Why OOXML will ultimately fail

In this industry there are short term wins and loses, and there are long term wins and losses. A small victory today for something you are championing can be matched with a subsequent setback somewhere else. The opposite is true as well. As we look back on any particular program we can see successes and failures, even if the overall result was positive. The IT world is not always for the faint of heart.

I think there are many reasons why Microsoft‘s XML specification for its own Office products, OOXML, will ultimately fail in the marketplace. Many people, including myself, have described how it is monstrously large, not particularly good XML and so not amenable to easy processing by standard XML tools, essentially a dump of Microsoft’s own product requirements and mistakes, and will, in the end, be fully implemented by Microsoft alone. If that’s not enough, it just won’t be one of the three main document formats that people use. Let me explain.

PDF files are all over the place. If you send someone a file and you want them to be able to nicely print it but not modify the original, send them a PDF. Adobe made a very smart move by moving the rest of PDF to AIIM and ISO in January. The open community process should work quite nicely for the further development and adoption of PDF. It’s not going anywhere and should easily stand competition from anything else in its niche.

The OpenDocument Format (ODF) stands on a sound technical and XML foundation, is being openly and actively developed by a community of global experts from many organizations in OASIS, is seeing broad implementation in independent ways from both open and proprietary sources, and is perfectly suited for being used in the next generation of online and hybrid online/offline office applications.

Its adoption rate is growing. Teenagers are using it. Politicians are using it. Some CIOs in Microsoft shops are using it at home when it comes time to spend their own money and technical expertise to pick products for their personal use. The huge and growing base of Open Office users are saving and distributing files in ODF format. The next generation of Lotus Notes, the so called Hanover release, will support it later this year.

We’re seeing top-down official policies on its use. We’re see bottom-up grassroots adoption, and we’re seeing flavors of everything in between. It will be the preeminent open document format for creating fully accessible documents. It’s an ISO standard today. It’s free. It’s open. It’s nobody’s lock-in strategy in a standards disguise. For some reason, people like that.

You know what I think is the biggest threat to acceptance of Microsoft’s new OOXML? It’s Microsoft’s own older binary formats for word processing documents (.doc), spreadsheets (.xls), and presentations (.ppt). These babies are going to be around for years and years and years.

There is a tremendous network effect with these formats since Microsoft has such huge marketshare with its currently installed products. If you send someone a .doc file, you can probably assume they have Word or WordPerfect or Open Office and so can read it. Increasingly I’m seeing email with files sent around in both these old formats and ODF format (that is, two attachments).

The transition we are seeing is from the binary file formats to ODF, with some conversion of older files. Over time, more and more files will just be created using ODF. The adoption of ODF is a true worldwide movement. It’s hard to fight a movement and the standard techniques to do so are pretty obvious.

This is how I think it will play out: PDF will continue to see widespread use, ODF uptake will increase at the expense of the Microsoft binary formats, and OOXML will be odd man out. Sure, we’ll see some people use it until they realize they can’t easily exchange them with others. Microsoft’s big challenge will be to get people to stop using the older binary formats and start using OOXML. This is a major challenge and is by no means certain. Indeed, the situation is confused because in Office 2007 there is even an option to use proprietary binary Excel macros extending OOXML.

As Pogo said: “We have met the enemy and he is us.”

For non-editable documents, use PDF. To use the modern XML format that is being adopted globally by individuals and organizations, use ODF. If you must use a Microsoft format, stick to the old binary formats so others can read them. I think OOXML is just too much trouble and I think the market will agree.

Addendum: While it is easy to think that Microsoft wants ISO certification for OOXML because ODF has it, what if that isn’t the real reason at all? If OOXML does manage to become an ISO standard despite all the objections, then presumably it will start getting mandated and the old binary formats will be deprecated. Then people will need to upgrade to the only software that will fully support OOXML, which is Microsoft Office 2007. Clever, if there is any shred of truth in it.

