It's been a week now since Microsoft announced its ODF/Office open source converter project — time enough for at least 183 on-line stories to be written, as well as hundreds of blog entries (one expects) and untold numbers of appended comments. Lest all that virtual ink fade silently into obscurity, it seems like a good time to look back and try to figure out What it All Means.
There are two ways to go about that task. One is the "have it your way," news channel technique (simply pick the channel that serves up your daily news just the way you like it, whatever that may be — liberal, conservative or just plain snarky). Nothing better than the Internet for that, where you can go shopping in the great marketplace of interpretation, and willful misinterpretation, that is the Web, and find more than you could imagine. If you do, you won't be disappointed with myriad ways that people have examined the entrails of the converter story to divine (or dictate) wha's up.
For example, there is metaphorical religious conversion theory, from Martin LaMonica:
Redmond has "road to Damascus" open source conversion
And differences of opinion about whether ODF supporters are jumping for joy or expecting the worst:
OpenOffice developers rejoice at Microsoft's OpenDocument Support
And, of course, there are plenty of theories about what Microsoft may really be up to. Here's a sampling:
Vaughn-Nichols: Microsoft not telling the whole truth about ODF translators
Pamela Jones: MS: OK. OK, so we’ll set up an “OS” project to build an ODF killer. Er, we mean translator
If you want to read just one analysis, though, by all means make it the acute, thorough and balanced report by Redmonk’s Steve O’Grady. His thoughts, as always, are a “must read.” You can find it, together with links to additional reactions to the announcement, here, and a follow up posting by Steve posted the next day here.
The second way is to go direct to the sources, and make up your own mind. Here’s my own take on what led to Microsoft’s decision, and what it’s likely to mean.
Although I’ve read many different takes on this story, I’m going to focus today on just one source: Microsoft itself, and more specifically, the July 6 press release, together with a conversation I had with Jason Matusow, Microsoft’s Director of Standards Affairs. I don’t rank as high in the blogging food chain as Steve O’Grady (he was one of two analysts that Microsoft briefed in advance), but Jason was good enough to call me the morning after the evening the press release was issued, to see if I had any questions.
Let’s start with the press release, which is useful for two purposes: the basic facts, and more intriguingly, the messages that Microsoft wants to deliver. Those messages are both specific to the facts at hand, as well as supplemental to a bigger picture, in that every press release provides the issuer with the opportunity to insert another piece into the mosaic that is the public image it wants to create.
Press releases are especially useful for this latter purpose, because they are extensively worked over and reviewed by multiple parties, and therefore are authoritative. Hence, while they are hardly the most objective sources in the world, they are highly indicative of what the issuer wants the market to think. To an extent, they are indicative of the bathwater that the issuer is drinking as well.
As I read the press release, Microsoft wants the following points to sink in regarding its new converter project:
1. The converters (one each, serially, for Word, Excel and PowerPoint) are being developed at the request of government customers.
2. The converters will be created within an open source project, for maximum transparency.
3. OpenXML and ODF were created for two very different purposes, and OpenXML is far superior to ODF. This will unavoidably result in some deficiencies in how well the converters will work.
4. This announcement is further evidence of Microsoft’s new commitment to “interoperability by design,” a four-pronged approach (only one of which involves an open process — standards).
Here’s how I see these messages fitting into the big picture:
1. At the government customer’s request: I have heard this phrased by two Microsoft sources as follows: “if even one citizen wants to send a document to a government in ODF form, they have to be able to deal with it.” The net desired impression, then, is that the need to accommodate ODF is minimal (so don’t take this as an admission that ODF is taking off), but when the customer asks, Microsoft listens.
2. Open source project: Microsoft deserves points on this one. They aren’t monkeying around, but are putting the code out front and largely in the hands of others, while still paying the bills. Is it perfect? Of course not. But neither is OpenOffice.org, where Sun pays the bills and supplies most of the programmers to write the code, and largely selects what code will be written. It’s only fair to be consistent in how we judge competitors.
