W3C home > Mailing lists > Public > public-qa-dev@w3.org > March 2007

Re: Bug in css validator?

From: olivier Thereaux <ot@w3.org>
Date: Wed, 7 Mar 2007 12:43:12 +0900
Message-Id: <811EC32F-DECA-418E-B61F-70BCDA447A03@w3.org>
Cc: CSS validator list <www-validator-css@w3.org>, QA Dev <public-qa-dev@w3.org>
To: Salvatore Tedesco <wavesun69@yahoo.it>

Hello Salvatore, all

On Mar 7, 2007, at 03:19 , Salvatore Tedesco wrote:
> Doesnt works - CSS validation (by link): http://jigsaw.w3.org/css- 
> validator/validator?uri=http%3A%2F%2Fwww.giuseppequaro.com% 
> 2Fhome.asp&warning=1&profile=none&usermedium=all

I've seen a few examples like this, where the XML parser in the CSS  
validator seems to find errors from outer space.
In this case, the fact that your ASP engine apparently serves  
different pages depending on the user-agent doesn't make debugging  
easy, but:

* I have looked at it with the CSS validator on another machine and  
it passes validation. No problem.

* when loading http://jigsaw.w3.org/css-validator/validator?uri=http% 
for the first time the validation result is positive (lots of  
warnings, but passes validation), but if reloading, then I get the  
XML parsing error
Line 41, Column 211
The value of attribute "alt" associated with an element type "img"  
must not contain the '<' character.
re-reloading gives the same error, unless you let time pass and try  
again. That's consistent with other recent bug reports, see

* retrieving the page with the CSS validator's user agent gives me  
the following line 41

<td class="td010101" rowspan="2"><img src="images/graph/ 
header1_460x250.jpg" width="460" height="250" title="Giuseppe Quaro -  
Official Web Site - Home Page" alt="Giuseppe Quaro - Official Web  
Site - Home Page"/></td>

The character on Column 211 is the first "-". Why does the parser see  
a "<" there, beats me.

I don't know if this is related to the broken pipe errors reported  
recently too, or if it is a bug in the XML parser shipped in the JDK,  
but I'm strongly suspecting something rather broken in the jdk 1.6,  
since the odd issues seem to have started when we switched. Should  
(could?) we switch back to 1.5 for the moment and test 1.6 on a  
different machine?

Received on Wednesday, 7 March 2007 03:43:23 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 16:54:51 UTC