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

(unknown charset) Re: [VE][247] Add Subject Here

From: (unknown charset) Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
Date: Thu, 17 Nov 2011 19:44:18 +0100
To: (unknown charset) "Jukka K. Korpela" <jkorpela@cs.tut.fi>
Cc: (unknown charset) www-validator@w3.org
Message-ID: <20111117194418138242.fc4de976@xn--mlform-iua.no>
Jukka K. Korpela, Thu, 17 Nov 2011 10:38:16 +0200:
> 2011-11-17 7:28, Leif Halvard Silli wrote:

I will post a reworked proposal based on your comments, in a separate 
letter.

>> you know. Hence the current text says 'sequence'.
> 
> OK. Or maybe 'construct' would be suitable?

'string' or perhaps 'code' might be better.

>>> but should not be used in HTML (as opposite to XHTML).
>> 
>> Please replace 'HTML' with 'HTML4', and please say 'XHTML and HTML5'.
> 
> Please may use the validator to check, say, HTML 3.2 documents, or 
> documents with a custom DOCTYPE, which need not have anything to do 
> with HTML 4. And "HTML (as opposite to XHTML)" includes HTML5 in HTML 
> serialization and excludes HTML5 in XHTML serialization (or syntax).

HTML5 does permit both <hr /> and <hr> inside the text/html version of 
the syntax.
 
>>> The validator treats it very differently
>>> from what you expect, and this causes many confusing error messages.
>> 
>> This seems like the most important point. May be it should say 'treats
>> it according to SGML rules, which in this particular case is quite
>> different from what most developers/authors expect, and this causes
>> confusing error messages.'
> 
> Sounds perfect to me.

Good.
 
>>> Remove the '/' character."
>> 
>> Here I would have written: In order to make the best use of the
>> validator's HTML4 validation abilities, you should remove the '/'
>> character, so as to not be disturbed by confusing error messages.
> 
> This has nothing particular to do with HTML 4 as opposite to other 
> SGML-based versions of HTML. The message should be adequate, or at 
> least non-misleading, even for a pure SGML document that has nothing 
> to do with HTML, since the validator is capable of handling them 
> (with restrictions), e.g.
> http://www.cs.tut.fi/~jkorpela/test/sgml.html

I'll try to incorporate this point in my proposal.

> I admit that the statement "Remove the '/' character" may sound rude 
> and too direct, but it's really the most practical advice here (and I 
> would put it first, really).
> 
> Perhaps it could be smoothened (and made more accurate) as follows:
> 
> "You should either remove the '/' character or convert the document 
> as a whole to XHTML (or XML)."

I will try to say something like 'For a more common interpretation of 
<foo/>, please use an XHTML1 or HTML5 doctype'.
-- 
Leif H Silli
Received on Thursday, 17 November 2011 18:44:54 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 25 April 2012 12:14:49 GMT