Re: The only name for the xml serialisation of html5

On Wed, 31 Oct 2007, Dan Connolly wrote:
> I'm interested to know what the editors (Ian, Dave) and other WG members 
> think of those arguments.

I discussed this with Hyatt just now. Our opinion is that the XHTML name 
isn't really that important, since we don't really expect the XML 
serialisation will be used much. Since the XML serialisation of HTML up to 
this point has been called XHTML, we think it makes sense to continue this 
naming scheme. We also think it makes sense to synchronise the numbering 
schemes of the HTML, XML, and DOM serialisations, thus leading to HTML5, 
XHTML5, and DOM5 HTML, as the draft says.

Further, since the name has been used for several years now there is a 
strong incentive for us to keep calling it "XHTML5" as we have been doing. 
In particular, Windows Live Search, Yahoo! Search, and Google all agree 
that the term "XHTML5" refers to the HTML5 spec, and Wikipedia even 
mentions XHTML5 as being the XML version of HTML5. Indeed, both the MIME 
type for the XML serialisation and its namespace have the term "xhtml" in 
them, strongly indicating that the language is XHTML.

We understand that there are various arguments that have been put forward 
to recommend against use of the term XHTML. One argument is that the 
"XHTML" name is intrinsically linked to the concept of modularisation. We 
disagree with this premise; XHTML1.0 (the only widely used version of 
XHTML) was no more modular than HTML5's XML serialisation. Even if it were 
so, though, we do not think that most authors would find the details of 
the language's schema construction to be of relevance in picking its name. 
Another argument is that HTML5 doesn't need a name for its XML 
serialisation. In practice we have found that actually we frequently refer 
to the two serialisation, and a name is thus very useful (and people 
naturally call the XML serialisation XHTML).

Regarding the possible confusion with XHTML2, we had thought that this 
confusion would be diffused by the rebranding effort that the W3C 
announced in March of this year [1]. We would point out that the confusion 
between XHTML2 and the original XHTML language (that is, the HTML4 
vocabulary cast as XML) existed before XHTML5 was even considered, so it 
isn't really an XHTML5 issue. We look forward to seeing the results of 
this rebranding effort.

In conclusion, since this is such an unimportant part of the 
specification, and since the arguments for changing the name are so minor, 
we do not feel that there is much reason to change the name.


Ian Hickson, for himself and David Hyatt.

Received on Wednesday, 31 October 2007 08:26:40 UTC