3. Different formats: Indeed the two formats were created for two different purposes, and I expect that there will likely be some inabilities for ODF documents to replicate, for example, all 200 Microsoft borders back through 1993. But I assume that there won’t be (or at least won’t need to be) any such problems in the other direction. The main difference between the two format approaches is that OpenXML is a format standard created to serve a single product line, while ODF was developed to enable the creation of multiple competing products, which is already occurring. Losing a few borders along the way is considered to be a pretty easy tradeoff if your goal is the latter rather than the former, because the anticipated rewards are very different. In fact, there is a place for both standards, and they should not be directly compared to each other any more than, say, a land line and a mobile phone should be directly compared, though you can talk into each of them.
4. Interoperability by design: Microsoft has realized that standards are not going to go away, and that customer demand for standards in general, and interoperability in particular, will rise rather than fall. It has taken a thorough approach to creating a new internal standards structure (interestingly, it has many lawyers, as opposed to just technical and business people, in key positions), and has constructed its four-point program to address that need.
It is important to note, at least to me, that Microsoft calls this program “Interoperability by Design,” rather than “Interoperability by Collaboration.” The salient difference between these two designations is that only one of the four roads to the interoperability goal (standards) that the program includes involves an open process. The others leave Microsoft in the senior, or at minimum parity, power position in negotiating the means of achieving interoperability — how, and with whom it pleases.
The official way that Microsoft phrases this “commitment to interoperability” can be found in the same press release (as well as in many other press releases, statements and documents), and reads as follows:
Ongoing Commitment to Interoperability
As demonstrated by the recent announcement of the Interoperability Customer Executive Council and the significant industry contributions to the Open XML file formats from leading institutions like the British Library and Apple Computer Inc. at Ecma International, Microsoft is broadening its long-term investments in and attention to interoperability across industries and platforms through such avenues as product design, collaboration agreements with other companies, standards and the effective licensing of its intellectual property. Additional information about Microsoft’s customer-focused interoperability commitment, including an open letter titled “A Foundation for the New World of Documents” by Chris Capossela, corporate vice president of the Microsoft Business Division Product Management Group at Microsoft, may be found online at http://www.microsoft.com/interop. (emphasis added)
Jason Matusow and I have debated what this means in several blog posts, the latest one of which is here, and you can read 274 comments on and off topic about the same piece at Slashdot.
Now to my conversation with Jason, which was pretty far ranging and candid. Jason said, and I believe him, that the real motivation behind the conversion project is the need to serve government users, and especially those in countries with strong commitments to use and honor ISO standards (ODF, of course, is now ISO 26300). That’s a credible reason, and if converters are going to be built anyway, as they are, Microsoft might as well be seen to be facilitating rather than holding back, and having at least some say in how the process develops.
I also believe that placing the project in an open source venue was a smart move, and an honest effort to be seen as not trying to play games. As Jason said — and who can question the statement — everything that Microsoft does is going to be questioned and attacked, so they decided to initiate the project in a way that would leave as little to question as possible. Of course, one can still poke at different aspects of how things are set up, but that’s inevitable, given that certain decisions have to be made, and when they are, they have to come out one way or another, each with intended as well as unavoidable potential implications. The choice of the BSD open source license is a good example of this, and you can find quite a bit of discussion on line about whether this was a good choice or a bad one, and what the motivations might be for so choosing. Jason answers a few questions on this topic in the comment thread at his blog.
On a related note, I asked Jason why there was no mention of the converter project in the May 19, 2006 Microsoft response to the Massachusetts converter RFI, given that the concept had obviously been kicking around for some time. He responded that final plans for the project had only come together in a detailed fashion in recent weeks, and that they did not want to be accused of making a “vapor ware” (my choice of words, not Jason’s) announcement that could be suspected as an effort to chill independent development efforts without a real intention of delivering on the promise. Again, that’s a reasonable enough explanation, even if other considerations might have been involved as well.
More intriguingly, Jason also noted that a decision like this is still difficult within Microsoft, with some constituencies hewing to the historical, proprietary way of looking at things, while others are arguing for a more adaptive, open approach.
I think that this is accurate as well, and have heard the same observation from various people I know inside Microsoft for the past year, and at each step along the way as Microsoft has loosened up in the ODF saga: first, on licensing terms, then on issuing its covenant not to compete, then on the submission to Ecma, and so on. It is logical to assume that just such a process of cultural shift would be required, that it would be difficult and slow, and that change agents would need real issues in the customer base to point to in order to carry the day.
