Re: <q>

2008/10/25 Preston L. Bannister <preston@bannister.us>

> Yep - knew someone was bound to miss the point.


I'm still missing your point. I'll explain why.


> In the usage I mentioned the semantics of <q> is exactly defined within the
> context between the server and a specific web application's client-side
> Javascript.


I think this would be true if your Javascript isn't operating on HTML. If it
is operating on HTML, then the meaning of the <q> element within that HTML
is defined by the HTML spec.


> This is in effect an entirely distinct namespace unrelated to the "Semantic
> Web" (whichever that is).


I don't see what this discussion has to do with the Semantic Web per se.
Admittedly, some HTML elements - <q> included - have associated semantics,
but normally the phrase "Semantic Web" refers to a web of documents using
RDF/OWL/etc to encode user-defined semantics or other semantics (e.g. Dublin
Core) not defined in the specification document of the (meta-)language (XML,
RDF, etc) itself. At least, that is my understanding.


> HTML through web browsers is and always will be (for the foreseeable
> future) a delivery mechanism for applications. Web application developers
> should and will make pragmatic use of the strongest capabilities of the
> platform, independent of dogmatic concerns.
>

I don't dispute this as such. I do dispute that using <q> the way you have
done is pragmatic. It will reduce the interoperability of your code (I pity
anyone trying to screen-scrape your sites), which is not pragmatic IMO.


> Your point would be valid if the semantics were meant to be global. That is
> not the case.


The semantics defined in the HTML spec are meant to be global across HTML
documents.


> To pick apart your analogy, yes - English is meaningless (not useful) if
> all of your audience speaks Spanish, only some speak English, and you want
> the entire audience to understand what you say.


In the case of <q>, all the latest major browsers (IE, FF, Safari, Opera)
now implement it. (To continue the analogy, they do, now, all speak
English.)


> Web application developers need to be practical, not dogmatic.


Given that all the latest major browsers do now support <q> - and given that
there are both block (<div>) and inline (<span>) elements designed for
delimiting content for author-defined styling, scripting, etc - I don't
think it's remotely pragmatic to use <q> for anything other than marking up
quotes.

Regards,

Sam