[sc34wg3] Almost arbitrary markup in resourceData

[Lars Marius Garshol]

* dmitryv@cogeco.ca
|
| I would like to concentrate more on specific suggestions:
| 
| TMDM:
| Define additional property for occurrence item: encoding type.
| Encoding type is enumeration:
| . Base64,
| . XML,
| . PCDATA

I think supporting the values XML and PCDATA is pretty much what is
being proposed, though the details are not yet worked out.

Adding "base64" to the list is an interesting suggestion. I'd be
interested to hear what people think of that.
 
| Also define property "namespaces" which will keep track of
| namespaces inside inline occurrences with XML encoding

The TMDM representation for inline XML will have to deal with this
somehow, yes. Whether it will be a property remains to be seen.
 
| TMQL:
| TMQL should be able to return inline occurrences of these encoding types.
| TMQL should allow 'to ask' about encoding types of specific occurrences

Seems reasonable to me, though perhaps not complete.
 
| TMCL:
| TMCL should be able to constraint possible encoding types for specific
| occurrence types.

Ditto.
 
| XTM:
| . Allow to define foreign namespaces in topicMap tag.

Already allowed in XTM 1.1.

| . Introduce new tag  <baseNameXML> and allow to define foreign namespaces in
| it.
| . Introduce new tag  <resourceXML> and allow to define foreign namespaces in
| it.

Not necessary. You can always define new namespaces anywhere you want.

| . Introduce new tag  <resourceBase64>

Hmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm. I think this is the wrong way to
do it, if we decide to support it. Let's see what people think, first.
 
| XTM deserialization
| During  XTM deserialization'global' foreign namespaces should be copied to
| each occurrence object with XML encoding.

It needs to be handled somehow, yes.
 
| Merging of inline occurrences with XML encoding
| Based on Infoset.

Actually, this one is really hard. There is no definition anywhere of
in what cases two XML fragments are considered to be equal. The
infoset does not answer this, nor does the XPath data model. XQuery
carefully skirts around the issue.

I think the definition we'll have to use is that if the Canonical XML
representation of the two fragments are equal, then they are equal.
 
| I think also that it is important to have ability to reference
| topics from regular XML documents. XTM namespace is just overkill
| for that. I would probably suggest defining small specification
| which is analog of XLink, something like TMLink or TMRef.

This has come up now and again, and I think it's a good idea. The way
to get started here is probably to just write up and publish a
proposal to see 

  a) what people say, and

  b) whether anyone picks it up and starts using it.

I'd certainly welcome it if someone were to sit down and do that.

-- 
Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
GSM: +47 98 21 55 50                  <URL: http://www.garshol.priv.no >