Ecma meetings in Toronto

Well we’re in the middle of our last face to face meeting for Ecma TC45 before the ISO vote on September 2. Novell is hosting it up in Toronto, and it’s actually given us a chance to meet some of the developers who are integrating Open XML into Open Office. They had initially worked with the XSLTs coming out of the ODF to Open XML translator project, but now it looks like they are building the support in at a more native level without the XSLTs. It was a really cool demo. They showed importing a formatted spreadsheet .xlsx file into Calc; a graphic heavy presentation .pptx file into Impress; and a rich wordprocessing .docx into Writer. There is still a ways to go, but there are really making impressive progress.

The main thing we’re currently working through is the list of comments that Ecma will submit to ISO as part of the fast trac process. Novell has some they’ve come across with their latest development effort, and there are also a few that folks presented based on public statements they’ve seen. Not sure how big the list will be, but there are some pretty good improvements in there that will be dealt with along with all the other national body comments at the ballot resolution meeting (BRM).

It should be a pretty exciting week or two. September 2nd gives us a great checkpoint to see where we are in forming a consensus for approval of DIS 29500 (Open XML). Basically we see who is already in favor of the standard and what fixes they’d like to see before it’s finalized, but we’ll also see folks who are still hesitant, and what fixes will make them change their mind. I think we’ll already be pretty close to the required number of yes votes, meaning that going into the ballot resolution meeting there will be a lot of momentum to get the issues fixed and move towards a final approval of the standard.

There were a couple interesting links from yesterday that I wanted to point out:

  • ODF / Open XML – Technical Specifications Mature Over Time – Jason Matusow takes a look at what it means for a standard to be mature, and how that should impact its approval. Clearly all standards continue to evolve and improve, so that should really be taken into account. This is why at Ecma there is now a lot of focus on the future maintenance of the spec.
  • Whitepaper on document format adoption – A report from IDC which drills into the current thinking in the industry around Open XML and ODF.
  • Last days for the Open XML ballot – Rick Jelliffe talks more about the closure of this stage of the fast track ballot. I’d be interested to hear what other people think of his suggestion to remove much of the informative text from the spec. I think it’s useful for implementers, but Rick has more experience in standards than I do.

I hope everyone is having a great week.

-Brian Quote of the Day:

CEGID – France

“With Open XML formats, office documents easily integrate into enterprise solutions, and management data is found within the files themselves. This format lets us combine structured and nonstructured documents and use office applications to pull information straight from ERP. As long as the content remains open and accessible, the Office Open XML format makes integration and compatibility between Office documents and enterprise applications easier. It is part of our long-term vision, because of its robustness and accessibility”

– Christophe Raymond – CTO

Comments (74)

  1. John says:


    Have you seen this?

    "The vote on OOXML looked fairly secured. Most in the Working Group in Sweden was against the vote to approve OOXML. The day of the vote, though, more companies showed up at the door. Some 20 new companies — each one payed about $2500 to be allowed to vote — and vote they did … for Microsoft. Most of the new companies were partners from Microsoft who suddenly out of the blue joined the Working Group, payed membership fees and voted yes for approval. From the OS2World story: ‘The final result was 25 Yes, 6 No and 3 Abs and this would from the start be a done deal of saying No! Jonas Bosson who participated in today’s meeting on behalf on FFII said that he left the meeting in protest and so did also IBM’s Swedish local representative Johan Westman.’"

    How can you guys justify this level of corruption?  It’s truely awful 🙁

  2. says:

    I think that is pure FUD. I don’t believe that story for a moment. First off the fact that they article says "the vote was fairly secured" as if blocking a standard is a good thing that should be secured is pretty telling of the source of this information.

    It’s funny that all these allegations come out, yet it was IBM who actually did the unethical things like writing one countries submission during the contradictory period, and actually changing up countries processes at the last minute. Noone talks about that though, instead they jump on these bogus stories about what Microsoft is doing.


  3. It seems that every country has different rules.   In Mexico it turns out that anyone that registers on a web page and emails their vote can vote.   This was presented as the "Democratic thing to do".

    I attended the meeting last week, and I do not think a single person in that meeting had actually read the spec or spent time with it, but people did repeat the same arguments that have surfaced everywhere (it was basically a repetition of Groklaw’s page and selected Rob Weir points) and a lot of statements that were just not true ("ECMA fees are half a million dollars to join" was my favorite).

    The Mexican vote at this point is going to be as educated as the result from the RAND() function, it will be bound by how many people registered and voted by email and not by any technical discussion.


  4. John says:


     You are just dismissing it as FUD?

     Which part are you denying?  That 20 new companies joined up on the day, all of which were Microsoft partners, and all of which voted for Microsoft?


  5. John says:


     I doubt that many people at all have read the 6000 page spec.  You seem to be dismissing the arguments against OOXML just because other people have made them before.

  6. Brian heeft vandaag weer een post over OpenXML met daarin een aantal interessante links: ODF / Open…

  7. Bruno says:

    John, regarding the Swedish vote, according to a slashdot comment regarding it, there is no problem because all members that voted were already members of SIS.

    "Any Swedish company can become a member of SIS buy paying somewhere around $300-$500 per year. To be allowed to vote in this particular issue an extra 15 000 Sek ($2500) was needed. So yeah, it is open for anyone with cash (but they had to be members of SIS since before."

    I guess IBM’s allies would prefer that members of SIS be denied the right to vote on matters that are before their own organization unless they vote IBM’s way.

    What I was in college, I was a member of a student union, but only attended meetings when matters were being voted on (e.g. electing office holders).  Although I didn’t attend the other meetings, I had every right to vote, and it’s the same thing with SIS.

    These companies were already members of SIS, they paid the required fee to vote on a particular matter, and did so.  The article presents zero evidence of any malfeasance.  That it implies that malfeasance occurred without any evidence shows that the article is FUD, by definition.

    BTW, the final vote was 25 Yes, 6 No and 3 Abs.  Supposedly the "suddenly out of the blue" companies numbered "20".  According to another slashdot post, one of the 20 new companies was Google, which voted NO.  Now, assuming that the other 19 of the "out of the blue" companies voted yes (which might not be the case), then if you take away the votes of the 20 "out of the blue" companies, the tally is 6 YES, 5 No, and 3 Abs.  So it’s not like there was some overwhelming mandate to reject OOXML to begin with.

  8. John says:


     4 companies, including IBM, left in disgust without voting.  It was most certainly originally overwhelmingly against OOXML.

  9. Me says:


    Since when do companies have "feelings"?   They are just godless profit making machines, when did they become sisters of charity?

  10. Joe Curious says:

    Follow-up here:

    Now it’s SIS system that is broken. Given that you have to be a member of SIS to vote, I don’t see how – after all $2500 is token money in case like this.