Addendum 2: The longer name for OOXML is “Microsoft Office Open XML” and it is shortened in various ways, including “OpenXML.”

Also See: An “OOXML is a bad idea” blog entry compendium


  1. Bob,

    I understand your point, but… I think your logic is flawed. I don’t believe there’s a real marketplace for OOXML or ODF. They are both a means to an end, not an end unto themselves. There IS a marketplace for office applications. Absent government “intervention” (not a great word, but you get what I mean), the dominate office applications would select the “winner” of the format battle.

    In some ways, this is VHS/Beta all over again. Btw, VHS didn’t win because of technical superiority – Beta was better from a technical perspective. VHS won (in the U.S. and Europe at least) because it won the battle for marketshare.

  2. Swashbuckler, the point is that the new market is exactly for the information in the documents and not the applications. When I can go into GMail and read a .doc attachment in Google Docs, I’ve moved beyond thinking that my primary unit of activity was about using a particular word processor. People will focus on documents and will read them or process them in a broad range of applications, and you and I can use different ones. The web has taught us this about standards and docs that can be read by virtually anyone in any browser (it’s not perfect). We changed from our pre-web behavior and we are witnessing a similar transition within the whole Office 2.0 space plus all the new desktop software that will exist to process our docs.

    We have moved from applications at the center of the universe to information there. We don’t all use the same application, but we use the same standards for the information. Then we have choice of whatever application we want. That’s cool.

  3. I’ll back Sutor up on this one, Swash. It’s about the information. Your well-meaning analysis rings true for 1995-2001…and interestingly there are some big vendors — particularly one who’s name rhymes with ‘Bun’ — who are very much of your mind, too.

    The bus has left the station. And there you are, standing with your bags in a cloud of dust (with only those ancient ideas for company).

    Bob’s Gdocs example is a good one.

  4. Microsoft strategy fails for another reason, the Office platform going forward. Ref BillG memo : http://edge-op.org/iowa/www.iowaconsumercase.org/011607/2000/PX02991.pdf

    OOXML should have been what RSS has been to web pages, a disruptor. Instead, it’s just angle brackets around 15 years of legacy : for instance, they encode text formatting runs in no less than 6 different and incompatible ways. Never should interns be allowed to design Microsoft file formats. At least, in the binary world all those mistakes were hidden or semi-hidden. Now they bite back anyone who would want to do anything substantial with it.

    To mitigate the issue, Microsoft could have shipped revised DOC/XLS/PPT files with XML streams inside for alternate consumption. That would have preserved the install base which now, as you said so rightly Bob, will have to deal with those alien DOCX/DOCM/DOTX/DOTM/XLSX/XLSM/XLTX/XLTM/XLSB/XLAM/PPTX/PPTM/POTX/POTM/PPAM/PPSX/PPSM files ! ;-)

    Everything is so much easier now with Office 2007. /sarcasm

    Office 14, the next release of MS Off-ice is all about applications running “on top of the MS Office platform”, and in particular integration with sharepoint. Microsoft is making here is the assumption that corporations are ready for another decade of writing and maintaining (Windows-only) applications with so many features as Word/Excel/Powerpoint only to add some more to it. Before that happens, OOXML must lose 90% of its weight. And it has to be self-descriptive, carefully designed XML, community-based, standards compliant, cross-platform, small and lean. Hell, it starts to look like ODF is the *right* format to do just that…

  5. Is there going to be a commercial bust-up related to ‘copyright’ and representing documents in free formats ?

    One of the advantages of the Microsoft approach (where you pay Microsoft for Office) compared with the rest-of-the-world approach (where Sun’s marketing budget keeps http://www.openoffice.org/ running, and you download what you want, source code and all) is that it makes a fund to pay upstream providers.