So my personal take is that we are observing a fairly consistent, significant and, well, fascinating evolution in the strategic thinking of one of the most powerful players in the IT industry. Placing this progression in tectonic terms, there have been some major shifts — earthquakes, if you will — at Microsoft in the past year, such as the organization of the new open standards (and even open source) structures within Microsoft, and the genesis and articulation of the Interoperability by Design public message. We have also seen smaller tremors and aftershocks, of which each concession in the ODF story is an example.
Obviously, there is still much unrelieved tension between the Interoperability by Design message and the real world of technology and customer expectations, both as respects open standards as well as open source software. Perhaps it is as accurate to call the Interoperability by Design program an articulation of an internal “belief system” as it is to see it purely as public marketing message, since I expect that there is passion behind maintaining this halfway house position between a proprietary world and an open environment. Corporate belief systems can be almost as strong as religious convictions, and no conversion is easy or succeeds at a uniform basis at the level of the individual.
Perhaps Martin LaMonica’s Road to Damascus metaphor is not so inappropriate after all, although I doubt that the establishment of the open source converter project will prove to have been the particular step in the road at which the revelation was delivered. But some day, I think that the remaining tension between Microsoft and the marketplace will need to be released. Whether that will be through a revelation that well-serves Microsoft and its customers, or through an earthquake (a catastrophic antitrust penalty? The rout of Microsoft products by Linux/FireFox/ODF and more challengers to come?) remains to be seen.
It will be fascinating to see whether the transition from Bill Gates to Ray Ozzie in the master architect’s chair will prove to be an opportunity to provide a smooth and easy release of this tectonic tension, or a ratification of the ancien regime that sets up the catastrophe.
Note: Joab Jackson at GCN.com just sent me this link to a Q&A of his own with Jason, in which Jason responds to some issues raised by David Berlind and Simon Phipps.
“In fact, there is a place for both standards, and
they should not be directly compared to each other any more than, say,
a land line and a mobile phone should be directly compared, though you
can talk into each of them.”
I beg to differ. Two standards for the same purpose is not only bad, but ultimately dangerous. ODF is Free & free, I can’t say the same about MS Open XML. ODF has no patents, I am not sure about MS Open XML. We have no guarantee that MS, some time in the future, will not change its standard and change its licensing terms along with it.
I thought you are FOSS supporter. After reading this article, I have doubts.
MS has to understand that the good old ways and means that got them to be the big monopoly is no longer going to work. This article is trying to justify some of their shenanigans and present MS as being a cooperative monopoly, far from it. Leopards don’t change their spots and MS is like a poisonous snake; you shouldn’t play with and the only way you are safe from is to kill it.
Well – let me clarify a few things.
First, is OpenXML “a standard?” (we’ll come back later to whether it’s a “good standard,” which is a different question, upon which I’ve written tens of thousands of words at this blog, all of which add up to “no”).
Yes, it is likely to become entitled to be called a standard, if Ecma approves it, and even more so if ISO approves it, the same way any type of car is “road worthy” if it meets all safety, emission and other tests. It doesn’t mean you have to like that car, but it has met the requirements that are established to allow you to drive it on public roads.
Second, yes, Microsoft has at least one patent it claims would be infringed by an implementation of OpenXML, and it has given a “covenant not to assert” that is pretty comparable to the one Sun gave regarding ODF with respect to its patents and OpenXML. You can read more on that if you go back to my blog entries of last fall.
Can it withdraw the standard? No, not once it’s made it’s commitments to Ecma. Can it still play some games? Absolutely, just like any other owner of a patent that relates to any other standard where they can claim infringement, and in other ways as well, since they control Office.
My main point is different – OpenXML really isn’t a “format standard” at all in the same sense that OpenXML is, so it’s silly to compare them. There’s never going to be a knock off of Office built to OpenXML, but OpenXML will make it easier for the myriad people that already develop tools, products, etc. for the Microsoft environment. I’d assume that those folks will be happier with OpenXML existing as a standard than they would be if it didn’t, as will Microsoft customers.
I could go on in greater detail, but that’s where I was headed.
– Andy
Standards that compete are only bad for the proponent of the loosing standard. I though the backers of FOSS wanted choice. Or does this only work when choice works against proprietary software?
