Re: [xml-dev] Are namespaces actually crypto-entities orcrypto-links? (w

[Marcus Reichardt <u123724@gmail.com>]

Not familiar with your proposal, do you mind sharing a link and the
reaction you described? Admittedly, I personally can't see the point
of namespaces that are automatically placed. AFAICS, they're
equivalent to opaque "xmlns:svg" and "xmlns:mathml" attributes
defaulted to a #FIXED declared value I guess, as in

    <!ATTLIST (svg|g|...) xmlns:svg CDATA #FIXED "http://www.w3.org/2000/svg";>

(assuming SGML and naming rules allowing the colon char as NAMECHAR)
but for what purpose when HTML5 doesn't do namespaces anyway, or only
to the extent necessary to support legacy apps such as in DOM
getElementByTagNameNS(), getAttributeNS() methods, or in the context
of legacy XLink attributes?

From <https://html.spec.whatwg.org/multipage/syntax.html#elements-2>:

> The HTML syntax does not support namespace declarations, even in foreign elements.
>
> For instance, consider the following HTML fragment:
>
> <p>
> <svg>
>  <metadata>
>  <!-- this is invalid -->
>   <cdr:license xmlns:cdr="https://www.example.com/cdr/metadata"; name="MIT"/>
>  </metadata>
> </svg>
> </p>
>

> The architectural forms approach of mapping terms in an SGML
> vocabulary to some other vocabulary, used in both Author/Editor and
> SoftQuad Panorama, was much better than the Xlink-style nonsense of
> foisting fixed namespaced names on people - but that argument got
> nowhere and went round in circles trying to get nowhere, unfortunately.

That was kindof the point I was trying to make elsewhere ;)

Only that I think a richer mapping at least involving specific target
attributes rather than mere renaming of elements and attributes would
typically be more useful IMO.