Those XML mention names

I mentioned to George Bina our discovery about the element names in the XML-mention domain. Mostly I was just expressing my surprise that no one of us had noticed before, presumably because no tools flagged those element names as being any kind of problem in a DTD. I confirmed for myself (both in and outside of DITA) that XMetaL certainly doesn't, and George confirmed that oXygen doesn't either. He thought that some earlier versions of Xerces might have given warnings.

He raised a good point about why this is correct behaviour for the tools: The spec doesn't say that these names are invalid, just that they're reserved for future W3C use. If W3C were to use such names in a DTD someday, we wouldn't want our tools rejecting their DTDs.

He also suggested (and I agree) that our decision to use/not use these names is therefore a risk/benefit analysis: What is the risk, should we decide to use the names "xmlelement", "xmlatt", etc., that W3C would come along one day and declare that the names are going to serve some official, other purpose?

Maybe we should weigh that against the aesthetics of these names compared with the "xm-" versions, before we make our final decision.

mag

dita message