W3C home > Mailing lists > Public > www-validator@w3.org > January 2011

Re: Bullet entity undeclared in HTML5 Meta description

From: Jukka K. Korpela <jkorpela@cs.tut.fi>
Date: Mon, 17 Jan 2011 00:14:21 +0200
Message-ID: <77B52806DAC94057A08C261CF733749C@JukanPC>
To: "Greg Lauver" <GregLauver@msn.com>, <www-validator@w3.org>
Greg Lauver wrote:

> Character entities in an HTML5 meta description tag: ampersand (&) is
> OK. Bullet () and other entities generate an error:
>
> Validation Output: 1 Error
> Line 5, Column 56: reference to undeclared general entity bull
> <meta name="description" content="This &amp; that &bull; Something
> else" />

This happens irrespective of context of use of entity references, when you 
use the so-called XHTML serialization of HTML5.

> <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">

This indicates you're using XHTML serialization. In it, you should not 
expect any entity references to be available except for those few that are 
predefined in XML (such as &amp;). See
http://dev.w3.org/html5/spec/Overview.html#writing-xhtml-documents
See also http://wiki.whatwg.org/wiki/HTML_vs._XHTML which contains loads of 
points to consider when using HTML5 as XHTML.

Apparently the experimental HTML5 mode of the W3C Markup Validator - a mode 
that is really more like a separate program performing special checking 
tailored for HTML5, rather than generic SGML or XML validation - treats 
entity references (except the XML-defined stuff) as undefined. This is 
somewhat drastic but understandable. The _good_ reasons for using XHTML (in 
general) normally mean that the data might be processed using a generic XML 
processor, which should not be assumed to know anything about the predefined 
entities in HTML5.

If you wish to keep using XHTML, you can replace the entity references by 
the characters themselves - after all, you are using UTF-8, so this is a 
matter of finding a suitable editor that lets you type any character. 
Alternatively, less legibly, you might replace the entity references by 
numeric character references like &#x2022;.

-- 
Yucca, http://www.cs.tut.fi/~jkorpela/ 
Received on Sunday, 16 January 2011 22:15:15 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 25 April 2012 12:14:45 GMT