I think the concern is this:
HTML was a well defined standard, but this did not stop MS trying to break that standard in many respects. Any implementation of OpenXML is going to be defined only by interoperability with MS Office, in the same way that MS continued to talk about "de facto" standards defined by IE compatibility.
Historically, MS has *always* used file formats as a tool to exclude competitors. People are therefore concerned that this will happen in future. The fact that lawyers are in many important positions merely deepens this, as it implies that the departments are going to need advice on antitrust concerns – something which we do not really associate with open standards.
Do you disagree with this?
Why “anti-trust”? If there is a “covenant not to sue” and the Standard will be in the hands of ECMA?
Microsoft’s covenants not to sue do not cover “enabling technologies which may be required to implement the technology”. What exactly does that mean? OpenXML uses Microsoft’s Open Packing Conventions which is also covered by patents, but I have seen no covenant not to sue concerning them. Would they be considered an “enabling technology”?
What makes you think that ODF is going to lose? It is not a matter of win or lose, it is a matter of MS flexing its monopoly muscle and throw its weight to discredit FOSS.
Both ODF & OpenXML are XML based, so they can be merged at any time. The only problem is when there are controdictory specifiactions. So the question is, if it had good intentions, why doesn’t MS join OASIS to contribute and influence the improvement of ODF? Why did they have to start from scratch? I guess there is something up their sleves.
“I though the backers of FOSS wanted choice.”
Of course we do. Not only want it, but strongly encouraged. But in this case we are talking about a standard not software applications. You can have as many app. using the same standard and still safe from lock-in or being depended on one vendor who controls the standard. Assume OpenXML is for free, could any FOSS organization comment and improve the standard? The way it is now OpenMXL is controlled by MS, It hasn’t been made clear by MS whether they will allow enhancnements or modification by other groups. On the other hand, ODF is open to all kinds of change as long as they are apporved by OASIS members.
Let’s not be naive, the track record of MS doesn’t give any assurance or confort to tust them.
The big quetion remians “why did MS leave OASIS and refuse to cooperate with FOSS”? I guess we all know why, don’t we?
The key issue is that OpenXML isn’t much different thatn their current DOC format, which has already has multiple open and proprietary implementations.
OpenXML 1.0 might be open, but there’s no reason to expect that OpenXML 1.1 will be or that it will be even remotely backwards compatible any more than you can expect the DOC format to be. Essentially, OpenXML’s openness is an illusion that may disappear within a single service pack that upgrades MS Office “to fix a critical security errror” or to conform to “Microsoft Genuine Advantage”.
With ODF 1.0, OTOH, the standard is fixed. If Microsoft conforms with it in the current version and doesn’t in the next, Microsoft is at fault and cannot claim ODF compatibility any more. If ODF 1.0 is upgraded, the new standard will be documented (in a forward and backwards compatible way if at all possible) and Microsoft can choose to support ODF 1.1 as well or not, but that support will be unambigious.
The current DOC format has been reverse engineered and incorporated into most competitors products and there is a covenant not to sue in place. Second if ECMA will administrate the OpenXML format and MS has only 1 vote in ECMA why would MS stray from OpenXML? What would there be to gain?
> The current DOC format has been reverse engineered and incorporated
into most competitors products and there is a covenant not to sue in
place.
Precisely. There’s no advantage or OpenXML, and two disadvantages: the illusion of extra openness and incompatibility with previous versions of MS Office.
> why would MS stray from OpenXML? What would there
be to gain?
They’d gain a competitive advantage over any other conforming implementation because they can support the OpenXML and Microsoft OpenXML format. They only need the OpenXML format in place to get their foot in the door to any organization that thinks OpenXML is a standard. Once they’re in, the inertia of that decision will keep OpenXML in place, so changing is relatively low risk. If people start to balk, no problem, promise to release OpenXML 1.1 to the standardization process over the course of a year or two.
For me, the charter of the ECMA technical committee is a source of concern. It says that ECMA OpenXML is supposed to be fully compatible with the format used by Microsoft Office. Unless they change the charter, they are bound to follow whatever Microsoft changes Microsoft puts into Office. Who controls Office?
Andy,
I agree with some of your points but disagree with the co-existence
of two standard, not when one is fully open while the other is
controlled by a monopoly.
