The definitive opening post on this is Andy Updegrove. Andy is an experienced standards lawyer and does wonderful analysis of the news to date. You would do well to start here because he focuses on the core of the Microsoft strategy starting with the press release and positions the reporting around it perfectly. (He also collects the relevant press articles.)
Jean Paoli (Scoble interview) and Brian Jones (blog) are proud technologists, so discount their blogs and save yourself some time if you're only interested in watching the standards strategy unfold. (This is not a snide remark -- how do you think they'd talk about their work? They're both appropriately strong supporters of this effort.)
The real fuss will begin once the patent license is presented (which both Jean and Brian have misunderstood in the past as well). Ecma International requires that participants offer reasonable and non-discriminatory (RAND) terms around intellectual property with respect to their standards.
It will likely be a royalty free license, because the current patent license around the proprietary specification is royalty free. That patent license, however, couldn't be sublicensed, so an implementer that wanted to license their implementation under the GPL couldn't. Indeed previous examples around the IETF SenderID standard would force users of other implementations to engage in a license with Microsoft which is a rather onerous problem for free and open source licensed software. Microsoft will of course hide behind the "it's our property" rhetoric, and label anyone that doesn't agree as either a "thief" or an "IP socialist". Standards can be a rhetorical game and Microsoft has a fetish about intellectual property (which may well reflect on their constant focus on competitors since IP is a vendor-to-vendor discussion).
As Stephen O'Grady points out in the ZDNet article, there is prior history here. The C#/CLI standards were also developed at Ecma International, a royalty-free patent license offered to essential patents, and Microsoft will still make no definitive comment about the mono project thereby keeping it in an ambiguous state with respect to its potential customers. While Novell continues to sell and ship mono, one can only presume there are lawyers from both sides snarling at each other in the ever present legal struggles between these two companies.
While standards exist to encourage multiple implementations, Microsoft wants to have the imprimatur without the attendant loss in control of what it perceives to be "its property".
It will be fascinating to see the early technical committee meetings. No changes will be allowed to the Microsoft formats of course, because they will claim that "changes are out of scope" due the millions of users presently using the format. (This is certainly how the Open Software Foundation played the GUI Wars around OSF/Motif in the early 1990s.) Indeed the point is to perfectly align the new format standard with the about to be shipped product.
The stress will be discovering how hard it is to maintain exact conformance. One bug fix after the standard locks but before the product ships could break the link between product and standard. Likewise, one necessary editorial correction to the document if the product ships first could render the product nonconforming with its own spec. There could be fascinating compliance angles to be played here by the other vendors.
This is not to be ignored. At least one first tier UNIX vendor in the mid-1990s discovered after they shipped, that their C-language compiler was no longer compliant to U.S. government procurement standards, and had to sit there quietly through a two year ship cycle hoping no one at their competitors or in the U.S. government would notice while they fixed it. Likewise, it would be interesting to know if Microsoft's .NET world conforms to the existing ISO standards based upon its own products. Indeed the reaction to the question alone would be worth an interview. (Are all conformance issues fixed in the .NET bug database? I bet we could check in the open source world of mono.)
The specifications review work will still (of course) be perfectly tedious. One interesting item will be the publication debates. Should the standard in all its complexity be published as the XML schema itself? The precedent was certainly set with the C#/CLI standards and the frameworks at Ecma International. But that forces a different set of difficulties on reviewers, and one has to have the tools that can run presumably on more than Windows. Which means there could be reference implementation code to leverage early, but at what cost and under what terms of use? There are many ways to skin this cat.
One of the more interesting Microsoft commentaries is from Jason Matusow (Microsoft Shared Source maven) . It's good that he's getting his opinions in order. Jason is a master of rhetoric, a brilliant presenter, and a glutton for punishment as the man from Microsoft that gets to present world-wide on Microsoft's open source and Shared Source strategies.
I've long argued that technology standards is a classical game of diplomacy where after the receptions and tea parties are over, the job is to defend sovereign territory while expanding one's area of economic influence. If you're really good, your best work goes unseen. Microsoft would do well to convince Jason to take a new position as technology diplomat to sell the Office standards pitch. It will require someone with his patience and skills and long experience presenting to governments and talking with the press and analysts to carry this forward properly.
Let the games begin.
"The C#/CLI standards were also developed at Ecma International..."
Weren't they mostly developed at Microsoft, "handed to" ECMA, and slightly polished by a committee for publication? After all, Microsoft invented both C# and the CLI, on which many parts of the .NET Framework, etc., depend. So it could hardly allow a non-Microsoft group to make arbitrary changes to them. Please correct me if I'm wrong about this.
Posted by: Tom Welsh | 22 November 2005 at 13:37