dita — archive

[Dana Spradley]

Looking over Don's notes from today's meeting, I paused over the following: >Discussion of Conformance section. >Sue-Laine asked for clarification of the goals. Dana asked about Must >requirements; Eliot: issue is processor requirements vs coding (data) >requirements (conventions for human understanding as well as machine >interoperability). If our coding standards are only intended to promote human understanding - and aren't sufficient to automate the interchange of new DTD/XSD modules - can they really be "must" requirements? I don't think so... Basically: if by following them in a fashion that DTD/XSD editors could automate, you can't ensure "plug and play" interoperability between DITA installations and across different vendors' DITA implementations - then they are "may" standards at best - or merely best practices that the TC itself follows in the DTDs and XSDs we release. Perhaps that is why they were originally called "design patterns," not "coding standards." --Dana