NOTE: The version of the covenant that is currently posted at the OSP page of the Microsoft Web site has been amended multiple times since this blog entry was written. Accordingly, it should not be regarded as a current or complete as to the OSP as it applies today. Just a few minutes ago, Microsoft posted its "covenant not to sue" implementes of its Office formats. I was able to get an advance copy, and here is a line by line legal analysis and comparison to the Sun covenant that supports ODF. Sun's wins by a mile.[Note: Footnote 3 to the Microsoft covenant was updated at 2:45 PM EDT on 11/23/05; see italicized text below]

I was able to get an advance copy of Microsoft's covenant not to sue, so that I could provide an analysis of it as soon as it was posted by Microsoft at a public Web page. Here it is, with footnotes indicating where my comments apply, followed by a comparison of the non-assertion pledge given by Sun to OASIS in favor of OpenDocument. As you will see, there are significant differences:

Microsoft Covenant Regarding Office 2003 XML Reference Schemas

Microsoft irrevocably covenants (1) that it will not seek to enforce any of its patent claims necessary to conform to the technical specifications for the Microsoft Office 2003 (2) XML Reference Schemas posted at http://msdn.microsoft.com/office/understanding/xmloffice/default.aspx (3) (the "Specifications") against those conforming parts of software products. (4) This covenant shall not apply with respect to any person or entity that asserts, threatens or seeks at any time to enforce a patent right or rights against Microsoft or any of its affiliates relating to any conforming implementation of the Specifications. (5)

This statement is not an assurance either (i) that any of Microsoft's issued patent claims cover a conforming implementation of the Specifications or are enforceable, or (ii) that such an implementation would not infringe patents or other intellectual property rights of any third party. (6)

No other rights except those expressly stated in this covenant shall be deemed granted, waived or received by implication, or estoppel, (7) or otherwise. In particular, no rights in the Microsoft Office product, including its features and capabilities, are hereby granted except as expressly set forth in the Specifications.

(1) From a legal perspective, the question is whether such a statement, made at a website, is binding on the party that makes it. The answer to that would be rather complex, uncertain, and untested. Note that the same issue arises with similar pledges made by Sun and IBM, so this is not unique to Microsoft, with one exception: Sun has made its commitment to OASIS under the OASIS IPR policy, backed by the member contract that Sun entered into with OASIS. If adoption follows with Ecma, as Microsoft hopes, it will be important for Microsoft to make the same pledge to Ecma in a legally binding fashion. I'll return to this issue later.

(2) While referring to the specifications at the Website is helpful in one sense, it is a risk in another sense, as that Webpage is completely under the control of Microsoft, which can change it any time, and as often as it wishes. Thus, until Ecma adopts the formats as a standard, Microsoft can do whatever it wishes, and implementers would need to follow, in order to maintain current compatibility, even if the formats go in a direction they don't like. Nor will they have advance warning, presumably, of such a change. Or, Microsoft could add new functionalities to its own Office products that lie outside the current formats, and not include them in the covenant, giving it an advantage over competitors. Note also that Ecma and its members could request changes to the formats. The net result is that implementing under the covenant will be risky under this language as written until (a) Ecma adopts the standard, assuming that happens, (b) Microsoft enters into a better version of this language in a binding agreement with Ecma (even though this covenant is more stringent than Ecma's own rules would require, and therefore Microsoft would be doing so voluntarily), and (c) Microsoft makes some sort of ongoing commitment to renew its covenant with each new release, to the extent necessary.

(3) It is very significant that the pledge, while "irrevocable," is identified to the 2003 XML Reference Schemas as posted. In order for this covenant to be truly useful, the pledge should extend to all future versions of the Schemas, especially if they are adopted by Ecma and blessed by ISO. Otherwise, the covenant is a dead end. [It has been pointed out to me that a statement on the Microsoft Web page that compares the new covenant with the prior license for the 2003 XML Reference Schemas includes the following statement:

"Microsoft will also be offering this same covenant with respect to the forthcoming specifications for the "Office 12" schema specifications." From a purely legal perspective, it would be desirable to refer to those specifications in the language of the current covenant, but it is reasuring to see that Microsoft has made a public commitment to offer the same commitment for the Office 12 specifications.]

(4) As with every other type of standard, this type of statement can be tricky. While it is the norm to make an intellectual property rights (IPR) commitment only as to what is essential to be used in order to implement the "required portions of the specification," it may be that there are features, capabilities, or other aspects or code that anyone would want to be able to use on the same free basis if they were going to implement the specification itself. I do not know whether that is an issue in this instance.

(5) This is what is often referred to as a "defensive suspension" term, and is neither unusual nor deemed to be unfair. The concept is that someone offering their IPR to the industry for free is entitled to a level litigation playing field if someone turns around and sues them for implementing the same specification. Note that this right only applies when Microsoft or one of its affiliates is threatened.

(6) This is standard "lawyer talk," and makes us feel better (especially those that work in large corporations that feel like litigation-magnets).

(7) "Estoppel" is a legal doctrine akin to entrapment. In the Rambus litigation, for example, estoppel was raised as a defense by some defendants, who said that Rambus could have spoken up, but didn't, thereby leading them to incur liabilities that could have been avoided if Rambus had been upfront about its IPR. Interestingly enough, for the same reason that implementers might feel nervous about the point that I made in footnote (1), Microsoft may not get much value out of this, since there will be no signed contract between Microsoft and the implementer.