    Now, for me, an ‘A’ is an ‘A’, and I’m not fussed about the details of the font; I’m an engineer, and we’re like that. But I also know that people like http://www.linotype.com/ actually sell fonts; in principle Microsoft could buy a few, bundle them with Office or Windows, and then their users could use these fonts in documents, send them to each other, and so on. OpenOffice.org/Linux users would have to find another way of paying http://www.linotype.com/ to interoperate legally. I know Shuttleworth is generous, but freeing the fonts might cost more than his mailing bill for the Ubuntu CDs https://shipit.ubuntu.com/ and might be more than he can carry.

    Same could happen for ‘arts and borders’ . Maybe ‘border 37′ is a picture of Mickey Mouse, used in Microsoft Office by kind permission of (and many dollars to) Disney Corporation. That wouldn’t be representable in an OpenOffice.org document, at least not in a way that was GPL-able.

    For me, the important thing about a document, or a presentation, or a spreadsheet, is that it should carry the information that I want to present. ISO ODF XML is great for that.

    But it’s not so for everyone. People tend to include commercial fonts, commercial clip-art, commercial pictures.

    How are we going to make that work, in the brave new ISO world ?

  6. How about security? Once Microsoft pulls the plug on the continuous stream of patches that are required to keep its product secure, there is a forced upgrade. If the CIO takes Microsoft for granted and doesn’t look at alternatives, the next office suite is Microsoft Office 2007. Considering all the little annoyances in Office 2007 that makes it inconvenient to write document in formats other that OOXML, then you have an upgrade to OOXML by default.

    This plan implies that Microsoft stands to benefit from its own lack of quality. But it is a fact that its products have built-in forced obsolescence. The upgrade will occur, it is only a matter of time.

    Another reason for upgrade is the desire to benefit from XML. If the CIO wants to extract metada from the document to populate a document management system, or implement an automatic process that extract portions of reports to repost the data elsewhere, he may have a business case for an XML format. Are you sure there is no such upgrade justifications hidden somewhere in Exchange or Sharepoint functionality?

  7. Bob, having participated in standards orgs, I certainly agree that standards are (generally) a good thing. So, argue on that basis, not the ODF is technically superior basis. The average user (which admittedly, is not kind of person who’s likely to be reading your blog) generally doesn’t care about what’s technically superior – I’ll go back to the VHS/Beta as an example.

    Sam, throwing out insults isn’t going to help your cause. Think of it this way: I work in the CTO’s office of a large software company. I’m probably more familiar with the ODF vs OOXML battle than the average worker in the industry, and I don’t have a dog in this fight. Neither the ODF crowd nor the OOXML crowd has convinced me that their chosen format is best, given all the things to consider. So, you might want to focus your efforts on honing the presentation of your position rather than coming up with insults.

    PoIR, you seem to be confusing open source and standards. There are applications from multiple vendors that support ODF and there will be (if there aren’t already) applications from multiple vendors that support OOXML.

  8. A perfect outline for all that’s right about ODF juxtaposed the shortcomings of MS-OOXML. Thanks Bob, for taking the time to post this (with links). I found Bob’s argument true to my experience when I recently wrote a word processor review for DonationCoder.com. Documents are the center of one’s information (at least archived), not the applications that create them. ODF has permanently ceded the big APPLICATION to secondary status. In other words, it’s just another tool as Bob points out, and OpenOffice is just one player on the field. With the “Big 3″ word processors (Word, OpenOffice Writer, and WordPerfect) — not to mention more online apps supporting ODF instead of MS-OOXML — a noticeable shift has taken place. Being vendor-neutral, applications that do not fully support ODF will begin to fade, and those that eventually fail to support ODF as their native file format (at least as an option — as OpenOffice has long done with .doc/.xls), they, too, will fade even faster. So I agree: even if Ecma 376 does miraculously achieve ISO certification, it will be short-lived and irrelevant to ODF’s future. Governments who have adopted ODF will share their experiences and results with others and Microsoft won’t be able to un-write the many cost analyses that show the savings and efficiency of using ODF.

