Coursey is wrong on Massachusetts

I just read an article by David Coursey, Massachusetts' Move to Open Format is Close-minded, and I'm afraid he has it totally wrong - too much time spent drinking from the fountain of Redmond wisdom, I fear, and not enough listening to David Berlind who understands the real issues perfectly. He criticises the proposal by the Commonwealth of Massachusetts to mandate a policy of using open formats for its business, saying

"I am not sure what the real problem is with using Microsoft file formats. No, they are not open, but they aren't completely closed, either. There are a number of non-Microsoft apps that support them. That makes Microsoft file formats "open enough" for many users."

What a short-term view. The real point is not what applications are available today, it's that allowing the use of formats that are under the control of a single party - without transparency of process or involvement from any other interested group - results in what I call "corporate Alzheimer's", where you are condemned to be unable to use your documents at some point in the future where the tools available today that access the format are no longer available and/or usable. This becomes even more of an issue once the format gets wrappered in DRM, which causes early onset of corporate Alzheimer's. That's the reason the National Archive of Australia was involved in defining OASIS OpenDocument - to ensure future historians are able to access digital source documents key to Australia's history. If we don't use open standard formats, we are doomed to forget.

Coursey goes on to say

"[Mr Quinn has] created the 2007 requirement for an open storage format to create an excuse for removing Microsoft Office from state workers' desktops."

My word, that is worthy of a Microsoft press release. Microsoft could most likely add the same level of support for OpenDocument as they have for previous versions of Word, for WordPerfect and for about 20 other file formats, and do it easily by 2007. Members of the OpenDocument committee tell me they put in a great deal of effort to ensure the format was capable of easy conceptual mapping to MS Office formats, not least because of the need to make the writing of conversion filters easy.

The truth is exactly the opposite of what Coursey asserts. Massachusetts are not anti-Microsoft when they make this decision, as they are at pains to explain - they are pro-openness. Any company that chooses not to support the open, standard format excludes themselves, they are not being excluded by Massachusetts.

The effects of allowing public administrations to use software that flouts standards are painfully clear, as the examples of the Copyright Office and of FEMA have made clear. Sun, like IBM (well done, Bob) have written to Mr Quinn in Massachusetts to endorse the decision, which is principled, wise, brave and most importantly pro- rather than anti-competitive. It seems so obvious that in the participation age documents need to be long-term readable in any word-processor that the only way to object is by invoking FUD and XML schema.

Coursey does get one thing right, though. He says

"I encourage Microsoft to meet Massachusetts' demand by opening its own formats or, alternately, teaching Office to read and write the OpenDocument format."

That's what we've all been saying for years, and the fact they have done neither (their formats are not open because of restrictions on who can implement them and because control is not shared) will be their downfall. Failure on both counts means only the latter is open to them, and they would be well advised to stop FUD-ing and get on with it

Update: Tim Bray has posted on his blog a copy of the letter Scott McNealy sent to Mr Quinn.

Comments:

I wholeheartedly agree with you. Thank you for posting this comment. eWeek did not for some (technical?) reason: It has never been in the interest of Microsoft to offer any interoperability. Most of these "other" wordprocessors have been hacking the MS format in order to open it. If you take a look at the web, most documents concerning MS formats have been produced by (sic!) Open Office developers!! Now MS is changing to XML, it is not in order to offer greater interoperability, it is to enforce its grip on the office market even more. This time by "free" licences and obviously patents. The FOSS community has already seen these "free" MS licenses are not free at all: so they rejected them. Is has come to this that you can only use your computer if MS gives you their OK (using license keys), now in order to access my own contents I have have to depend on MS too?? I don't think so. And so does Mr. Quinn. It's that easy.. And BTW, MS office isn't that great at all. Never liked it. I use LyX instead. Produces decent HTML, PDF and more. Very small files too. And it doesn't crash as much, let alone the productivity boost..