MS has a long history and we
shouldn’t forget it.
They took Kerberos, extended it and kept it
closed from everyone else. It wasn’t illegal but wasn’t ethical
either.
They rejected the Internet until they realized it was not
in their best interest.
They rejected TCP/IP and developed their
own NetBios protocol. They had to abandon it after realizing they
will be left out of the Internet protocol.
They took W3C HTML
standard and extended it and never supported the W3C standard until
recently with IE 7 and only because they started to realize they are
losing their clout and after they created the browser incompatibly
mess we have suffered form for so long.
They took Java/Script and
corrupted the whole standard. If it wasn’t for Sun suing, we would
have had a bigger mess. They created .Net framework which is mostly
Java to bring in very modest improvements over Java.
They are
doing the same thing with ODF. They tried to discredit it but
couldn’t. They tried containing it but they failed. they tried
deception with XML shell around binary format, they were exposed. Now
they are trying to sabotage it by creating their own implementation
of standard XML format.
MS has been touting inter-oprebability
with FOSS lately. If they were truthful about it, why didn’t they
stay with OASIS? They could have submitted their enhancements and
requirements for approval but instead, they elected to take their
ball to play their own game. It could have been much less costly to
them than creating their on standard. It would have been received as
a good gesture by the industry instead of being outcasted.
Like
in previous experiences, MS has an agenda up their sleeves. They know
they are going to lose their monopoly eventually but they are not
going to give up easily. MS will play their game and FOSS is not
going to join them. MS doesn’t play fair and we couldn’t blame them. MS
is a corporation that abused its monopoly and it is time for it to
pay for their illegal behavior.
I think the concern is this:
HTML was a well defined standard, but this did not stop MS trying to break that standard in many respects. Any implementation of OpenXML is going to be defined only by interoperability with MS Office, in the same way that MS continued to talk about "de facto" standards defined by IE compatibility.
Historically, MS has *always* used file formats as a tool to exclude competitors. People are therefore concerned that this will happen in future. The fact that lawyers are in many important positions merely deepens this, as it implies that the departments are going to need advice on antitrust concerns – something which we do not really associate with open standards, more with the same old Embrace, Extend, Extinguish.
Do you disagree with this?
Sorry not to respond to this earlier.
I think that it’s very significant that lawyers are closely involved (hence the mention). I don’t think that it’s primarily antitrust concerns that are involved, though, for several reasons.
First, those that I personally know and have met are not antitrust lawyers. So if that was the primary concern, you’d choose them differently. Second, I’m sure that Microsoft has a dedicated antitrust team, although it would certainly make sense to have at least one good antitrust attorney on the standards team. And finally, there are many more day to day patent, copyright and licensing issues that come up upon which critical decisions must be made. Antitrust issues are _big_ issues, and tend to involve either major, egregius actions, conspiracies, or ongoing courses of conduct, and, I think, would originate as a result of decisions made outside of the standards department.
– Andy
I am not sure when competition came to be seen as a bad thing.
Two competing standards are fine. If one is dramatically inferior
to the other, the inferior one will lose. (Of course, if it is
only somewhat inferior, other factors may contribute to a success)
Having now spent a good bit of time “under the covers” of Open Document
Format, I can assure you it has some definite flaws. With effort
and luck, these can be corrected, but it is not so amazingly well
designed that it couldn’t stand the heat of a bit of competition.
Unfortunately, I am not sure that it will get any real competition,
because OpenXML is so specifically designed, nobody but Microsoft will
follow it (except for add-ons for MS Word and such, where it will be
handy to have a more open design). So, we are going to have to
fix ODF without the benefit of much competition, I’m afraid.
That GCN Q & A settles nothing for me, Andy. Jason avoids commenting on why the plug-in is a separate menu item and doesn’t just add another file type to all the open, save & defaults dialogs like the ODF Foundation plug-in does.
Simon
I disagree with Vaughn-Nichols on one point. He says that the translator is going to be a plug-in, and that most users will never bother to install it. However, the main group the translator is aimed at is government organizations, and they all have IT people who can do the installation.
I think the ISO standard for OpenDocument format is ISO 26300, not ISO 23600.
-Gavin
Right you are, and thanks for noting my slip. I’ve corrected it now.
– Andy