The current state of Configuration Management Standards - an opinion by John Roberts.

At the present time we have two CM handbooks, several CM standards, and over fifty initiatives such as ITIL and CMMI, all contain conflicting, incomplete, and sometimes mutually exclusive CM requirements … none of which contain a standard framework for implementing CM that can be contractually implemented and monitored. The problem started years ago when the military cancelled CM mil-standards in their rush to become more like “commercial” industries … the same industries with all those cost overruns and product recalls you hear about in the news today. So Data Managers were laid off, CM staff cut back, CDRLs reduced, and few CM requirements put into contracts. It did not work, so now the military is trying to put some teeth, “shall statements” and new CDRLs, into the current contracts and standards … which is like trying to put a square peg into a round hole.

Here is an example of how big this problem has become: The cancelled MIL-STD-973 contained 680 shall statements, 32 must, and 184 will. Too many for sure, but at least it provided a standard CM framework. EIA 649-B contains only 5 shall statements, 34 must and 37 will. So we went from 680 to only 5 shall statements. It is no wonder that the Army released their own standard. The current standards are built on trusting contractors and subcontractors to implement proper CM. But it is impossible for them to implement a CM program with only 5 firm CM requirements, regardless of help from a CM Handbook. And it is also impossible for Government and Military customers to monitor contractors for CM compliance. The problem is compounded by the fact that most contractor CM plans are written at a very high level and commit to nothing, allowing their management to cut back on CM without violating the contract. You can deligate authority for CM to contractors, but ultimately customers are still accountable and responsible for proper CM on their programs.

The very same people that called -973 a bad standard and pushed for the removal of shall statements from -649, which created the current mess, are now busy writing and reviewing the new “shall statement additions” for 649. That is very sad, hypocritical, and a conflict of interest in my opinion. We need a real update, not misguided fluff. Dropping Configuration Control from -649 was a bad idea. Introducing "aliases" such as the term "variance" added nothing but confusion. And changing CM Planning into the laughable Configuration Management Planning and Management … as if adding a second “management” to Configuration Management added any real value ... is outright silly! The good news is the person in charge of this new standards project is extremely intelligent, committed and hard working. If anyone can pull it off, he can. We need a single good CM standard and Handbook that can be contractually invoked, has shall statements, and can be easily tailored for use by all industries. When complete, it should be turned over to an international standards body to form the basis of a new CM ISO. But what do I know … I only have 35 years of CM experience.