RE: Data models?

The model is a grammar that generates data structures as instances of that
model.

The web is a graph of resources connected by URI-references embedded in the
resources.

The model only knows from resources, URIs and URI-references.

The structure is all the growing expanse of site-full Internet.

Not the same thing.

And if you look carefully at Gregory's proposed changes to 1.5, you will see
the same distinction.  The model means you can recognize structure in the
data.  A data structure says you present the data in the structure.

In HTML tables, there is a tree structure in which the data is presented,
and a
matrix structure in which it is to be interpreted.  Maybe that's a clearer
example.  Columns are implicit structure in HTML tables, they are not part of
the data structure until you know the model.

Al

At 04:34 PM 2001-08-24 , you wrote:
>> > So do we have any examples of instances where structure is encoded in
>> > something other than markup?
>>
>> PDF is such an example. The structure is encoded in the logical structure
>> tree, an internal data structure in the PDF file.
>
>So should we use "data structure" instead of "data model"? Thus:
>
>1. the hierarchical structure of the content is represented unambiguously in
>the markup or data structure.
>
>
>Or is this redundant?
>
>It seems to me that what we are getting at here is this:
>
>1. The content has some intrinsic structure to it.
>
>2. This structure must be reflected in the structure applied to it, either
>through markup or some other process.
>
>So in the example above, "the hierarchicial structure of the content" refers
>to the intrinsic structure of the content, and "the markup or data
>model/structure" refers to a structure imposed on the content from outside.
>And we are saying that these must match.
>
>Am I getting warmer?
>
>Chas. Munat
>

Received on Friday, 24 August 2001 16:46:17 UTC