    Anyway, does anyone have a list of company names and votes in English? I cannot find one.

  11. Curious Joe says:

    By the way, just what is "Farance, Incorporated" entity on US vote? I cannot google up *anything* about it that is not related to voting "no" on OOXML standartization. And Doug doesn’t seem to know this either.

  12. Bert says:

    Brian: Blocking comments as in Sweden is evil and you know that it happened. OOXML is as bad as that.

  13. says:

    Bert, take a reality check. Look at ODF and the control Sun and IBM are taking over that. They don’t even let folks talk about interoperability in the OASIS meetings. They booted folks like Gary Edwards who was one of the few original participants still left on the committee. They also ignored suggested improvements by the Novell folks because it didn’t fit with their goals.

    In Sweden, the IBM side started bringing in people who hadn’t even read the spec and just wanted to vote "no". Google joined at the last minute with the sole plan of blocking it. We asked a number of the companies we knew who want to see Open XML adopted if they would also participate and show their support.

    Another thing you’ll probably notice is that from what I’ve seen so far, the same comments are coming in from every country, where in many cases 90 plus percent were from IBM. They are attempting a denial of service attack on the various national bodies in the hopes to prevent them from reaching any sort of consensus on approval.

    Once the comments are all in, Ecma will sort through them and start working on proposed resolutions. I’m hopeful that many will be duplicates, so the numbers won’t be quite as high as they will initially appear to be.


  14. Rob Weir says:

    Brian, I’m afraid your "reality check" has bounced for lack of sufficient facts.

    Let’s review.

    You say that OASIS won’t allow talk about interoperability?  Funny, we were just talking about it on today’s TC call. Also, note that the ODF Adoption TC is hosting an "ODF Interoperability Camp" in Barcelona in around 3 weeks, at the conference, with 8 or so ODF vendors planning on attending.  I know that Microsoft knows about this conference.  Your Stephen McGibbon is registered to attend.

    Gary Edwards booted out?  That’s also news to me.  You can see the membership roster right here:  Gary remains a dues-paying representative of the Open Document Foundation.

    We ignored suggestions from Novell?  I think if you check the minutes, you’ll find that we debated Florian’s proposal for a month, with a lot of list traffic and meeting time spent on this.  His proposal was voted down 2-7 in favor of an alternative list enhancement proposal proposed by KOffice.  We discuss, we let everyone talk and then we vote. Is there something wrong with this?  Does Ecma do it differently?  I honestly don’t know, since their meeting minutes and discussion list archives are private.

    Comments on OOXML constitutes a "denial of service attack" on ISO?  Give me a break.  The question should, be why Ecma’s TC45 did not find these errors during their review?  Submitting a 6,000 page specification full of defects under Fast Track pretenses — that is the Denial of Service.  

  15. dmahugh says:

    Rob, why did IBM not find any of these errors during the review process?  And how do you explain that so many countries have the exact same comments that you submitted to V1, word-for-word identical?  Did they all copy you, or did you copy them?

  16. says:


    You might want to talk to Gary. Everything he’s been blogging about lately is actually pretty much the opposite of what you’re saying.

    Why didn’t IBM join Ecma if it has such a huge interest in the specification? We didn’t join the OASIS TC because we didn’t care. That’s also why we didn’t try to block it at ISO. It wasn’t something we were involved with. If you are going to try to pretend that your interest in blocking Open XML has more to do with making it a good standard than it has to do with blocking it from being an equivalent to ODF in terms of ISO, you’re not fooling anyone. I took a look at your blog lately and it’s like watching Fox news. It’s ridiculous. I understand you guys are the underdog, and its fun being in that role but this is really getting bad.

    And to say we launched a denial of service attack by bringing a 6,000 page spec is spoken from the true lips of a non-implementer. ODF is underspecified just about everywhere you look. That’s why noone can interoperate with it. The Open XML file format has been under development for almost 10 years. It then went through a round of standardization and documentation at Ecma for a year where it grew from 2,000 to 6,000 pages because folks like Apple and Novell (who are actually implementing the thing as you’ve noticed) needed more information. It then became a standard and as you know the fast track process is set up so that SC34 doesn’t need to create the standard, just do a final review on it.

    Again, if you care so much about changing the spec, why didn’t you just join TC45? My guess is because you only care about blocking the thing. That’s why the comments you log aren’t from the perspective of an implementer, they are from the perspective of someone who had a list of terms and issues they thought would be controversial, and they went through and did a search on the spec to find every occurrence of those terms and issues. Effective for your goals, but overall it’s just sad.


  17. Rob Weir says:

    Doug, I did find many of these bugs during the time OOXML was under review by Ecma.  I’ve been posting about these problems since last July, well before the specification was approved by Ecma.

    I’ll make no apologies for my technical comments.  They were reviewed and approved in the US by a committee that included yourself and over a dozen MS partners.  Hundreds of these comments were accepted by the US NB.  You had your chance to object to them, but you didn’t.  Soon will be your chance to actually address them.  This is progress, yes?

    Brian, you are responsible for the FUD you post, not Gary.  You need to verify your facts yourself, not rely on others. It is enough that you rely on me to debug your OOXML specification.  I’m not going to be fact checker for your blog as well.

    As for whether IBM is an implementor or not, let’s just say that IBM, and Lotus before that, have a long history of attempting to support the numerous poorly documented and flawed Microsoft Office file formats.  I doubt it will end just because Microsoft has put angle brackets around their binary formats.

  18. n4cer says:

    People in glass houses shouldn’t throw stones. Lotus is responsible for some of the flaws in the MS’ formats, and it’s not like there was open documentation available when Lotus, WordPerfect, and others were market leaders and MS had to add compatibility for their respective formats.

    Competitors conveniently act as if MS is the only one to ever have proprietary formats. IBM and others in the industry have created (and still create) proprietary software and hardware, and MS could likewise argue about some of IBM’s boneheaded hardware decisions that they’ve had to support over the years.

  19. Red Weird says:

    "let’s just say that IBM, and Lotus before that, have a long history of attempting to support the numerous poorly documented and flawed Microsoft Office file formats"

    Wow. Talking about IBM and Lotus implementing poorly documented Microsoft formats… Ok, I’ll bite.

    I know Microsoft documented BIFF format back in 1994(!), here: Did you think OOo folks (and numerous others before them) had to reverse engineer everything? And where and when was 1-2-3 file format documented? I know Excel had to implement 1-2-3 quirks, like infamous year 1900 bug, for compatibility reasons (only to get booed for that decades later by anti-OOXML-and-anything-else-Microsoft zealots). Can you name similarly scoped bugs that IBM and Lotus had to implement? Go on, I’ll wait right here with my bowl of popcorn.

