Inheritance and other buzzwords

[james anderson]

Rick Jelliffe wrote:

> I think the wd-namespace will work as advertised, but I think a lot of
> people
> will try to use it for more than the uses that it pupports to address.

the issue is that, if completely specified, it would accomplish the things which
you, yourself say need to be accomplished. one would not need to "leave them to
he programmer". if not, then inconsistent interpretations will arise and the
problems, which you note, will prosper and flourish.

> Following are
> some general comments.
>
> > in previous remarks, it has been explained that the wd expressly avoids
> > specifying a semantics.
>
> 1) Some people claim that the difference between a hyperlink and an entity
> declaration/reference is that there is something voluntary or contingent
> about a hyperlink while an entity declaration/reference expresses a
> more fixed and necessary relationship.  If we accept that for a second,
> then I think Andrew's comments are clearer: the schema nominated
> in the namespace declaration is not "specified" rather it is "identified".

since all of this only matters anyway  in the context of validation and
entity/attribute defaults (in general, when the intent is to ascribe a behaviour
which is specified elsewhere - otherwise you could go ahead and name your
entities anything) the only time it matters whether you have an unambiguous
name, or not, is when the intent of the "identification" is to "specify".

> ...
>
> So namespace PI is more like a hypertext link rather than an entity
> reference.

which link has to be guaranteed to lead to an unique location should you follow
it.

> ...
>
> 2) I think the term "scope" shouldn't be used here: all namespace PIs have
> scope over the entire document, not over particular entities.

that's the probem. if they have indefinite scope, then you get ambiguities. if
they have dynamic scope, then you don't. to be more complete:1. the schema
identified as a namespace source has indefinite extent. that is the entities
(elements, attributes, pe's(? not sure), (ge's), notations, ...) defined within
it are valid within the process which references the schema.
2. the namespace name (the binding of the name to the namespace-region of
schema) has indefinite extent. (ie. the universal names are universal - and not
just within the document, but within the process)
3. the prefixes name (the binding of the prefix to the namespace region of the
schema) has dynamic extent. that is, within its physical entity and any entities
referenced from there.


> And an element
> type name without a prefix has no binding to a schema using the namespace
> mechanism (it still could use the standard XML markup declarations, or
> architectural forms, or ICADD fixed attributes, or other home-made systems
> though.)

then it has a binding. that it is implicit does not make it go away.

>
>
> 3)...
>
> With the wd-namespace PI the number of possible things that can clash
> is reduced to the prefixes themselves and to unqualified names.

with the correct scoping and default rules, it should be possible to eliminate
these clashes.

[does anyone have a denotational definition for xml wrt the dom? it would make a
discussion like this so much easier...]

bye for now,



xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)