Geeks With Blogs
Brad Corbin's Blog .Net and CRM Reflections of a Reformed Procedural Coder

You have got to go check out these documents just posted by Microsoft on their press release page.

Sure, its 200 pages of legalese, but it really is fascinating reading. Basically, Microsoft is claiming that they have bent over backwards to be in full compliance with the 2004 European anti-trust decision, but that the commission has been deliberately vague about what documentation is required and frequently changed their mind, increasing the work required for Microsoft to deliver the required documents. MS also basically says that the commission doesn't know what the heck they are talking about when they claim that the documents are incomplete or unusable. 

Here are some juicy tidbits:

A summary of the Commission's “Statement of Objections”:

6 The Statement of Objections falsely and unfairly accuses Microsoft of failing to comply with Article 5(a) and Article 5(c) of the 2004 Decision. The Statement of Objections rests this accusation on two claims. The first claim is that Microsoft has refused to provide the full scope of "Interoperability Information" (as defined by Article 1(1)) of the 2004 Decision) that the 2004 Decision requires. That is, the Statement of Objections asserts that Microsoft has refused to document the full range of Windows technology required for compliance. The second claim relates to the usability of the documentation of Interoperability Information that Microsoft has provided. According to the Statement of Objections, it is too hard for competitors to use the documentation to develop work group server operating systems that will be able to compete on an equal basis with Microsoft work group server operating systems.

The first objection is interesting, but you gotta love that second objection. Ooo, the technical documents are too hard.

10 ... In large part, the Commission’s criticisms ignore the inherent complexity of writing specifications for software as complex as the Windows server operating system communications protocols. The task required many person years and resulted in more than 12,000 pages of specifications written, of necessity, for an audience of software engineers skilled in the relevant art. Of course, the specifications are detailed and demanding; they are not "unusable."

The revised set of technical documents was completed by Microsoft by the December 15 deadline, but the commission released an additional “Statement of Objections” a week later before they had looked at them!! 

138. Despite Microsoft’s substantial and serious efforts to reach the ever-receding horizon of the Commission’s demands for achieving compliance, the Commission was not to be deterred from its pre-holiday rush to impose a punitive daily fine on Microsoft. The Commission issued the Statement of Objections on 21 December 2005.

And here's my favorite line in the whole thing:

12 ... In short, the Statement of Objections claims Microsoft has failed to create Technical Documentation that the Commission did not read, and for which no competitor has sought a license, all to address a problem about which no customer has ever complained.

From later in the document:

178. One of the Trustee’s principal criticisms of the Technical Documentation is that it "assume[s] prior knowledge of the Microsoft environment."111 This criticism maintains that the Technical Documentation, to be sufficient, must enable a person with no prior knowledge of the Microsoft network and programming environment to implement the work group server operating system functionalities described by the Technical Documentation. As stated by Professor Finkelstein, this criticism of the documentation is "[o]ne of the most surprising of the criticisms advanced by the Trustee" and "can only be regarded as naïve and hardly a good working basis for the tasks that the documentation envisages."112 Indeed, the thought that any company seeking to develop a work group server operating system comparable to a Windows server operating system product would attempt this task using only engineers who knew nothing of the Microsoft environment is entirely divorced from reality.

I'd give this a big “duh”. If you were going to make a direct competitor to Windows, you think you'd want to use people who were somewhat familiar with Windows technologies. There were even some complaints that the documentation wasn't “free-standing”, ie, it made references to other existing external documentation:

180. The Finkelstein Report further notes: "Interestingly where Microsoft has deemed it necessary to copy information from other sources, competitor evaluators have complained."117 Whereas competitors like Sun complain that Microsoft has included too little external information,118 Oracle complains that Microsoft has included too much.119 As far as Microsoft’s critics are concerned, Microsoft is "damned if it does and damned if it doesn’t." Such criticism is obviously biased and unreliable.

Oh man, it keeps getting better. I gotta quote this one in its entirety:

194. The Trustee’s 30 November 2005 Report notes that it was based on a review of "around 10 working days,"133 and the 15 December 2005 Report was issued around 10 working days later. Clearly, the Commission’s rush to issue the 21 December 2005 Statement of Objections, without even reading the version of the Technical Documentation that was made available by Microsoft on 15 December or asking the Trustee to review it, meant that the Trustee’s criticisms were based on a very limited review. The Trustee’s criticisms all must be viewed in this light.