  20. Anon Coward says:

    <blockquote> Quote of the Day:  

    CEGID – France

    “With Open XML formats, office documents easily integrate into enterprise solutions, and management data is found within the files themselves. This format lets us combine structured and nonstructured documents and use office applications to pull information straight from ERP. As long as the content remains open and accessible, the Office Open XML format makes integration and compatibility between Office documents and enterprise applications easier. It is part of our long-term vision, because of its robustness and accessibility”

    – Christophe Raymond – CTO</blockquote>

    A *community member* wrote that? You have got to be fucking kidding me! That’s a goddamn press release. No human being who is a member of a community would actually write that kind of crap as a contribution to, or as a discussion point for, the community.

    And you’re putting this out as a good example of the sorts of things that people are saying over there?

    Man, either you’ve got a hell of a lot of people over there talking to each other in language even more impenetrable than this (which would be seriously *weird*) or you just have zero clue as to what makes up a proper community.

  21. John says:


     Ah, I had suspected that the whole thing about Gary Edwards being kicked out was being made up.  I asked for some evidence in another thread but nobody replied.


  22. says:

    Rob and John, I’ve linked to the Gary Edwards posts before, and I guess I should have again as clearly anytime something doesn’t back up your position you assume it’s made up… I won’t make tha mistake again

    Stephen has some links and quotes from Gary here:

    Rob, thanks for the other comments as well. I think the ODF guys would rather you spent more of your time helping to improve their spec rather than "debugging" Open XML. I think the Open Office guys would rather IBM contribute it’s work back to the project rather than taking an early fork of Open Office into workplace and then keeping any changes proprietary. You’re motiviations are clear, and it’s entertaining to watch the drama.


  23. John says:


     I am confused sorry.  You said that Gary Edwards was kicked out.  The link you posted doesn’t say anything about Gary Edwards being kicked out of OASIS, that I can see.

     As far as I can tell, my assumption that you made up the bit about Gary Edwards being kicked out is still correct.  Have I missed something?

     Brian, if Microsoft just played nice for once, then people wouldn’t need to have to fight you guys so much. It would be a win all round.


  24. says:


    Before you go any further in accusing us of not playing nice, I think you might want to look around as there is a wolf in sheep’s clothing (maybe even a couple).

    Go check out Gary Edwards links:

    There is one titled "Is Sun Friend or Foe?" from July 12, 2007 where he states:

    "It was at the height of our List Enhancement battle with Sun that OASIS stepped in their threat to boot the OpenDocument Foundation.  OASIS carried out that threat in May.  The lesson we learned is clear and unequivocal.  Opposition to Sun, in either the marketplace (da Vinci) or in the OASIS ODF TC, can be quite hazzardous to your health."


  25. Ed Brill says:

    Hi Brian, I posted this early this morning but perhaps it didn’t submit.

    You wrote, re Sweden:

    "In Sweden, the IBM side started bringing in people who hadn’t even read the spec and just wanted to vote "no". Google joined at the last minute with the sole plan of blocking it. We asked a number of the companies we knew who want to see Open XML adopted if they would also participate and show their support."

    My comment was, the list of companies that voted in the SIS is public, referenced in comment #1 here. Other than Google, all the late joiners were Microsoft partners, and it’s not like Google is some kind of IBM stooge, either. So, where were all the people "the IBM side started bringing in"?

  26. Andrew Sayers says:

    I found that little back-and-forth with Rob Weir to be more enlightening than ten blog posts from any of the people involved.  There’s been a tendency for people on both sides not to directly address the arguments of those that disagree with them, so that discussion was a real breath of fresh air.

    I’m sure it feels awkward to be talk to people that have so little faith in you, but those of us listening from the sidelines get a great deal out of it.  Some more of the same would be hugely appreciated.

    – Andrew

  27. hAl says:

    [quote]Doug, I did find many of these bugs during the time OOXML was under review by Ecma.  I’ve been posting about these problems since last July, well before the specification was approved by Ecma[/quote]

    As IBM being a long time member of Ecma can I ask you how many comments did you or IBM submit to the Ecma TC45 during Office Open XML developement within Ecma ?

  28. Brian,

    I was nominally part of the "kicking out" that you describe, and I have to say you are wildly exaggerating it.  This was basically an argument over dues, with the OpenDocument Foundation claiming a number of unrelated people as "employees".  They invited me in under these auspices, but I never had a chance to participate before OASIS cracked down.  At that point, Gary and others got grumpy because they had done a lot of work for the TC, and OASIS extended an offer to many of them for a free membership, but some were ticked off at how the whole thing was handled.  It had nothing to do with a disagreement over the specs.

    As for the comment about the Novell suggestions being ignored, I can assure you they were discussed and discussed and discussed.  Finally, the one list proposal was rejected in favor of an alternative list proposal.  It was messy, as all committee work is, but clearly followed through on the wishes of the majority, who certainly were not mostly IBM folks.

    Just thought I should set the record straight.  I still feel like Microsoft has made a huge blunder by ignoring comments early on and then pushing forward with all flaws in place since then.  I am grateful to Microsoft for documenting OOXML, but I have been beating my head against a wall for months trying to implement code using OOXML, and I have to tell you, it is a bear.  I hope that the comments are acted on, but I also hope a TC takes some time to think through the complex dependencies, which make it extremely hard to create or manipulate content.  ODF has been much easier to work with, but I would rather support both.  I am just unsure sometimes of my ability to wade through the OOXML specs, and failing to do so would hurt me as well as Microsoft.  For the sake of us developers, could you all get past the political battle and clean up the mess some?

    – Ben Langhinrichs

    Genii Software

  29. Dave S. says:

    There’s a rumor about a memo to Swedish developers. Anyone care to post a copy?

  30. Andy says:

    There is no point in denying the facts about Sweden. I get reports about committee stuffing and irregularities on a daily basis. Its big news over here, really no orchestrated campaign behind, few blogposts and many journalists investigate it.

    This is what betanews write

    As you know betanews is your inoffical press agency that always takes the utmost spin.

    "Illuminet CEO Jonas Bosson was one SIS member who was urged to join at the last minute by one of the leaders of the Foundation for Free Information Infrastructure, to help balance an anticipated Microsoft-driven surge of five to seven members."

    No one had to "urge" Jonas. After all he is the Swedish president. Betanews tries to sell a different reality but the story is out.

    We understand that you try to depict opponents as fanatics. However, the safest method to survive the storm for Microsoft would be to let ECMA retract fast-track, and then use a normal ISO standard review process for OOXML to gets its ISO approval and resolve the problems.

    Bona fide, that is all. I don’t mind a Committee of neutral experts which reviews comments that were submitted by all parties. Your problem is that neutral review would ultimately lead to disapproval.

