15 Aug 2000

© © 2000 Applied
Information
Science

• accepted unquestioningly by everyone
• complete in concept and methodology
• compatible with all vendors and tools
• ergo ~ sine qua non to modern information systems engineering.

Well, some of us are a little slow on the uptake. I've read the standard doc since version 0.8 or earlier. I've lurked in the news groups, endured the seminars, tried out the tools, chatted up the self-proclaimed experts. But I still don't get it. 

In my analysis, the UML is a loose amalgam of a lot of older, good ideas and some poor ones. But it is not a revolution; not a comprehensive way of discovering, specifying, designing, and building software; not a guarantee of bug-free code generation at the push of a button; not even a decent way to document most software projects.

I'm not trying to knock Rational Corporation or its three amigos who unified their personal styles to arrive at the UML. Although Rational is perhaps at times irrationally exuberant in promoting UML, nonetheless consumers of any so called information technology have a responsibility of their own to understand what they use, to not march over a cliff blindly.

[much more to come ...]