[rfc-i] draft-hildebrand-html-rfc

Julian Reschke julian.reschke at gmx.de
Tue Jul 24 07:45:23 PDT 2012

On 2012-07-24 16:19, Joe Hildebrand (jhildebr) wrote:
> On 7/24/12 12:55 AM, "Julian Reschke" <julian.reschke at gmx.de> wrote:
>>> Where exactly is it not predefined? HTML4 includes a full set of
>>> character entities for Latin-1, see
>>> http://www.w3.org/TR/html4/sgml/entities.html. I couldn't imagine that
>>> some of these were abandoned in HTML5. XML is a different story of
>>> course, but here we are talking about HTML, no?
> http://goo.gl/VDCD6
> Contains &nbsp;

Yes, but that only applies to the HTML serialization, not the XML 

>> No. As far as I can tell, we're discussing XHTML.
> &#xa0; always works, so we don't have to close the loop on HTML dialect
> yet.

Yes, of course.

> That said, I'm going to argue strongly for HTML, with some slight
> modifications to preferred serialization to increase the number of tools
> that can be used.  The biggest reason is going to be marketing, not
> technical - and I expect that to upset some folks, for which I apologize
> in advance.
> The IETF brand has been ravaged by decades of obstinate conservatism in
> the look of our primary output.  To the next generation of potential
> protocol developers, this not only calls the quality of our technical work
> into question, but also is an indicator that they're not welcome into our
> old-boys club of cryptic tooling.
> In the world outside the IETF, HTML5 is non-controversial - in fact it's
> assumed as a starting point for a lot of new development.  The fact that
> it's going to keep evolving over time is seen as a good thing.  I believe
> to combat the image problem that we have we should take a small amount of
> risk in order to move to where the world is going, rather than once again
> picking a decade old technology because we understand it slightly better.

Well, XHTML5 is HTML5, too, just a different serialization.

That being said: it doesn't really matter. Conforming HTML5 can easily 
be transformed into XHTML5. But if we do that, we'll need to change some 
of the rules currently in the draft, for instance with respect to 
closing elements (for some elements in HTML, such as <link>, it's not 
even optional).

Best regards, Julian

More information about the rfc-interest mailing list