195. A striking example of the Trustee’s inability to conduct a thorough analysis under these circumstances is the Trustee’s so-called "sufficiency test," on which the Trustee heavily relies for his conclusions. In this asserted test of the sufficiency of the Technical Documentation, the Trustee attempted "to specify the design of programs intended to implement" the functionality of "adding a new user."134

196. According to the Trustee, his "unsuccessful work" on this effort took the following time:

• 12 hours on 8 December;
• 10 hours on 10 December;
• 8 hours on 11 December; and
• 12 hours on 13 December.135

Based on his inability, in only 42 hours, to design an implementation of the add user functionality for a server operating system that would be "capable of providing the equivalent functionality as a Windows server to the Windows clients and/or the other Windows servers which might form a part of an organisation’s network," the Trustee concluded that the Technical Documentation had "failed" the sufficiency test.136

197. The Trustee’s Second Report asserts that "to receive and process a request to add a new user" is "one of the simplest possible operations in a distributed system environment."137 As the Finkelstein Report explains, this assertion is a major error:

      "Any consideration of the sufficiency test must . . . be immediately prefaced by a fundamental reservation. . . . . The test may at first blush seem simple because client-side ‘add-user’ is indeed very straightforward; just use the LDAP (Lightweight Directory Access Protocol, an IETF standard) interfaces or the many other routes to solving the same problem. But this reflects a misunderstanding of the selected test, and our view speaks directly to the contrary; the test chosen is one of the most difficult that could have been selected."138

Can't you just right-click and choose New -> User ?  :)

The Trustee apparently wanted a few more pretty pictures in the documentation, too:

206. Complaining that the Technical Documentation does not contain enough graphical illustrations, the Trustee states: "Simple inspection of available textbooks covering many of the aspects of system operation and interoperation show that illustrations are widely used by technical authors to augment textual presentation."155 This remark reflects a fundamental misconception. The 2004 Decision does not require that Microsoft write a textbook. As reflected in the Hirst Report, numerous textbooks that offer detailed background information on Windows server operating systems and the Microsoft network and programming environment are readily available.156 The Technical Documentation is designed to provide a reference to supplement the knowledge of engineers already skilled in the art of developing work group server operating systems. Although in fact Microsoft Press already makes available a number of books (which the Trustee does not appear to have used), Microsoft is not obligated to create instructional textbooks for engineers who lack this foundational knowledge and skill.

Wow. And I thought legal documents were boring!

(Bold and Italics are added by me, in case that wasn't obvious. Any underlining is from the original document.)

After getting all this crap from the commission, Microsoft still agrees to even make the original Windows source available to developers trying to implement these server technologies:

225. These events, together with those of the previous months, made Microsoft more than fearful that the Commission is not interested in Microsoft’s compliance, but rather in painting a false picture of Microsoft as a renegade that refuses to comply. Microsoft concluded that even stronger measures were necessary to demonstrate its total dedication to compliance to the Commission and, if necessary, to reviewing courts.

226. On 25 January 2006 Microsoft therefore took the unprecedented step of voluntarily offering to license the actual Windows source code for the communications protocols covered by the 2004 Decision for licensees to use as a reference in assisting them to develop their own implementations. This source code is the ultimate and decisive disclosure of the relevant technologies for the Windows server operating systems. Indeed, it is the actual embodiment of those technologies, and therefore constitutes some of Microsoft’s most valuable intellectual property.

I don't normally read legal documents like this, but this is really interesting!

Posted on Friday, February 24, 2006 8:38 PM | Back to top

Comments on this post: Tasty tidbits from Microsoft's EU legal correspondence

# re: Tasty tidbits from Microsoft's EU legal correspondence
Requesting Gravatar...
I can't believe you would question Trustee Neil Barrett's competence! According to his biography that comes up in Google, the man has given expert testimony in cases involving paedophilia and is regularly asked to conduct "penetration tests"! I think it's pretty obvious that he's a skilled software engineer.
Left by John Hensley on Mar 14, 2006 7:29 PM

# re: Tasty tidbits from Microsoft's EU legal correspondence
Requesting Gravatar...
Believe it or not, a 12,000 pages "documentation" that contains examples and "implicit knowledge" instead of specifications *is* "too hard" for external developers to work with. Maybe you should read the Finkelstein report (commissioned by MS) itself instead of MS's excerpt. In it, Finkelstein himself says that MS may be just too incompetent to produce a specification others could use to re-implement MS's services. Yep, that's what he says, right in there. Just read the damn thing.
Left by multi_io on Apr 04, 2006 9:37 AM

# re: Tasty tidbits from Microsoft's EU legal correspondence
Requesting Gravatar...

I don't think that the Finkelstein report supports your position at all. I have read most of it, and the executive summary pretty well states the overall opinion of the entire piece:

0.3 We conclude that the interoperability information as provided by Microsoft meets current industry standards, particularly in such a complex domain. We believe that it has provided complete and accurate information, to the extent that this can be reasonably achieved, covering protocols, dependencies and implicit knowledge. We observe that the WSPP documentation goes significantly beyond what is 'on-the-wire' so as to assist skilled and knowledgeable software developers to develop a Work Group Server operating system that can work on an equal basis with Microsoft’s own products within the Windows domain architecture. There are of course further improvements and clarifications that can be made; it is standard practice that these are performed in response to user queries.

Is there something specific in Finkelstein that supports your position?
Left by Brad Corbin on Apr 07, 2006 8:10 PM

Your comment:
 (will show your gravatar)

Copyright © Brad Corbin | Powered by: