Re: The illusion of simplicity and low cost in data design and computing

I don't recall a way to say that some element's content uses a notation, in
XML. The scope of NOTATION in XML relates to external non-XML entities.
There is no semantic that says, unless I have forgotten, that if an element
has an attribute which specifies a notation, it applies to the content of
the element (rather than to some link or reference.)

Even in XML Schemas, the datatype type NOTATION has the caveat "For
compatibility (see Terminology (§1.4)
<https://www.w3.org/TR/xmlschema-2/#terminology>) *NOTATION* should be used
only on attributes and should only be used in schemas with no target
namespace." Which makes it clear that it is a sunset feature limited to
replicating what DTDs provided.

In any case, even embedded NOTATIONS were supported, you need to look in a
schema or markup declarations so it is not inline. (By embedded I mean the
content is in an element, by inline I mean that the the markup to say what
the notation is at the same location.)

XML's problem as a syntax is that it does not make enough use of delimiters
to signify things inline, locally. This forces unnecessary resort use of
extra layers.

For example consider the attribute Eq "=".  It is a useless thing in a
position that could carry lots of meaning. Instead you could use that
delimiter position to indicate that an attribute was some kind of ID or
anchor (say with "=#") or some kind of IDREF or target (say with "#=")
(which I have in my RAN experiment).

So lets imagine some future XML with some delimiter convention to say
whether the attribute applies to the current element orkto its value (say
"=" and "=="). That would provide an inline way for a system to know that
there is an embedded notation rather than, say, a link.

(Now in fact there is a syntax for inlined embedded notations as element
content: you could use a PI instead of the text node. But that prevents XML
annotation of portions of the embedded notation, so cripples a key value of
markup languages, that you can tag things. And for things like XSLT, a PI
is not part of the value of an element, so rather than being a mechanism
for labelling data content with a notation, it removes it.)

Cheers
Rick






On Mon, 15 Aug 2022, 2:04 am C. M. Sperberg-McQueen, <
[email protected]> wrote:

>
> Rick Jelliffe <[email protected]> writes:
>
> > XML provided the first (headers, doctyoe, namespaces, etc) but kinda
> > fails to provide the second (standard inline conventions to introduce
> > embedded notations). Which is why it struggles whenever it has to
> > contain non-xml data inside: hence see ludicrous approaches like
> > requiring an external schema to say some element value is bin64.
>
> Isn't that what NOTATION declarations are designed to do (and in fact
> do, when people use them)?
>
> --
> C. M. Sperberg-McQueen
> Black Mesa Technologies LLC
> http://blackmesatech.com
>