W3C home > Mailing lists > Public > public-html@w3.org > September 2007

Re: The only name for the xml serialisation of html5

From: Leif Halvard Silli <lhs@malform.no>
Date: Sun, 30 Sep 2007 16:18:55 +0200
Message-ID: <589c5bb396ed280ac9e1e890e8392983@10013.local>
To: Dean Edridge <dean@55.co.nz>
Cc: Adam Nash <adamn@wirespring.com>, public-html@w3.org

2007-09-30 06:06:27 +0200 Dean Edridge <dean@55.co.nz>:
> Leif Halvard Silli wrote:

>> For compatibility with the Web, we have to serve XHTML as text/html - at 
>> least to some UAs.
> Arrrr..umm..., well, not really, when you attempt such a practise your site 
> becomes invalid HTML  [2]. [ ... ]

Indeed. This incompatibility is being hidden for the users - and page developers - though.  For example the W3 Valdiator will by default deem Wordpress.org in accordance with the code syntax - which is valid XHTML 1.1! Only when you force it to be validated as served, namely as HMTL (of some sort ...), is it [deemed invalid].

I want HTML 5 format names that make this very clear!

>> HTML 5 is meant to put an end to that hack.
> I don't quite know what you're getting at there. That's not what HTML5 is all 
> about really. HTML5 is primarily HTML "text/html" [...]

Indeeded, «HTML5 is primarily HTML "text/html"». That is exacty what I wanted to convey. 

> Perhaps you were referring to the fact that HTML5 exposes how silly it is to 
> send a "so called" XHTML document over the wire with a "text/html" mime type 
> whilst thinking it is valid XHTML, and makes it quite clear that XHTML has a 
> mime type of  "application/xhtml+xml" or "application/xml".

Here you are close to interpreting me correctly. However, e.g. Wordpress.org is not a «so called" XHTML document». It _is_ a valid XHTML document - according to its syntax. The problem is that judged against how it is being served, it isn't a valid _text/html_ document.

> To use XHTML 
> according to the HTML5 spec you simply use a XML mime type(as above), the 
> actual markup is identical for most web pages.

Technical how-tos, aside, the question should be: What would be the most helpful name?  I think the answer is that the names of the two HTML 5 formats should help convey two important things, namely that a mark-up document must be

1. served according to the format it has. 
2. coded with a syntax that is in accordance with the format it is going to be served as.

Unfortunately, as I demonstrated with Wordpress.org, the 'XHTML' name is not able to help make this clear. And adding '5' to 'XHTML' would not make that any clearer  - but instead simply add other unclarities. Yes, the XHTML name has lead people to code differently. But not to serve accordingly.

Therefore, the logical thing would be to name the two formats according to how those formats should be consumed/served. That would help us draw the necessary connection between coding syntax and the consumption/serving.

Hence my proposal: For the SGML-inspired custom format, 'HTML 5 text' or 'HTML 5 html' - where I like 'HTML 5 text' the best. For the XML based format, either 'HTML 5 application' or 'HTML 5 xml' - where perhaps 'HTML 5 xml' is most likely to go home - allthough 'HTML 5 application' is fine with me. I would not be against naming the two formats  'HTML 5 html' versus 'HTML 5 xhtml' either, actually. And perhaps that is what you would find most logical of these variants?

[deemed invalid]: 
<http://validator.w3.org/check?uri=http%3A%2F%2Fwordpress.org%2F&charset=%28detect+automatically%29&doctype=HTML+4.01+Transitional&group=0>
-- 
leif halvard silli
Received on Sunday, 30 September 2007 14:19:14 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:08 GMT