    MS-OOXML reminds me of someone who says DVD is good enough for him. Even TV is good enough. Now you want him to ditch his vast collection of movies and convert or buy Blu-Ray discs? And a Blu-Ray player? No recorder? And a HD TV? How much is that going to cost him? What advantage does he gain from the move? He might get some nice graphics and features, but the cost will never bring more value. (Rewind this media format argument to CDs-VHS-LPs-Floppies, etc.) At some point, the consumer of the document wants a measure — any measure — of control. The OpenDocument Format provides the answer in the world of documents.
    Swash, don’t take offense at Sam, he doesn’t mean any. That’s just his wicked sense of humor (see his PlexNex blog for his broad and light-sided take on lots of subjects).

    Chris, as for fonts, that’s pretty clear, isn’t it? None of the new Microsoft fonts are great onscreen. Segoe UI is fine within the UI, but Tahoma is more efficient (another MS font). I buy Linotype fonts occasionally and have used FrutigerNext LT for years in every application, including OpenOffice, Opera, XYplorer file manager, WinRAR, and so on.

  9. Forget about practical considerations. OOXML is built upon government manipulation, deceit, technical sabotage, lobbying, and maybe (just maybe!) even bribery.

    MA Governor-Elect Names MS Anti-ODF Lobbyist to Technology Advisory Group

    The Sorry State of Massachusetts

    Microsoft offers schools in Mass. free software [to sabotage ODF adoption]

    Microsoft plays Massachusetts Senate card

    Microsoft playing three card monte with XML conversion

    Q&A: Former Mass. CIO feels ‘bittersweet pride’ after battles with Microsoft,

    ,—-[ Quote ]
    | As CIO of Massachusetts from February to November last year, Louis
    | Gutierrez had to endure most of the brunt of Microsoft Corp.’s political
    | wrath over a state policy calling for the adoption of the Open Document
    | Format for Office Applications, or ODF — a rival to the software vendor’s
    | Office Open XML file format.
    | […]
    | Do you see any reason for there to be two standards? If you were
    | starting blank-slate, there certainly would not be value to creating
    | two separate standards. Over time, it has sometimes been useful to
    | have the competition of two standards to keep both sides honest.
    | But I don’t see particular value in the long-term co-existence of
    | two separate standards.


    From the previous CIO….

    ,—-[ Quote ]
    | Quinn: Almost to a person, to anybody involved or who knows about
    | the ODF issue, they attributed the story to Microsoft, right, wrong
    | or otherwise. Senator Pacheco may be a bully but I do not believe he
    | is disingenious and would stoop to such a tactic. Senator Pacheco and
    | Secretary Galvin’s office remain very heavily influenced by the
    | Microsoft money and its lobbyist machine, as witnessed by their
    | playbook and words, in my opinion.
    | Quinn: I believe that the ODF decision will stand. I believe MS
    | will continue to do anything and everything it can to stop it. And I
    | know my seat wasn’t even empty and they (MS) took another shot at
    | the title, to no avail. This horse is out of the barn and I see no
    | way for it to go back in. Remember, all we are asking for was and is
    | for Microsoft to commit to open and the standards process; so
    | everyone looks really bad if the plug gets pulled at this juncture.


    How on earth would ISO approve something that the world rejetcs, even without making amendments? Bug and proprietary ‘extension’ in a standard?!?! Royalies? It’s not even a free format then, let alone an open one.

    Do the world a favour, Bob, and continue to work on public awareness. I am convinced that Microsoft will also continue to work… on deception and dirty tricks.

    From the antitrust case Iowa… (blast from the past to fuel your endeavours)

    Microsoft’s switcher tactics and WordPerfect

    Microsoft offered money to companies to destroy Lotus disks

    Brads (of Microsoft): lets tilt lotus into into the death spiral

  10. PolR said “Considering all the little annoyances in Office 2007 that makes it inconvenient to write document in formats other that OOXML, then you have an upgrade to OOXML by default.”