Posted by Hans Bezemer on September 08, 2005 at 08:41 PM PDT #

"If we don't use open standard formats, we are doomed to forget." I work at Microsoft, though in the core XML team, not Office. I don't have any particular opinion on whether Office should support OASIS ODF in the short term. In the long term, asssuming it becomes a REAL standard, with multiple interoperable implementations by mature products, used by large numbers of people, I agree that Office should and presumably will accomodate that business reality. We can (and have) argued about whether being an OASIS standard makes it a real standard or not. I'm a bit skeptical , recalling WS-Reliability, or WSDM ... or any number of other OASIS "standards" with limited market acceptance. Would anyone disagree that if ODF becomes a 'standard' in the sense that anyone can send anyone an ODF document with confidence that they can read it, it will be because of a LOT more factors than the OASIS imprimatur?

My main issue with the quoted assertion is that \*all\* XML formats that have a clearly documented schema and a stylesheet to translate them to HTML can be considered to be based on open standards - they're based on the core XML Recommendations that really are supported by almost everyone today, with proven interoperability. Nobody can DRM them, lots of applications (including the browsers on the vast majority of desktops today) can display them nicely, and several products allow people with varying skills to edit them. Of course, most people will never see the XML and will interact via MS Office, OpenOffice / StarOffice, or one of the various other desktop suites. But those librarians of the future can crank up the old XML tools to read them long after we and our employers are dust, just as they can in your vision of the ODF future. In other words, the standards that provide the secret sauce here are the core XML ones, not the ODF, or MS Office de facto standards, or whatever. There are reasons to prefer a single universal format over an arbitrary number of XML formats, but most of the arguments in this debate do not go there.

But there are also reasons for NOT attempting to force the world into a one-size-fits-all office document format even if it comes from a truly open process that creates a genuinely interoperable standard. For example, notice that- there are at least two other OASIS standards that meet specialized needs that overlap to some extent with ODF - DITA and DocBook. A policy that mandates \*any\* openly documented XML format rather than trying to force everyone into the One True Document Format gracefully accomodates those who need something like DocBook for their high-end technical documents, as well as those whose simple needs are met by XHTML anf for whom ODF is overkill. Furethermore, it appears from the outside that the MS Office XML document formats are optimized for somewhat different use cases -- MS Office (both the software and the format) seems more accomodating to operational business documents wtih strongly typed data bound to external databases and services, hence it needs the somewhat complex features of the W3C schema in which they are defined. OO/ODF, on the other hand, seem to be oriented more towards traditional textual/tabular documents; being based on the RELAX NG schema makes this cleaner than using W3C Schema would, but at the cost of making it harder for data-oriented users to integrate with the large selection of tools that support the W3C Schema. I'm not by any means arguing that the MS/W3C approach is better for everyone than the OO/ODF/RELAX approach, only that it meets a set of mainstream business use cases more easily.

Finally, all this talk about Microsoft's imminent downfall reminds me of another recent situation where the protagonists declared "Mission Accomplished" once the easy part was over. The hard part in Massachusetts will be to get all those folks who have been "liberated" productively up and running on a new and (currently) unfamiliar system, and to keep them happy in a world where innovation is continuous and user expectations are ever-rising. The people at the keyboards don't care about abstractions such as "open standards" vs "integrated innovation", they care about getting their work done with as little techno-hassle as possible. When the dust settles, its their hearts and minds we all have to win.

Posted by Mike Champion on September 09, 2005 at 02:55 AM PDT #

You've made some good points here, Mike, but I have a few simple counter agruments: Most of my friends who have switched to OSS like Firefox and OpenOffice have done it happily with very little learning-curve, both inside and outside the office. That is because MS Office tends to be false feature rich to the point of obfuscating the common features. Most office paper (real or vitrual) tends to be simple and list and table oriented; most people stick to the stock features that a light-weight word processor can offer. The switch in all cases was simple and effective. Secondly, and I think most importantly, the issue here is not a technical one, it is a vendor-locking one. People argue constantly about market share (i.e. WinXP unfortunately came with my latest computer, but the first thing I did when I got it was install Linux; chalk one up for MS), but now matter how you slice the market, computers are increasingly connected to each other, and even if other platforms are not gaining market share (since there is clearly no definitive data), MS machines are talking to a greater number of other platforms, and visa-versa. Issues arise because Firm A has a tightly integrated MS system and independant partner Firm B has a Linux server from which Firm A needs a service. MS monoliths integrate horribly with other platforms becuase of the MS desire to dominate the market (win hearts and minds, as you put it). If I choose a monolithic MS "solution", will I have access to the services I need unless the outside party also happens to have the same monolithic MS "solution"? Probably not. This is not in the spirit of Free Speech and the open flow of information; rather it is an inhibition. I work in a library, and we don't keep legacy hardware on hand for practical reasons. Unless MS is still supporting ancient formats a couple decades from now (highly unlikely) or keeps their source code in escrow and these intrepid librarians are willing to hack their legacy software to run on the platforms of the future, the old documents will be dead. A well-defined and documented public format increases the odds that a format will still exist and be readable in the future and on other platforms. Lastly, I don't want to win anyone's hearts or minds. I want them to be able to effectively use their computers, however that happens best. This, I think, is a major philosophical point to consider.

