CM wiki web
FAQ: Lies, Bloody Lies, and Terminology
Nobody can predict what other people exactly mean by acronyms such as RCS, VCS, SCM, CM...Worse, few people care; yet all these terms should bind to some kind of information management. Commonly, these acronyms expand as: Revision Control System, Version Control System, Software Configuration Management (or sometimes Source Code Management), Configuration Management, or Change Management?...
It is in fact likely that the use of Source Code Management at least is a deliberate attempt to throw some confusion, to mock any benefit that could be brought by Software Configuration Management on top of Revision Control; or maybe even worse: an Orwellian example of Newspeak. Why isn't there a standard terminology? There is, or actually, there are. Many. The problem is that this is the result of an evolutionary process, one much faster than the evolution of species. There is no much sense in defining globally what a wing is: during evolution, it must have been selected at a stage when it wasn't useful to fly, maybe e.g. as a radiator—penguins actually use their wings to swim. Standard definitions make thus the implied assumption that the evolution has ended, at least from their point of view, but they seldom state it. And standards get old fast, and are seldom revised. In fact, this structural failure of standards is one of the reasons for the success of Open Source, seen as a dialectical anti-thesis over standardization. Believing that information management can only build on discrimination, I suggest to give a different meaning to each of these different terms. I base my discrimination on the differences between:
- Control and Management
- Revision and Version: revisions being exclusively sequential in time, on one and the same line, and versions being a priori parallel