    It’s a little worse than that. To run VBA macros now, not only the filename must end with “M” but the Windows operating system you run has tighter screws. Windows Vista, to my knowledge, does not allow to run unsigned VBA macros. This is a BIG deal for many corporate deployments out there that did not bother this before. We have a case here where Vista+Office2007 is the worst possible combination.

    PolR said “If the CIO wants to extract metada from the document to populate a document management system, or implement an automatic process that extract portions of reports to repost the data elsewhere, he may have a business case for an XML format.”.

    Here too, something changed in Office 2007. If your document is password-protected then you cannot extract the metadata because the whole package is encrypted not, not just the corresponding OLE streams. This is a breaking change. In fact, to be accurate, it’s even worse. The document type itself changes when you password-protect a document : from ZIP to OLE. Before Office 2007, the OLE document would remain an OLE document. Just wait until those using all sorts of CMSes out there figure that out.

  11. Myself, I think Microsoft’s “integration” – the heart and key of which is the MS OO XML file format – will cause Microsoft to fail, quite ignominiously. The reason? IBM used to have quite a tidy little applications and system stack aka silo for its customers at one stage. All nicely and tightly integrated.

    The customers? All nicely and tightly sown-up and accounted for.

    What did IBM find? Once there was a choice, the customers moved elsewhere.

    IBM learnt a very expensive lesson over those years. Microsoft’s lining itself up to learn the self-same lesson. To argue otherwise, unless there is compelling reason to believe otherwise, involves special pleasing for Microsoft – but Microsoft, unfortunately, is a past master at that.

  12. Bob, I’m suprised you didn’t mention the real disruptive document format, which is HTML. Everyone has a viewer for it already and, as you mentioned in the comments, Google Docs provides an editor for it too.
    I wrote more on this recently:


  13. Swashbuckler wrote “PoIR, you seem to be confusing open source and standards. There are applications from multiple vendors that support ODF and there will be (if there aren’t already) applications from multiple vendors that support OOXML.”

    I don’t quite understand where this point come from. Nowhere I did imply single vendor support of either standard. My point was there exist a reason for CIOs to introduce OOXML based on a combination of security considerations and typical Microsoft Office upgrade processes. I needed to single out Microsoft Office because of its large installed base.

    Stephane Rodriguez, I really love your posts. All these details you reveal in blogs all over the place mean hidden costs to users and third party developers. People that currently built plans to adapt to Microsoft roadmap or evaluate Office 2007 vs whatever alternatives they want to consider need to know these things and factor these costs into their analysis. I really wish this data is made available in one place for easy reference by all these planners.

    When I read your post in this thread, I wonder if Microsoft is pulling another DR-DOS on the market. It looks like they make life miserable to non-Microsoft products and users of currently reversed-engineered Microsoft technology they want to deprecate while the combination of Sharepoint/Office 2007 will appear to work flawlessly.

  14. I believe that, as various governmental units continue to adopt ODF instead of OOXML, which forces their contractors also to adopt ODF, there will be a trickle-down effect. Looking at the quality of conversion between the two formats alone tells me that NO ONE will want to use both. So then, when the leading technology, defense, and construction contractors are told that they can use ODF or work elsewhere, what do you think they will choose?

    If you are a subcontractor of one of these companies, that means that you also have to choose ODF. If you are an employee of any of the above, or if you desire to be, you will choose ODF for your resume format, meaning that recruiting firms and their software suppliers will need to move over also.

    In other words, assuming that the movement in government continues the way it has around the world, by the time Office 14 rolls around, Microsoft will *have* to join the rest of the world if it wants to sell its software.

    For consumers, having a choice of software that is fully compatible with a standard format (I know that hasn’t happened yet, but it will) is a good thing, because that means that they can choose their software based on *price* and *features*, rather than what name brand is in use in their employer/school/city. I see this as the best possible outcome.

Comments are closed