RE: Subset Data Model

From: Uche Ogbuji [mailto:uche@ogbuji.net]


The reason for syntactic compatibility is extremely simple: you can use the large existing body of XML 1.0 processors to process MicroXML.  Notice very carefully my use of "XML 1.0."
..................
XML is defined by a single spec: XML 1.0.  It is not defined by Infoset, XPath, XSLT, XQuery, XDM, or anything else.  Therefore your basic premise, that if FOO is not compatible with someone's pet data model, then FOO is not XML, doesn't pass muster.

-------------------------------------------

The above is not my argument.  I am completely uninterested, and am not discussing the theory of what is or isnt something.  I am trying to ask questions about the goals of a technology and if the proposed new technology solves those goals.
If the target audience were those who are experts in XML I would agree            with you, But again, I am not talking about theory I am asking about practicality and usefulness.

The target audience of uXML (from people who claim to speak for it) seems to be those who are specifically NOT experts in XML hence the "subtleties" of the distinction between the various XML specs is extremely significant to achieve the target goal (which is btw ... what ?)
Simple question:  Can you describe simply to someone who is not comfortable using XML but would be comfortable using uXML exactly what tools he/she would be able to use which would work ?  You had to be very careful (your words) to say "XML 1.0".    Someone who knows the difference between "XML 1.0" and say , an Elephant doesn't seem to be the target audience of uXML ... hence I disagree that my arguments don't pass muster.

Maybe we could start with

A) WHO is the target audience for uXML ?  (So far the best I have heard is those who think XML is too complicated).   That seems a fair audience ... so continue ...
B) What knowledge do we assume they have ?  (By #A I presume very little about XML if anything).   Hence subtleties of what is or is not "XML 1.0" is irrelevant.   What works or what doesn't work is relevant, IMHO.  Or rather how can we make life easier for people in #A.     If your argument is "IF you knew anything about XML you wouldn't ask these stupid questions" is not a great solution IMHO ... This is specifically FOR people who don't know anything about XML and easily equate XML, Infoset, Namespaces, XSLT , XDM, XPath etc.  The people who do know the details don't need uXML (IMHO).

C) Would the process of explaining uXML to them along with what tools will work with it be easier or harder than explaining how to write XML and what tools will work with that ?  Is that story bound to be easier or harder in the future as we write uXML specific tools ?