  31. says:


    It was my understanding that before some of the Microsoft partners came on there were a number of folk IBM brought in who eventually left in protest. Either way, that wasn’t my point. My point is that it shoudl be fine for people to encourage other folks they know to come out and provide support in the review of a standard. Bob Sutor blogs all the time asking people to write to their representatives, and he even recently asked countries to consider becoming P members in ISO. I’m not sure how it’s OK for IBM to lobby but not for Microsoft.



    I didn’t think I was exaggerating Gary Edward’s statements. In fact his statements are more hard core than my summary. You might want to talk to him if you think he’s not giving it to us straight.

    Also, we are definitely looking into all the comments coming in from folks. The specification will continue to improve over time, as you would expect from any standard. Keep the comments coming! 🙂


    Dave S.

    No clue…

  32. Jeffrey says:


    A short while ago the swedish members of the SIS committee totalled only 9 incuding MS, Sun and IBM.

    Of those 9 they actually voted 5-4 for the approval Of Office Open XML.

  33. Sam Hiser says:

    Ben is basically correct about the OpenDocument Foundation & Gary except on the important flourish: that the CHANGE, not upholding, of the rules had EVERYTHING to do with our position on interoperability.

    Why would one imagine that the Foundation’s many employees were permitted pass-through attendance to the TCs for years (and LEAD some of the Sub-Committees)? Then, after we beat Sun’s plugin in the Massachusetts bake-off with the da Vinci proof-of-concept (Summer 2006) and then published our interoperability roadmap in Dec 2006…

    "<a href="">Interoperability”>">Interoperability: Will the Real Universal File Format Please Stand Up?</a>"

    …then the corporate membership rules of OASIS were changed ("clarified") vis non-profits (Dec 2006) and it all seems like Microsoft is running OASIS — or Sun.

    Right after the Contradiction Phase in Feb, the new OASIS rules were enforced at Sun’s request (we have the smoking gun) when OASIS "reminded" us of the rules for membership.

    Funny thing is, members of the OpenDocument Foundation never block-voted. We are on record for voting thee different ways on two-way votes. Sun was threatened by our interop proposals and "kicked us out". No one, including Gary, ever said HE was kicked out of OASIS.

    But we were excommunicated by rule-change.


    As for your post: it’s full of denial. Most likely the proprietary dependencies which are fundamental to the your spec — and the lack of a reference implementation — will be not resolvable at the BRM.

    Consequently, you will be chucked out just like we were.

    (Here’s to company. My wife asked last night if I’d like some more peas. I said, "No, <em>with comments!</em>")

  34. Brian –

    It is, of course, acceptable for anybody to lobby and encourage their supporters.  I think the sense is that this has gone beyond that level to include Microsoft either paying for memberships or promising some sort of quid pro quo in exchange for votes.  I have absolutely no idea whether that is happening, but sometimes the appearance of impropriety is as bad as the impropriety itself, and Microsoft is pushing very aggressively and with little subtlety in this process.  I think, and I am probably not alone, that Microsoft has lost perspective and is endangering its goodwill in the community by pushing too hard.  I don’t see IBM or Sun doing that, and in fact it feels from here like Microsoft is simply trying to make it look like IBM is the major opponent to OOXML.  News flash, it isn’t working, and Microsoft stands to win the battle but lose the war.  Just my opinion, of course.

    – Ben Langhinrichs

  35. John says:


     Just to be clear – you had gotten mixed up before, and Gary Edwards was not at all kicked off the technical team?


  36. orcmid says:

    Wow, it got a little warm in here, aye Brian?

    Meanwhile, you asked a question.  I think the informative material is very valuable, and you might consider providing an "Office Open XML Annotated" as a separate document.  (Some of the ISO standards appear in books in just that way.  Plauger’s book, The Standard C Library is my favorite example.  It’s not entertaining, but its attempt to demonstrate requirements tracing is admirable.)  

    Another way would be to make some sort of non-normative appendix that is keyed to the normative sections, and that might be better for consistency.  It might also be relatively mechanical to do and maintain with your document-base approach.  

    I think I favor the last idea most, if you must separate out the material, because it preserves the document-base scheme and is the most likely to be reviewed and commented on and improved as developers work their way through it.  

    There is a great post, today, about the power of comments in code:  I think that applies here manyfold.  If ISO can stand it, I would like the non-normative, informative material still collated into the document but set off more visibly and cleary for what it is.

  37. Andrew Hilton says:

    John, I think it’s clear he was taken off the technical team.  What is your point?

  38. nksingh says:


    Have you seen Rob Weir’s blog, or Bob Sutor’s lately?  They are IBM representatives and neither of them seems to have a day job other than bashing OOXML these days.  I think it’s pretty transparent what some folks at IBM are up to.  

  39. John says:


     Sorry that I’m being really dumb here, but what is your source that he was taken off the technical team please?


  40. John says:


     There aren’t that many geeky people that aren’t bashing OOXML these days.  Have you seen slashdot or osnews recently?

     (A skewed demograph I’ll admit)


  41. Fernando says:

    He is back!! Ben Langhinrichs, the IBM partner, is here again spreading FUD.

    "I think the sense is that this has gone beyond that level to include Microsoft either paying for memberships or promising some sort of quid pro quo in exchange for votes.  I have absolutely no idea whether that is happening… I don’t see IBM or Sun doing that"


  42. Fernando – That would be the Ben Langhinrichs who is also a Microsoft partner, and has a substantial business selling products that allow people to use Microsoft Exchange along with Notes/Domino, used by many of the larger companies who have migrated from Domino to Exchange as part of their migration strategy.  Clearly has a real bias for IBM, right?

    My company is an ISV, emphasis on "Independent", and we sell products that manage to piss off both IBM and Microsoft, depending on the day.  I am personally involved in developing products that will help IBM and others that will help Microsoft.  Unless you are a strong believer in the "If you aren’t with us, you’re against us" camp, and think some sort of loyalty oath to Microsoft should be mandated, there is no reason to disparage my feedback.  Of course I am closely aligned with IBM on some matters, since I have made a large part of my business in the Notes/Domino space, but take a look at the CoexLinks product we sell which allows companies such as Honeywell, US Air, RSM McGladrey and Microsoft itself (when they acquired Groove) to interoperate.  Think IBM is enthusiastic about that?

    Brian – I agree that Rob Weir and Bob Sutor are very actively involved in promoting ODF and bashing Office Open XML.  Rob, in particular, makes an excellent case for issues with Open XML.  But how is that any different from your promotion of Open XML and Jason Matusow’s and Doug Mahugh’s and Wouter van Vugt’s. etc.?  As I said above "It is, of course, acceptable for anybody to lobby and encourage their supporters.", which writing in support or opposition on your blog counts as doing.  Sending letters to the Swedish Microsoft Gold Partners promising extra marketing funds, as alleged in the IDG article: is not the same thing at all, and I have not personally seen any evidence of IBM or Sun or anybody else doing the same.  If I did, I would blame them publicly as well, and I have actually refrained from writing about this letter on my blog until I see some more substantial evidence.

