- From: Dr. Olaf Hoffmann <Dr.O.Hoffmann@gmx.de>
- Date: Tue, 19 Jan 2010 20:41:55 +0100
- To: public-html@w3.org
Leif Halvard Silli: > > It strikes me that this is an example of a language which we could have > developed within HTML5, if Toby's Decentralized Extensibility proposal > had been implemented in HTML5. > > http://www.w3.org/mid/1263511821.18556.5.camel@ophelia2.g5n.co.uk > > Then we would not need develop new elements. How would you have looked > upon such a solution? All the stuff which relates to RDFa and @role > could have been added to those elements instead. > > OTOH, we have - as you have noted - XHTML, which allows you to use > validly use specific elements. And, we can serve XHTML as text/HTML. > So, with HTML5 we get two kinds of text/HTML: "real" text/HTML and > XHTML text/HTML. The latter gives us much more freedom than HTML5. What do you think, what I'm currently doing? ;o) I retired most old text/html stuff already and move to XHTML+RDFa and SVG: What you can currently do is to use XHTML+RDFa http://www.w3.org/TR/rdfa-syntax/ served as application/xhtml+xml (and as text/html tag soup only to outdated browsers without XHTML capabilities, they do not care either about this stuff, therefore not important, that this is the wrong type). Or alternatively SVG tiny 1.2 (what is not worse than (X)HTML concerning semantics for poetry ;o) And referencing a vocabulary like LML or simply refer to the few poetry elements from DAISY as a minimal solution for many/most applications. The main advantage by using text elements from SVG is, that they have no semantical meaning, only define a graphical representation with text alternative, title, description, metadata for any element. Therefore it is simple to add the semantics with the role attribute for example without abusing any elements for anything and to add every meta information, what is considered to be important at the place, where it belongs without any daring feats. And because HTML5 notes, that you can embed SVG, this markup is applicable just in case, that everything else should fail with HTML5. (well, for the text/html variant there seems still the problem with the not existing namespaces - but if a user-agent really wants to interprete such semantical information there will be a way if it enters the SVG area - who cares about nasty details now? It needs to do this anyway for all this RDF stuff indicating CC or GNU licenses often present in SVG nowadays ;o) Olaf
Received on Tuesday, 19 January 2010 20:20:08 UTC