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

Re: Linking 'location' and 'technique': models and examples

From: cstrobbe <Christophe.Strobbe@esat.kuleuven.be>
Date: Tue, 5 Dec 2006 12:54:43 +0100
Message-ID: <1165319683.45755e0360533@webmail2.kuleuven.be>
To: public-wai-ert-tsdtf@w3.org

Hi Shadi, All 

Quoting Shadi Abou-Zahra <shadi@w3.org>:
> Hi Carlos,
> 
> Using ID and IDREF attributes seems much more difficult to process.
> Also 
> more difficult to read and so more prone to bugs. Unless there is a
> case 
> for significant benefits, then I suggest we don't complicate things.
> 
> For this group our primary target is to describe the relationship of
> a 
> sample to the respective technique. So nesting (different types of) 
> location pointers under each technique makes most sense (option A).

That's actually option B ;-)

> However, since option A is quite restricted to the (current draft)
> WCAG 
> 2.0 model, we may choose to go for a less technique-oriented
> approach. 

The downside of a technique-oriented approach (option B) is that it 
does not make much sense when creating test samples that do not map to 
a technique documented in WCAG (we do that often in BenToWeb) or when 
creating test samples for a set of guidelines that does not link to 
techniques (I assume that's Shadi's point).

> Option B is probably more verbose since the technique would be
> repeated 
> in each (different type of) location pointer but I could live with
> it.

I assume this actually refers to option C ("techniques" element inside 
the "location" element). The repetition of the technique for each 
location is a clear downside. On the other hand, as Carlos Velasco 
pointed out, we could require that test samples are "atomic" in the 
sense that each sample contains only one occurrence of the failure or 
technique, hence avoiding repetition of the technique.

Any other thoughts on this?

Best regards,

Christophe

-- 
Christophe Strobbe
K.U.Leuven - Departement of Electrical Engineering - Research Group on 
Document Architectures
Kasteelpark Arenberg 10 - 3001 Leuven-Heverlee - BELGIUM
tel: +32 16 32 85 51
http://www.docarch.be/ 

Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm
Received on Tuesday, 5 December 2006 11:55:02 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:18:34 GMT