Now let's see how the Sun covenant compares:

Sun OpenDocument Patent Statement to OASIS

Sun irrevocably covenants that, subject solely to the reciprocity requirement described below, (1) it will not seek to enforce any of its enforceable (2) U.S. or foreign patents against any implementation of the Open Document Format for Office Applications (OpenDocument) v1.0 Specification, or of any subsequent version thereof ("OpenDocument Implementation")(3) in which development Sun participates to the point of incurring an obligation, as defined by the rules of OASIS, to grant (or commit to grant) patent licenses or make equivalent non-assertion covenants. (4) Notwithstanding the commitment above, Sun's covenant shall not apply and Sun makes no assurance, covenant or commitment not to assert or enforce any or all of its patent rights against any individual, corporation or other entity that asserts, threatens or seeks at any time to enforce its own or another party's U.S. or foreign patents or patent rights against any OpenDocument Implementation. (5)

This statement is not an assurance either (i) that any of Sun's issued patents cover an OpenDocument Implementation or are enforceable, or (ii) that an OpenDocument Implementation would not infringe patents or other intellectual property rights of any third party. (6)

No other rights except those expressly stated in this Patent Statement shall be deemed granted, waived, or received by implication, or estoppel, or otherwise. (7)

Similarly, nothing in this statement is intended to relieve Sun of its obligations, if any, under the applicable rules of OASIS. (8)

Here's how the Sun covenant measures up:

(1) This correlates to my footnote 6 to the Microsoft covenant, but see my comments in footnote 5 below for an important distinction between the two covenants.

(2) Only lawyers that live on the other side of the looking glass could make a statement like this. Sun reserves the right to enforce its unenforceable patents? I can't guess why this is in here, but I think it washes out, since by definition, if you were sued by Sun, you need simply point out to the court that the suit was, by definition, groundless.

(3) Sun here makes the commitment that I note is not made by Microsoft in my footnotes 2 and 3 above. Note also that the version of the specification to which the covenant applies is on the OASIS Web site, and is therefore not under Sun's control.

(4) What this means under the current OASIS IPR policy is that if Sun is part of the process that creates the next version of ODF, it will renew the pledge as respects the new version. If it quits OASIS, or chooses not to participate in the next version, then it's commitment would continue to the existing version, and to those parts of the first version that persist into the new version, but not as to any new material that may be added. Presumably one reason this may have been added is so that if (for example) the members of the revision Technical Committee wished to add a description of a Sun proprietary product to ODF that was not there before, Sun could drop out if it was unwilling to contribute that feature. Of course, by doing so, it would also lose the ability to influence the further evolution of ODF, so there is an offsetting incentive (although not a guaranty) for its continuing participation.

(5) This defensive revocation right mirrors that of Microsoft, as noted in my footnote 6 above, but with one important exception: while Microsoft has reserved the right to protect itself, Sun has reserved the right, if it wishes, to be a "patent policeman" that could assist any implementer of ODF.

(6) See footnote 6 above. The same comments apply here.

(7) See footnote 7 above. Again, the same comments apply here.

(8) This language supplies another layer of protection, secured by Sun's membership agreement contract, not present in the Microsoft covenant. Ecma adoption may provide equivalent protection if it adopts the Microsoft formats.

It is important to note that both of these covenants suffer from a common failing, which is that, as between the patent owner (Microsoft or Sun) and an actual implementer, there is no well-funded White Knight to swoop in to defend the small ISV. Similarly, there is no commitment to make the same covenant binding on someone that might buy the patent (the same problem relates to standards generally).

The upshot is that the Sun covenant is far superior in several important respects to the Microsoft pledge.

This raises the question: Why? Certainly, Microsoft must have expected that a comparative analysis like this would be done almost immediately, so it must have had a strong incentive to match the Sun covenant as closely as possible, and it decided not to. Again: Why?

There are three possibilities. The first is that it has evil entrapment plans afoot, but I really don't think that is likely to be the case, and certainly not in each instance, since it would be rightly pilloried for doing so. The second is that it hasn't gotten far enough through the knothole to bring itself to go as far as Sun did.

The third is that it has made the calculated decision that this is as far as it needs to go to obtain the objective that it is trying to achieve, which is to head off ODF at the pass.

Which is it? My guess is that it's a combination of 2 and 3. I'm told by those I know in Microsoft that making such a covenant was a difficult and contentious decision internally, and it would be tough to sell internally more than the absolute minimum necessary to arguably do the job.

Will it be enough? We'll find out. But on my review, there's a lot of light standing between the two covenants which will provide plenty for people to talk about. Microsoft knows its customers well, though, and it’s the customers ultimately that will say whether Microsoft bet its hand conservatively and still won, or didn't act boldly enough to walk away with the pot.

Given that Microsoft has already collected some endorsements from European governments in favor of its formats, backed by the Ecma announcement (its not coincidence that Ecma is a European standards developer), its clear that its already test marketed its bid with its customers.

The bid now passes across the table. Who's going to make the next bet?

Note: This entry does not constitute the rendering of legal advice; please consult your own attorney before making any decisions involving Microsoft's intellectual property rights.

subscribe to the free Consortium Standards Bulletin