W3C home > Mailing lists > Public > www-validator@w3.org > August 2013

Re: [VE][html5] Add Subject Here

From: Jukka K. Korpela <jkorpela@cs.tut.fi>
Date: Wed, 21 Aug 2013 08:48:45 +0300
Message-ID: <521454BD.8030904@cs.tut.fi>
To: Windra Asrah <windra.asrah@gmail.com>
CC: www-validator@w3.org
2013-08-12 8:07, Windra Asrah wrote:

> Validating http://windraseptamadya.blogspot.com/
> Error [html5]: ""

The validator now says "Passed, 1 warning(s)". Did you fix the issue?

Unfortunately, you message spent about a week on a W3C server before 
getting distributed to the recipients of the mailing lists. I'm afraid 
the list is losing much of its usefulness unless the problems causing 
such delays (which are not uncommon, though a delay this long is rare) 
will be fixed.

> & did not start a character reference. (& probably should have been
> escaped as &amp;.)

This is a longstanding issue. The authors of the specifications and the 
authors of the validator seem to disagree on whose rules are to be 
applied. Please remember that the validator is experimental software 
that checks against an mutable draft, presumably some relatively recent 
version of HTML 5.1 Nightly.

> But the wrong is in my href, not in the inner of this HTML

By HTML rules, character references are interpreted in href attribute 
values as elsewhere. The dispute is what happens when something starts 
like a character reference but does not constitue one.

> Allowed and valid:
> <a>&amp;</a>

Yes, in all versions of HTML.

> Allowed, but invalid:
> <a>&</a>

Allowed and valid in all versions of HTML except those using XHTML 
syntax, such as HTML5 in XHTML serialization ("XHTML5").

> Allowed and valid:
> <a href="abc.com/a&b <http://abc.com/a&b>">&amp;</a>

I'm not sure what you mean by that, but href="a&b" would create the 
disputed situation in HTML5. Some people say it's OK. Some say it's not. 
In any case, browsers deal with it without problems. And both browsers 
and validators deal with href="a&amp;b" without problem. The issue is 
whether authors should be "forced" to use the latter in order to pass 
validation.

> Suggestion:
> in href and src, no &amp;, but &.

Validators cannot (or should not) decide on such matters. They are 
expected to check against rules set elsewhere, not (re)define the rules. 
In this case, it is practically impossible to have the rules changed, 
because character references have always been allowed in attribute 
values, and disallowing &amp; would invalidate *illions of existing 
pages that work on all browsers (as far as this feature is considered).

Yucca
Received on Wednesday, 21 August 2013 05:49:13 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 14:18:09 UTC