|
Hi Roger,
I think that XML Schema is more the implementation of a model, more precisely:
1. There are several data models for XML: the infoset, the PSVI (Post-Schema Validation Infoset), the XDM, etc. I would call these models rather than XML Schema itself, even though PSVI is defined in the XML Schema spec. XML Schema converts infosets to
PSVIs by annotating trees with types. All these models view data as trees, but some are richer (e.g., with types or with sequences of trees) than others.
2. Just because something is implemented with XML Schema doesn't mean that it's the same data model. You can store tables or graphs or cubes in XML (imagine RDF/XML, XBRL, etc), i.e., store them as underlying trees, but that doesn't make them any less
tables or graphs for a human user. It only makes Mike's quote even more important, because in some cases, even though data is stored in XML, it's more *useful* to view it as tables or graphs or cubes, etc, but it's not *incorrect* to view it as trees one layer
down.
My 2 cents :-)
Kind regards,
Ghislain
From: Costello, Roger L. [costello@mitre.org]
Sent: Wednesday, January 25, 2017 2:56 PM To: xml-dev@lists.xml.org Subject: [xml-dev] Quote of the day The test of a model is not whether it is correct, but whether it is useful. [Michael Kay]
An XML Schema is a model.
Therefore, the test of an XML Schema is not whether it is correct, but whether it is useful.
/Roger
|