xml-dev - Re: XML terminology [was Re: Pontifications on the Perversity of Pedant

[Rick JELLIFFE <ricko@geotempo.com>]

Jonathan Robie wrote:
> 
>  Peter Murray-Rust wrote:
> 
> > I agree that for newcomers to XML the terminology is probably
> > counterproductive. I would never dream of trying to explain to a chemical
> > software engineer that <molecule> was a tag containing an element type
> > name. I might well say "your tags have to nest properly". But for formal
> > discourse among XML developers we have to use precise terminology because
> > otherwise we have confusion. We have seen this in the past on both XML-DEV
> > and XML-SIG where imprecise use of terminology led to different approaches
> > to documentation, and thereby to implementation.
> 
> Unfortunately, I'm not sure that the terminology is as precise as it should
> have been. If it were, there would have been no question about how to
> determine the name of an element or attribute, what - if anything -
> constitutes a type, etc.

I think SGML got it right with "generic identifier".  A "type
identifier" is the thing that should appear in a type declaration; the
"element identifier" should appear in an element declaration; in the
case of DTDs these two functions are combined (hence "element type
identifier" which reflects the dual purpose of the declarations).  

But there is no reason why an "element identifier" in an element
declaration needs to be a GI: it could be an XPath (in some schema
language) for example. If XML terminology is clarified, I hope "Generic
Identifier" is resolved.

By the way, I am notsure that "node" is an adequate substiture for
"element" at all. If an element is a node, what is an attribute? If we
are talking trees, it must be a node too. If we say it is a "property"
or an "attribute" we have left simple trees and are in AVT or some other
specialist jargon anyway.

Rick Jelliffe

***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************