W3C home > Mailing lists > Public > public-wai-ert@w3.org > October 2006

Re: validity levels

From: Johannes Koch <johannes.koch@fit.fraunhofer.de>
Date: Tue, 24 Oct 2006 12:21:15 +0200
Message-ID: <453DE91B.2050705@fit.fraunhofer.de>
To: public-wai-ert@w3.org

Carlos A Velasco schrieb:
> Shadi Abou-Zahra wrote:
>> If the test is "does the Java unit compile" or "is the Java
>> code valid" then these warnings you describe *are* subclasses of 'pass'.
>> What are you testing for?
> 
> If you take, let us say, a c/c++ compiler with different optimization
> levels, warnings could be a subclass of "pass" or "fail":

or cannotTell

> it is not
> clear to me which one to take.

IMHO compiler warnings are a pass of the "but you should do better" kind.

> Going back to Johannes example with the CSS validator: you can see this
> warning:
> 
> # used by org.w3c.css.values.CssColor
> warning.out-of-range: %s is out of range
> 
> for a color value out of range. Although it may be interpreted as a
> "pass", it could also be "fail". I am in favour of having warning as an
> option.

I am with Shadi for not having a top-level validity level "warning". I 
don't know, when this message is created. But if the problem makes the 
CSS invalid, it should be a (subclass of) fail, if it does not affect 
the validity, it should be a (subclass of) pass. If you can't make a 
pass/fail statement, it should be a (subclass of) cannotTell.
-- 
Johannes Koch - Competence Center BIKA
Fraunhofer Institute for Applied Information Technology (FIT.LIFE)
Schloss Birlinghoven, D-53757 Sankt Augustin, Germany
Phone: +49-2241-142628
Received on Tuesday, 24 October 2006 10:21:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:18:27 GMT