W3C home > Mailing lists > Public > public-wai-ert@w3.org > June 2005

Verbose reporting

From: Charles McCathieNevile <chaals@opera.com>
Date: Thu, 02 Jun 2005 18:04:50 +0200
To: "public-wai-ert@w3.org" <public-wai-ert@w3.org>
Message-ID: <op.srq2yccrwxe0ny@widsith.local>

It seems like Sidar are in the middle. Hera collects information at a  
checkpoint level, although in some cases there are several tests done per  
checkpoint. The goal is to collect more detailed information, for the kind  
of use cases Shadi described.

One of the important questions, then, is whether the following is  
reasonable

<e:Assertion r:about="#passed">
   <e:testSubject r:resource="#subject"/>
   <e:result r:parseType="Resource">
     <r:type r:resource="&ns;pass"/>
     <d:description r:parseType="Literal">
     <p xmlns="http://www.w3.org/1999/xhtml"><a  
href="http://example.org/tests">Automated Test Cases</a> 1, 2, 4, 5, 9, 10  
all passed. No regression problems.</p>
   <e:testCase r:resource="&tests;1" />
   <e:testCase r:resource="&tests;2" />
   <e:testCase r:resource="&tests;4" />
   <e:testCase r:resource="&tests;5" />
   <e:testCase r:resource="&tests;9" />
   <e:testCase r:resource="&tests;10" />
   <e:mode r:resource="&ns;automatic" />
</e:Assertion>

<r:Bag r:ID="subject">
   <r:li r:resource="&subj;A" />
   <r:li r:resource="&subj;B" />
   <r:li r:resource="&subj;C" />
   <r:li r:resource="&subj;D" />
   <r:li r:resource="&subj;E" />
   <r:li r:resource="&subj;F" />
   <r:li r:resource="&subj;G" />
</r:Bag>

This is already a lot less verbose than what Hera produces. Compress it  
and you have quite a small amount of data (although clearly non-zero). It  
is also possible to make an aggregate result, such as "no regression  
properties", or "WCAG 1.0 double-A conformance" and say

<e:Assertion r:about="#compact">
   <e:testSubject r:resource="#subject"/>
   <e:result r:type="&ns;pass" />
   <e:testCase r:resource="&tests;nr" />
   <e:mode r:resource="&ns;automatic" />
   <s:seeAlso r:resource="#passed" />
</e:Assertion>

Do these approaches break anything?

I think you can have multiple subjects and tests pretty meaningfully. I  
don't think you want to have multiple results in teh general case (I will  
write about a specific case where you may want to do that in another  
context) and I don't think it makes sense to have multiple modes mixed  
together.

cheers

Chaals

-- 
Charles McCathieNevile                              chaals@opera.com
          hablo español - je parle français - jeg lærer norsk
   Here's one we prepared earlier:   http://www.opera.com/download
Received on Thursday, 2 June 2005 16:05:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:18:25 GMT