    Be careful not to fall into the "My company, right or wrong!" mindset.  Part of the reason I interact on your blog more than some of the other Microsoft blogs is that you have appeared somewhat less biased, and more willing to discuss ideas.  Microsoft is treading on very thin ice, and believing in Open XML as an ideal is different than supporting unethical behavior   in the pursuance of that ideal.  Obviously, you are a Microsoft employee, so I don’t expect you to bash Microsoft, but you might think about equating IBM’s behavior with Microsoft’s unless you have some more solid evidence.  If you do, of course, I’d be interested in seeing it.

    – Ben Langhinrichs

  43. says:


    Please notice one key difference in IBM’s approach and Microsoft. IBM is spending much of their time trying to find ways to block Open XML (not make it better). If they really wanted to make it better they would have joined Ecma. They are spending far more time doing this than they are in doing anything good for ODF. It’s basically Rob’s full time job, and he has a bunch of other folks researching and sending in info to him.

    The only time we’ve ever said anything bad about ODF is when people complained about Open XML’s existance. I explained how ODF and Open XML were both developed in parallel, and also pulled together a couple basic examples of why ODF just wouldn’t work as a default format in Office. This was just something I did on the side though. I lead a team that works on Office 14, I still serve on Ecma working on Open XML, and I try to manage this blog. I have no time (and Microsoft hasn’t asked me) to look for all the glaring holes in ODF.

    I was talking with some folks who have worked on Open Office and document translations the other day about the two formats and they wondered why I didn’t rip on ODF more in my blog. They started listing off a bunch of weak points (no definition for list styles; major holes of underspecification; tables; spreadsheets; functions; interop; etc.). I said that I just didn’t really think it was worthwhile to rip on ODF as I’m not really opposed to it. I care more about Open XML, and as long as people have a choice between the two I’m happy.


  44. says:

    Also for folks who are worried about any type of shenanigans, here is a post from Jason explaining more about the articles on Sweden’s vote:


  45. Brian,

    I understand and respect your position.  The problem is roughly the problem faced by a third party political candidate in the U.S. presidential elections.  When the system is heavily, heavily dominated by one side, that side doesn’t need to rip on the opponent, just make sure they don’t ever have a chance.  Microsoft didn’t attack ODF for much the reason they don’t attack the Mac OS.  It provides a useful cover for a dominant market position.  I am not blaming you, as I would do the same, but it does explain the different positions.

    What it does not explain is the incessant ranting about IBM.  They are hardly the only people complaining about Open XML.  They don’t even originate many of the complaints.  You talk about Rob Weir and how many people echo his complaints word for word.  What you ignore is that various of "his" complaints were started by others, many non-IBMers.  For example, I personally wrote about the deprecated elements such as autoSpaceLikeWord95 in my post Self deprecating standards ( )  in October 2006, and Rob first mentioned them in January 2007, linking back to my post.  So, when all those people echo Rob Weir, they are echoing me as well.  But do I hear Microsoft complaining about Genii Software’s manipulating the system?  Of course not.  If you like, I can point out other small independents who have voiced concerns which have been echoed, not started, by Rob Weir or Bob Sutor.  Microsoft clearly has a strategy of acting as if this is all about IBM vs. Microsoft, because who is going to pick one big money grabbing corporation over another.  You do it as well.  IBM may well be against Open XML, but it is not their fight.  If anything, it has been more actively fought by developers who will have to live with this standard, but that might be too easy for the public to get behind.

    Get your own standard fixed.  It is badly broken, and people like me will keep complaining until it is, or quit in disgust if it is not.  I could care less about the ISO process except that it will tend to fix in concrete what might otherwise be changed by more rational hindsight.  Open XML will still get used by zillions of people regardless of whether it is an ISO standard, but people like me will be a heck of a lot more likely to build products to enhance yours if you work for us instead of against us.  I thought Microsoft was pro-developer!

    – Ben Langhinrichs

  46. John says:


     When you talk about people being able to chose between the two, does that mean MS Office will support ODF as well?  It would be great to one day be able to send a file from koffice (so an ODF file) to my professor and know that he can read it in MS Office, without downloading new software, plugins, etc (He’s not that great at computers heh).


  47. John says:


     Remember in the second post you dismissed the whole Swedish thing as being just FUD, and that you didn’t believe the story for a moment?

     Apparently Microsoft have now admitted that:

    -In a letter from Microsoft, our business partners were informed that they were "expected" to participate in the SIS meeting and vote yes. As a compensation they would get "market benefits" and extra support in terms of Microsoft resources.

    -This was a mistake and the letter was sent by a single employee on his own initiative without sanctions from Microsoft. He also quickly realised his mistake and tried to recall the letter.

    (Translated from Swedish, here: )

     I know you’ll just say that you don’t believe this either and that this is just FUD as well, but the evidence is looking interesting.


  48. says:


    I’ve always appreciated your feedback, especially as we have different views on things. It’s always a good discussion.

    I don’t mean to imply that other people don’t have issues with the spec. It would be insane to assume that in 6,000 pages people can’t find things they wouldn’t want to see fixed. I also think that most if not all of the issues are very easy to address, and are quite trivial in most cases.

    I talk with developers all the time building solutions on these formats. The iPhone, Mac, Windows, Palm, and Linux all have applications that run on them and support Open XML. People are implementing it. Those developers think some of these complaints like the "doThingLikeOldWord" are just silly. They just ignore those tags. There are about 200 doc settings properties in WordprocessingML and 95% of them are fully defined. There are a handful of legacy settings where Word had a layout bug that they fixed at one point, but since somebody might have a document they want to display using that bug we put an option in place for it. It doesn’t make any sense to specify the bug in Open XML (I don’t even know how you’d specify them), but we thought that since we still use that flag we should at least include it in the spec with a rough description of it. It’s no big deal, and is probably about the last thing you’d want to implement anyway. If this really is a huge problem with the spec though, let’s discuss ways to fix it. I’m open and I’m sure Ecma is too.

    IBM is actually going to every single national body though and lobbying with these exact same issues. It’s not an attempt to fix the spec, it’s an attempt to block it from standardization. I believe that other people have very good intentions. The Japanese and British for example have logged some excellent issues and Ecma has every intention of working towards a resolution on them. The IBM actions though are shameful. They don’t even allow the national bodies to form their own list of issues based on what that countries needs are. They are gaming the system.



