|
I've been involved with the development of two standards recently: the CMDBf standard and Service Modeling Language (SML). Each has followed a similar path, starting as a brainchild of an industry consortium and then moving on to a public standards body. While establishing a standard is an exciting goal in that it moves the IT industry forward, the process to achieve that goal can be tedious. Given the varied interests of the parties involved, the fact that the process works at all is quite remarkable.
An industry consortium is usually made up of vendors, who work together to put forth a proposal. The group is chartered for a specific purpose and has a limited lifespan. Membership is usually by invitation only. There are six vendors in the CMDBf: CA, IBM, BMC Software, HP, Fujitsu and Microsoft. The charter is a private agreement that is written by lawyers and signed by each of the participants. The agreement is tricky because the intellectual property issues are complex.
Few people understand how risky a consortium like the CMDBf can be. For instance, a participant could put something into a consortium deliverable for which his or her company holds a patent. Five years later, after the standard has been accepted and widely used, the patent owner notices that it could demand royalties. By that time, the patent ownership might even have changed hands so that the owner demanding royalties might not have participated in the consortium.
Lawyers justifiably worry about that sort of thing. As a counter measure, a clause ceding ownership of patented contributions is often included in the charter. This immediately goes under the legal microscope because no one wants to cede ownership of anything he or she doesn't have to. Eventually, it boils down to some courage and trust from all the participants.
After a consortium has taken a draft to a certain point, they can hand the draft over to a public standards body. For example, the SML consortium handed a draft to the W3C (the World Wide Web Consortium). Public standards bodies usually do not have any special governmental standing, but they do have public rules and procedures for endorsing standards that are agreed to by all their members. The more prestigious and numerous the members, and the more rigorous the procedure for endorsing a standard, the higher the standing of the standards body.
A body like the W3C has strict operating rules on the openness of the vetting process that each of its standards undergoes before acceptance as a standard. For example, the deliberations of the SML standard working group are publicly posted on the W3C web site, SML Working Group, and from there, anyone can post a bug that the working group is obliged to examine. Only after an extensive period of scrutiny will a standard be accepted. That means that almost anyone with an opinion on SML will be heard and given attention. This makes for robust and error-free standards.
The problem with rigorous public vetting is that it is painstakingly slow. Every voice must be heard, and there are many to hear. This is why industry consortiums are so important. The goal is to bring together the most interested parties together, lock them in a series of rooms, and produce something in a limited period of time that will be a good starting point for the public process.
SML has moved on to the public stage, the CMDBf is close to it. All this may be dry stuff compared to building products, but without it, products-and IT as a whole-are held back from becoming all they can be. |