Posted by Matthew Fitzgibbons on September 09, 2005 at 08:57 AM PDT #

Mike:

The problem with all your arguments is you are arguing from the position of an XML professional, not a productivity suite user. I heard Jean Paoli do this when the previous MSFT XML format was launched, and it was a diversionary tactic then too (rather than the innocent doc-head/data-head thing of 1997). Here are some responses from your comment to illustrate:

  • The truth is, word-processor users are not the sort of folk who use arbitrary XML dialects. You say:

    A policy that mandates \*any\* openly documented XML format rather than trying to force everyone into the One True Document Format gracefully accomodates those who need something like DocBook for their high-end technical documents, as well as those whose simple needs are met by XHTML anf for whom ODF is overkill.

    but what 99% of word-processing users want to do is just be able to exchange documents readily and indefinitely. The way to achieve that is to have a standard, not to have infinite choice. There are indeed a subset - a shrinking subset - of users for whom a custom schema, or a standard like DocBook, is appropriate. Those people don't use OpenOffice.org <em>or</em>MS Office (although I recognise MSFT wants to grow into that market). The vast majority of WP, spreadsheet and presentation tool users just want to be able to author, open the thing anywhere, be able to pass to anyone they need to and know they will be able to do these things forever. A choice of XML dialects is not a need let alone a priority and subjugating document interoperability to schema choice is the opposite of what they want.

  • I realise it's MSFT's desire to confuse office productivity tools with development platforms with rich clients, but that really doesn't help anyone with the work they've traditionally obtained a productivity suite to perform. You say:
    MS Office (both the software and the format) seems more accomodating to operational business documents wtih strongly typed data bound to external databases and services, hence it needs the somewhat complex features of the W3C schema in which they are defined. OO/ODF, on the other hand, seem to be oriented more towards traditional textual/tabular documents; being based on the RELAX NG schema makes this cleaner than using W3C Schema would, but at the cost of making it harder for data-oriented users to integrate with the large selection of tools that support the W3C Schema. I'm not by any means arguing that the MS/W3C approach is better for everyone than the OO/ODF/RELAX approach, only that it meets a set of mainstream business use cases more easily.

    This is exactly the point - I want a word-processor, not a general-purpose web services hosting XML engine. Do that to add value, fine, but your swiss-army-knife needs to also support the majority case - the common-or-garden user with his/her interoperability need. Interoperability doesn't exist if I have to get an XML consultant in to my local school to set up a schema mapping for them.

  • Then, your main assertion:
    My main issue with the quoted assertion is that \*all\* XML formats that have a clearly documented schema and a stylesheet to translate them to HTML can be considered to be based on open standards - they're based on the core XML Recommendations that really are supported by almost everyone today, with proven interoperability.

    Come on now, those of us who got XML started in the market remember that bogus argument. Yes, XML provides a standard for structuring data, the same way ASCII presents a standard for displaying it. And similarly, the data represented can be closed, proprietary, encrypted and unreadable. Just look at the binary objects in the XML format you just retired in Office 11 to see what I mean. You say "Nobody can DRM them" but that's exactly what's happening - encrypted data between XML tags is a part of many applications of XML.

  • Finally, you say:
    The hard part in Massachusetts will be to get all those folks who have been "liberated" productively up and running on a new and (currently) unfamiliar system, and to keep them happy in a world where innovation is continuous and user expectations are ever-rising.

    Please stop asserting that Massachusetts action is about blocking Microsoft. It's not. They will buy from you if you meet their requirements. It's about one of your customers telling you that all your XML technology is fine but what they really want is interoperability based on a standard not controlled by any vendor.