    I don’t know what your expectation is of my awareness of how thing go around the globe, but let me clear things up.

    1. I don’t participate in OASIS, and so my only awareness of what goes on there is from people I talk to and things I read. Gary Edwards, who served on the TC more than any other member leading up to the original ODF standardization made the claim (backed up by Sam Hiser on this blog) that the foundation was booted because of their push for interoperability. If you think Gary and Sam are lying, take it up with them.

    2. I didn’t originally believe the Swedish story because I know for a fact we have strong policies in place around not pulling any shenanigans when it comes to the ISO meetings. We know how closely people watch what we do and we’re very careful to not do anything innapropriate.

    I don’t think there is anything wrong with asking parterns who we know are also in favor of Open XML to participate. This is just basic lobbying. The thing that is innapropriate is offering some kind of payback in return. I would have hoped we would never do that, but according to Jason someone screwed up, but quickly corrected their mistake (before the vote even took place).

    Why would I say I don’t believe that? I don’t understand your last statement. I can’t tell if your just trying to be a smart ass or if there is something more to what you’re saying.


  49. Brian,

    I am not personally very worried about the deprecated elements – they just happened to be the first item I found from my blog.  I can also find posts pointing out positive things about Open XML and negative things about ODF.  Neither is the point.  The reality is that Open XML is difficult to read programmatically, not terribly hard to create from scratch, but virtually impossible to modify with any degree of certainty.  The developers you are talking about are all reading, or perhaps creating, but few are modifying in any meaningful way.  The dependency chains, or whatever you want to call them, are horrible and inadequately documented.  Therefore, applications which are simple with ODF, where the dependency chains are extremely straightforward, become almost impossible with Open XML.  As you may know, I have been working on a product called OpenSesame, which was originally meant to handle both ODF and Open XML in separate but parallel versions.  ODF has been very easy to work with, with the only limitations being issues with the implementations of merged table cells and that sort of thing.  Even formulas in spreadsheets have been fairly easy to work with, given what we are doing.  Open XML was a non-stop nightmare until I put that implementation on hold due to the issues involved.  I have stopped and started the effort a couple of times since, depending on how fed up I have been.  As of now, it is on hold until further notice.  The resistance by Microsoft to address the complexity issues earlier, and the strong impression of future resistance to change as shown in this push towards ratification recently, do not give me much hope that a reasonable compromise will be reached at a BRM.  These are not minor issues, but major issues, and the specs can’t just be tweaked, they need to be rethought.  In much the way AT&T couldn’t seem to shake its background as a monopoly when it tried to go into the PC business, Microsoft is having trouble shaking its comfortable proprietary stranglehold on office documents.  But remember what happened when Lotus couldn’t handle the transition to Windows?  Right, Microsoft took advantage and charged into the gap with a better Windows spreadsheet.  Microsoft is looking to duplicate Lotus’ failure by assuming that past monopoly would ensure future success, and is similarly practically guaranteeing that result by not stopping and doing this right.  It is like Microsoft is saying, "Nobody tells us what to do", which is just exactly what Lotus did.  Have you used Lotus 123 recently?  Have you used WordPerfect recently?  Why not?  Both thought they could shove their old successful and dominant product into a new paradigm, without adequately taking the new paradigm into account.  Microsoft is following that playbook perfectly.  You own the market now, so why not simply shove your old formats into this new XML paradigm without really understanding it?  Because the market doesn’t leave time for failure.  Even if you delay and delay and fix Open XML eventually and inadequately and then fix Microsoft Office to work with the revised Open XML in, say , Office 2009 or Office 2010, it will be too late.  Making this a standard won’t make it right.  WordPerfect for Windows was still a DOS product kludged into a Windows product.  Lotus 123 for Windows didn’t "get" the paradigm until way, way too late, and is now relegated to a has been.  Both products had similar dominance at one time.  So what do you know that they didn’t?

    If that thought doesn’t shake your confidence one iota, it should.  It is that unshakable confidence that leads people astray.  Do really think Open XML is the right XML standard?  Do you really think it is built to take advantage of the new paradigm?  I am not asking if you think ODF is perfect, or whether you would prefer to see Open XML win, but just whether you really look at that huge, complex mess of a spec and think, "Yup, we really got this new paradigm better than anybody else."  I certainly don’t think so.

    – Ben Langhinrichs

  50. John says:


     I’m not saying that Gary and Sam are lying, I’m just simply asking for a link to where they said that.  I followed the link you gave before, but I couldn’t find a mention of it there.  I’ve honestly tried to phrase my question for a source in the most neutral way possible.  I don’t understand how we are getting lost in communication here.


  51. says:


    did you see my comment where I linked to Gary’s articles? Here’s what I said:

    Go check out Gary Edwards links:

    There is one titled "Is Sun Friend or Foe?" from July 12, 2007 where he states:

    "It was at the height of our List Enhancement battle with Sun that OASIS stepped in their threat to boot the OpenDocument Foundation.  OASIS carried out that threat in May.  The lesson we learned is clear and unequivocal.  Opposition to Sun, in either the marketplace (da Vinci) or in the OASIS ODF TC, can be quite hazzardous to your health."


  52. says:


    This is now becoming a very good discussion. There are a couple things in your last comment that are important to deal with.

    The first is the issue you have around over confidence and whether or not Office will fail. I have differing views on this, and I think that the Open XML work we’ve done, and more importantly the work we’ve done around custom defined schema is very powerful and will take us a long way. We’re heavily focused on how we can redefine what an Office document is, and what kind of a role it plays in business. Integration of custom business data into Office documents is huge.

    Your second issue is more general around whether or not Open XML is the right format. I think there is a need for multiple formats. Open XML was necessary for the majority of our customers, as many of them don’t yet care about XML, and we absolutely need to meet their expectations in terms of how documents will work. I would have loved the opportunity to design an XML format from scratch without having to account for legacy behaviors, but that just wasn’t an option.

    The last point is the one I’m most interested in though. What is the application you’re building and what are the problems you’re having? From the short description I get the feeling you’re working with SpreadsheetML. I think that consuming SpreadsheetML isn’t too bad, and I think generating it actually rocks. The more difficult pieces come in trying to do specific modifications to an existing file. Is that the issue you’re dealing with? What are the areas you’re finding most problematic? Do you have some blog posts or discussion somewhere else where I can go read about them?

    I would love to have these discussions. It’s much more interesting than the typical "leap year bug" or compat settings stuff.


  53. John says:




  54. Ano Nymous says:

    Ben, calculation chain in OOXML is load-time optimization. If you don’t write it, or don’t write parts of it, it will still work – the only thing you will lose is time spent when recomputing calculation chain after your document is loaded. That means you either spend effort to learn how to write optimized files, or you don’t and output less-than-optimal files.

