. They basically just tell Word to use an old buggy layout behavior that was fixed. We thought it would be better to at least put them in the format so folks who understood that behavior would know what flag to use. OpenOffice has the same type of settings, they just kept them all (even the more useful ones) out of the spec completely (which I don’t think was the right decision from an interoperability point of view).

    So, do you think it is still important to provide the documentation? Or would it be better to deprecate it? Or do you want to see both things happen?


  52. H.Kwint says:

    Hi Brian,

    After reading all the comments and your answers, I’m left with three questions:

    -1) You indicate Microsofts concerns with billions of legacy documents was one of the reasons for creating OOXML, and not being able to use ODF. If Microsoft is concerned with all those legacy documents and people not being able to open them in the future, why doesn’t it open all old .doc , .xls etc. formats? That would be the best assurance those people will not lose their old formats. I think opening up those old formats would be a better assurance and less work than the efforts to partially cover these old formats (because that’s what compatibility mean I assume?) in Microsoft’s OOXML scheme. I’m sure a whole developer community would be created around legacy Microsoft Office file formats if Microsoft were to release them under an open permissive license, lots of ‘conversion’ applications would be created while not even costing Microsoft a single dollar. Microsoft customers would never have to worry about their legacy-format documents again, since there would be a whole ecosystem to help them which whatever they wanted to do with their legacy documents. That would be true backwards compatibility. I can’t see how OOXML is ‘compatible’ with those old legacy documents, except for vaguely describing in OOXML terms what is in binary form in those formats.

    -You said above:

    2. If OOXML is not an ISO standard the alternative is worse.  Multiple interoperable standards = choice.

    You sound like your opinion is choice is a good thing. I think you are confusing with applications, when it comes to applications choice is a good thing. When it comes to standards however, choice is a bad thing.

    Though I’m interested in both open source software and open standards, I work as a technical drawer though, and I can tell you, when it comes to standards, you don’t want to have choice when choosing a standard. Multiple interoperable standards cost and have costed business billions, and those costs could have been prevented if multiple interoperable standards didn’t exist.

    Let’s try to explain that. The metrical and imperial system are interoperable and indeed provide a choice. Nonetheless, confusion between them during the design-phase caused a NASA-Mars-rover to fail, which costed them about 1 billion dollar. This wouldn’t have been necessary if only 1 standard would have existed, and if choice was absent.

    At some chemical plants, DIN-flange-dimensions and ASME-tube dimensions are used, mainly because of ‘legacy’ reasons and the wish to keep compatibility. While most of the time the DIN- and ASME standard are interoperable and ‘provide a choice’, business could have saved billions if there was only one choice, and choice between DIN- and ASME was absent. Now, must time is wasted by deciding what ASME tubes (inches) fit on what DIN flanges (millimetres), and this is an error-prone process. Also, a lot of manufacturers both have to sell DIN- and ASME dimensioned tubes and flanges. Business could have saved a lot of money if choice of standards was absent.

    To give a software example: there’s plenty of choice when it comes to video and audio standards.

    For example, look at:

    However, it costs the Windows user – a Microsoft customer – lots of dollars of license fees for all those different ‘partially interoperable’ standards, and Windows users need zillions of programs to be able to use all these ‘interoperable staandards’. In fact, all these ‘interoperable standards’ is the worst nightmare of every Linux-user, since most of those interoperable standards are not supported on Linux.

    If there were only a few standards life would be easier, and less time would be wasted on supporting all those different standards.

    So, choice of different standards is a bad thing for business and consumers and costs them billions of wasted dollars, like proven in the engineering field before. That’s why more and more companies are switching to ISO/EN standards, instead of having to support British, American and German, Swedish and French standards. While they had the choice between all those different ‘interoperable’ standards, they would rather have one single standard, since that’s more convenient, less error prone and less expensive.

    Nonetheless, Microsoft is trying to push for an ISO-standard in an area in which an ISO standard exists. While I can understand ODF and OOXML are different, the goals of the accompanying ISO standards are not, they try to be a ‘document standard for office applications’. Though Microsoft might not be able / willing to change ODF via OASIS (to me it seems like the latter is the case, but who am I?), there must be possibilities to try to change (append to) ISO26300. In the case where OOXML becomes ISO-standardized, it will partially overlap the existing ISO 263000 standard, because both describe a document standard for office applications.

    It will again cost Microsoft’s customers billions of dollars. Why is Microsoft trying to make life more expensive for its customers? Billions could be saved if there was only one single ISO standard of office applications.

    Now, here’s the question: Why does’t Microsoft try to save its clients money, instead of making sure their clients waste more money like which is the result of trying to get OOXML standardized at ISO?

    For your spelling in the comment-box problem, I can really recommend Mozilla Firefox to you, it includes spelling checks in comment boxes (even in different languages), and it has prevented me from several spelling errors (though you might see it didn’t catch all spelling errors of course) in this comment already.

    3.) Instead of you and Rob Weir (and I for that part) arguing with each other, couldn’t the two of you better spend your time by sitting around the table and discussing how these two file formats – ODF and OOXML – can be merged into ‘one fits all’ standard? It would certainly be better for Microsofts customers; their money would be better spent if you were trying to save them money instead of being paid by Microsoft and explaining them why Microsoft standard policy is costing them more money.