  • OpenDocument already has multiple implementations and this isn't just a pseudo-standard like the patent-encumbered WS-\* trojans:
    In the long term, asssuming it becomes a REAL standard, with multiple interoperable implementations by mature products, used by large numbers of people, I agree that Office should and presumably will accomodate that business reality. We can (and have) argued about whether being an OASIS standard makes it a real standard or not.

You have many fine, Microsoft-serving arguments why you want to belittle your customer and call them a dinosaur. Your customer is telling you their requirements, and your partners - IBM, Sun - are telling you we agree. What's with the defensiveness? Why the blindness to the obvious need?

Posted by Simon Phipps on September 09, 2005 at 11:15 AM PDT #

I agree with the thrust of most of Matthew Fitzgibbons' points - most people don't need the high-end features of Office, people (especially in libraries do need to access information long after the systems and applications that create it are in the landfill, open and documented formats are the best way to ensure this, and I also want users "to be able to effectively use their computers, however that happens best." Since I was a Mac bigot before I went over the the Dark Side, I'm not so sure about some of the specific points - I could never get used to the Linux UIs, and my skepticism about OpenOffice probably comes mostly from how poorly it is supported on OS X (yes Simon I tried NeoOffice/J several times and it ran like molasses on my poor little 12" PowerBook). My point, is that it is \*XML\* not the specific ODF format that offers the "open and documented formats".

Simon takes issue with that because XML is only accessible to XML geeks. There is some truth to that, and the basic reason that I \*did\* go over to the dark side was to sign on with a crew that was determined to change that. Also, there is no disputing that one can encode proprietary data in XML, but remember I'm not shilling for MS Office's format(s) here but for XML itself as the appropriate standards on which forward-looking IT policies should be grounded.

He seems to think that the idea of XML itself as the basis for interoperability is a dated and bogus argument; I was struggling to come up with a pithy response until I saw Jon Udell's take on this debate http://weblog.infoworld.com/udell/2005/09/09.html#a1298 - that said it far better than I can: "I tend to regard two formats related by a transformation as, effectively, the same thing. But that's precisely the point. It's just data. Exposing it as XML matters more than how exactly we do that."

Posted by Mike Champion on September 10, 2005 at 02:29 PM PDT #

Well said, Simon! Please send this good piece of info also to the guyst at PFF http://weblog.ipcentral.info/archives/2005/09/ma_open_source_1.html keeping in mind that these guys have asked the EU antitrust court to join the trial and side with Microsoft :)

Posted by Stefano Maffulli on September 11, 2005 at 06:03 PM PDT #

Actually, Mike, "he" is me - this is my blog :-)

I do not disagree with the fact that XML forms the basis of interoperability - I have been arguing that since 1997. But XML is as you say only the base level. It is a data formatting standard which provides <em>developers</em> with the base on which to perform transforms, engage in decoupled data manipulation and promote application independence for data - all good things, if perhaps open to subversion by use of binary elements.

But while use of XML gives developers a starting point, it does nothing for <em>end-users</em>. For interoperability, end-users need an agreed XML vocabulary to be used by multiple applications. And for vendor independence, they need that vocabulary to be under independent, industry-supported stewardship. In the case of productivity applications, that need is only partly met by the forthcoming Microsoft document format, which is solely controlled by Microsoft and can be changed arbitrarily for competitive advantage.

To argue as you seem to that XML <em>alone</em> delivers interoperability to end-users is indeed a bogus argument that I've not heard seriously used since the 90s. I read Jon Udell differently to you - I saw him saying that XML makes format choices trivial to the developer and thus implying that <strong>there's no reason for Microsoft to hold out apart from lock-in</strong> in this case. Your customer, along with a growing number of others, is asking for vendor independence for their data - as necessary as bread. Would you offer them stone? A developer technology does not necessarily set end-users free.

Posted by Simon Phipps on September 11, 2005 at 08:23 PM PDT #

Post a Comment:
Comments are closed for this entry.
About

Thoughts and pointers on digital freedoms and technology markets. With a few photos too.

Search

Archives
« April 2014
MonTueWedThuFriSatSun
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
    
       
Today