    Writing software is not an easy task and it never was. Just deal with it, ok? 🙂

  55. Daron says:

    Why is it that even Microsoft Certified Gold Partners, some of whom have businesses dependent on Microsoft Office, won’t naturally vote in favour of MOOXML without ‘market benefits’?

    Anyway, my Swedish isn’t great but it looks like the SIS has been annulled as certain parties tried to multiple vote.

  56. Francis says:

    Ben–I would also be interested in examples of the "dependency chains" that are tripping you up in OOXML (and not in ODF.)

    1. Are they complex internal structures (i.e., within a file, that are difficult to parse, interpret, and encode correctly?)

    2. Are they dependencies on the applications themselves (i.e., logic inherent in and only privy to Office?)

    3. Or are they dependencies on the operating system and associated APIs (e.g. on Windows GDI and spooler.)

    Where the dependencies originate matters–it means that different solutions are in order. (E.G., #1 -> change the actual file formats, #2 -> make necessary information public, #3 -> modify the applications.)

  57. Dave S. says:

    Ben –

    "So what do you know that they didn’t?"

    I think they know three things.

    1) They know they can communicate between the apps group and the OS group so that any rich advantage needed from the other group can get done.

    2) They know they can afford really awesome lobbyists to ensure that any resistance is quickly met by a fast education on what the fallout would be if Microsoft lost local support.

    3) They know how much it costs a municipality which is considering an alternative to go through every single publicly owned computer and see if there is any unlicensed MS software on it, though I may have read too much into articles such as –

    For some reason Lotus just didn’t know they should do this.

  58. Dave S. says:

    I’m still pondering this – if the legacy formats were fully documented and adopted by ECMA then couldn’t the MS-XML format be compared to those to see if there was complete correspondence?

    It would also provide reference for those pesky likeWord_usta_do problems as one could just look at the way Word usta do it.

    It would also allow organizations to retrieve from the legacy documents the information they were looking for without forcing them to convert via Office 2007.

    Instead, their own converters could generate meaningful XML-based metadata – such as whether the document had been archived and where and who was currently responsible for the document.

  59. Ano Nymous – LOL.  "Writing software is not an easy task and it never was. Just deal with it, ok?"  I love it.  As for the optimization, I just know it doesn’t load in the one and only implementation, Office 2007.  I don’t even know for sure if what I mean by dependency chains is what you call calculation chains, but whatever they are, they fail (or I do).

    Francis – There are a number of specific issues, which mostly fall into #1 and #2.  I generally don’t use any other API’s aside from the compiler and an XML parser I created from HTML Tidy a while back, which allows good cross platform capabilities and very high performance due to some enhancements I made.  Anyway, there are some complicated structures, but time and patience and asking for assistance would get me through that.  There are still areas there that could be greatly improved.  The second part is harder to define.  The OOXML spec itself claims various things, such as VML being deprecated, but the documents I wind up dealing with keep having VML in them.  Do I write to the spec or to the actual output?  Over years of working with Lotus Notes and its arcane CD record format, I have learned that you have to write to the application, not the spec, because the spec may or may not be right, but the application always is.  So, aside from the sheer density of the OOXML specs, there is a weird mix of under and over documented logic.  There is no point to all the pages which define page borders, but then there is a whole lot of quiet when it comes to how to really use a the ids that Office likes to throw into everything.  It is also very hard to either stop or simulate the tracking of changes, and precious little meta discussion of where change ids come from and when you can simply use your own logic and when you need to follow Office’s internal logic.  And even where the spec is clear, it is not clear whether Office is really following the spec, since things don’t seem to work the way they are specified.  It is not even that the spec is violated, simply that there are rules which seem to be unspoken.  Those would fall into your #2, I guess.

    Another issue, and I really didn’t think it would bother me when I started, as I have worked for years with all sorts of other people’s spaghetti code, but the weird abbreviations and shortened names make it very, very hard to intuitively see what is right or wrong.  In ODF, the names tend to make sense, and that turns out to be much more helpful than in OOXML, where they tend to feel like reading a book without the vowels.  You could probably figure out a lot of the meaning in context, but the effort involved leaves you exhausted.  That is the problem with developing with OOXML – it just leaves you uninspired and exhausted.  How do I put that into a comment for the committees to sort out?  Yet it is a huge stumbling block.

    Anyway, I have spent way too much time on this today, and need to get back to work.

    Brian – You can read about OpenSesame in my blog.  See for a quick search.  It mostly refers to ODF, as that is the focus of the Notes/Domino audience for my blog, but the same functions and features could/should be available for Open XML.

    – Ben Langhinrichs

  60. nksingh says:


    I’m a frequenter of OSNews (slashdot is a bit too inane for me) and I can tell you that most people on that site are simply not objective one way or the other.  There are also lots of people who act like they know more than they do.  

    Look at a Windows Home Server thread there.  Most anti-MS OSNewsers believe that its storage system is purely a RAID or LVM setup without taking to time or mental effort to realize that neither RAID nor LVM have the same behavior as the WHS storage pool.  If they can’t reason about something comparatively simple like the behavior of a storage system, how can someone on OSNews or Slashdot understand something intricate and esoteric like CJK line spacing rules (i.e. autoSpaceLikeWord95).  

    One thing you see really often from the anti-OOXML side is rhetoric without balance.  They rage against things such as bitmasks and leap year bugs and autoSpacing and all this other crap.  Some people make slide decks with this stuff on it, but it’s not like they are really trying to implement the standard and it’s not as if they really understand deeply the points they are trying to raise.

    Stephane Rodriguez is the sole exception among the really visible anti-OOXML posters.  He actually has some knowledge and did raise some valid points with technical depth beyond simply parroting Rob Weir, or GrokDoc, or whoever produced the body of barbs floating around the cloud.  Stephane has actually played around with the documents.  Unfortunately his style of commentary is not constructive and his personal combativeness and tendency towards calling people "bitches" is not conducive to making a format everyone can use.  Perhaps his feedback can be used to fix flaws in Excel that cause problems in maintaining the calc chain (perhaps a timestamp or hash should be added and the chain can be recomputed on the ly based on it).  

    The only people who should care about this standard are implementors of document solutions. The average reader of Slashdot, Ars Technica, or OSNews really do not need to have an opinion on this since the OOXML and ODF wars will not affect their lives or their businesses to a large effect.  They do care about it, though, because so many of them want to stick it to the man by finding some way, any way, to thwart Microsoft in something.  How many times have people mentioned business practices in what should be a technical discussion?  How many people are panning OOXML and promoting ODF with no plans of personally implementing either?  What exactly is the goal the anti-OOXML people are aiming towards?  Microsoft is not going to implement ODF exactly the way OOo does (the spec leaves a lot of wiggle room for app-specific options… and no one is obligated to follow OOo’s lead.  See if Lotus interoperates perfectly…), so you won’t get document interoperability for free with this move.  This whole movement strikes me as a guttersnipe from a group of companies exploiting the general anti-Microsoft milieu in some quarters.  

    When this whole to-do is over, IBM may just drop the OSSers like a bad date and proceed to take a bunch of government contracts with a Lotus suite that supports ODF in name only.  And when you ask Rob Weir to open source his stuff or reveal Lotus-specific format information, he’ll probably repeat the line about how it contains "old legacy code which is covered by licenses and patents outside of IBM’s ontrol."  

  61. John says:


     I find it hard to put myself in your shoes to understand where you are coming from.

     You see Microsoft as this big company that tries to do something good, but always seems to be disliked because they happen to be the guys on top?  Complaints about their bad behaviour etc are because people don’t understand the issues, and technical complaints are just people not understanding the technical issues.

     You don’t understand why people are attacking OOXML and feel that it should be a technical discussion only.

     I can’t speak for anyone else, but my view on it is that Microsoft just keeps on playing dirty.  But you guys don’t see it as playing dirty – but as just business.

     Has Microsoft ever played nice?  Seriously – have you guys ever actually implemented an open and free standard, correctly?  Without proprietary closed extensions, without serious compatibility bugs that go on unfixed for years, and so on?

     I’m sure that there are examples, so please point them out, but I cannot for the life of me think of any.

     All it would take for Microsoft to get a much better reputation would be a few good acts.

     I know people here attack IBM and Sun, and they both do some crappy things.  But at least they are also trying, even if it’s only because it is in their interest.  Nksingh points out the IBM may just drop OSS – but so what?  That’s the wonderful thing about OSS – once it’s free and open, you can’t take it back.


  62. Andrew Sayers says:

    John and nksingh,

    I think you’re falling into a bit of a false dichotomy here – in fact, there’s a wide spectrum of approaches in Office Open XML commenters.

    The loudest, most repetitive members of the Slashdot crowd remind me of fans at a football game – cheering on their team while poking fun at the other team and its supporters.  Their collective volume have some psychological effect on the game, but that’s about it.

    Somewhere along the spectrum from that are the likes of John and I – people that haven’t used Office Open XML in a program, but want to know more about the issues surrounding a story that’s of significant importance for the industry.  We occasionally have a minor effect on things when we come up with good questions, and some of us will probably go on to do something directly useful some day.

    Further along the spectrum are the likes of Ben Langhinrichs and Stephane Rodriguez.  They are implementers that can have a significant effect on proceedings, because as well as asking good questions, they’re able to provide really interesting answers.

    These are three discrete points on a spectrum, rather than the only three types of people in the world.  Lumping one person in with another based on a simple heuristic like "do I agree with them?" blinds you to the important, subtle messages coming from different people.

    – Andrew

  63. John says:


     You’ll always get mob mentality – and double so for the ‘good cause’.

     Take any cause that people fight for and you’ll get nutters or just ignorant people in it – saving our planet (some members of green peace ), vegetarians, saving animals (PETA etc), global warming, most religions, and so on.  That doesn’t mean that there points are any less valid.

     People do make good points on slashdot – there’s a lot of smart people there.  But I think that if the overall message is something that you disagree with, then you can just read the nutters and convince yourself to just dismiss the whole group.  Just the same with any other group of people in the world.


  64. nksignh – Since this seeems to be the FUD mantra du jour from Microsoft, I guess I’ll refute it here the way I did on Doug Mahugh’s blog.  Your snide remarks about "when you ask Rob Weir to open source his stuff or reveal Lotus-specific format information" reveals either ignorance or an attempt to confuse people who don’t happen to know better.  IBM Lotus Notes data formats have always been public, since the very beginning.  The Notes C API toolkit, available as a free download without highly restrictive agreements, describes and documents the Notes rich text format which underlies most of Notes storage, and all the other storage formats other than two, MIME and ODF, that are documented elsewhere.  So, Rob Weir or any other IBMer would not try to restrict access to those formats, but have made them available for free and for generally free use for years to everybody.  So, did you not know, or was I correct that this is the FUD mantra du jour?

    – Ben Langhinrichs

  65. Christian says:

    The Notes NSF-file format seems to be completely closed. The fact that there is a C API toolkit helps nothing.

    MS could say that there is WORD automation and that can be used to access Word.

    Is there any information (of course reverse enginered!) about the Notes NSF-format?

  66. Christian – Don’t be ridiculous.  The NSF format isn’t particularly relevant, as it is the data format you are trying to access.  That is fully described, showing the data structures, etc.  It is nothing like Word automation, because you can take the raw data and decipher it without using the C API if you like.  The data formats are fully  documented, with every byte accounted for. – Ben

  67. says:

    Hey Ben,

    The binary formats (even the new xlsb from Excel) are documented and the documentation is freely available:

    Back to Open XML though. I know this is a lot of work, but if you could start a running list of all the problems you face as you’re implementing Open XML, that would be great. The more specific the better. In the mean time I’ll look through your post on nested tables and see if I can help there too.


  68. Brian –

    Thanks.  I’ll try to create a comprehensive list whenever I come up for air.  And I am fully aware that the binary formats are documented, but thanks for the link.  I agree, back to Open XML, as it needs all the attention it can get.

    – Ben

  69. nksingh says:


    I’m not part of any OOXML marketing machine.  I’m just an interested observer, who is merely a recent college graduate.  

    I’m sorry for being incorrect on the availability of SmartSuite’s format documentation.  I didn’t check, but I based my comment on one of Rob Weir’s (or maybe Bob Sutor’s) blog posts a long while ago responding to people asking for Open-Source Notes.  

    I apologize for any anger I may have caused.  I appreciate in particular your willingness to provide constructive criticism rather than blind sheep-like repetition of GrokDoc points (that stuff is boring).  Sure it’s good to accept legitimate criticism from all quarters, but Brian’s team is made up of human beings so they are not likely to react well to criticism lodged with hatred.  

  70. says:

    That said, I have had to develop pretty thick skin over the past couple years… 🙂

  71. luke says:

    Brian, please stop misusing the term FUD. It is not another word for lies or false information. In fact, FUD is hardly relevant in the subject of XML file formats.

    Thank you.

  72. Andrew Hilton says:


    Arguably the main tool of anti-OOXML campaigners is that of FUD.  Reasoned logic is rare, more common is a scare campaign centred around the notion that MS is inherently bad.  

    Brian has frequently been the target of that term (and many much worse allegations), so you can hardly blame him for having a go.

  73. Doug Mahugh says:

    It’s been quite a year for those who have been blogging about the Open XML file formats. Here’s a look