W3C home > Mailing lists > Public > public-bpwg-pro@w3.org > March 2008

Re: Draft text on sampling and test scope

From: Alan Chuter <achuter@technosite.es>
Date: Wed, 26 Mar 2008 14:13:14 +0100
Message-ID: <79cab0900803260613k398c8ee3t76e965f71c28dba5@mail.gmail.com>
To: "Mobile Web BP Pro list" <public-bpwg-pro@w3.org>

The Basic Tests don't mention sampling. It seems to be implicit that
100% of resources should be tested. With non-automatic tests this is
not practicable. also, the device space makes this impossible. We
can't test all pages on all devices, so a sampling scheme is needed.

What other dimensions are relevant to sampling? Other than URIs and devices.

For URI scope, I suggest using the UWEM sampling scheme (this much is
already in the draft). But something is needed about what devices to
use, if any other than the DDC. This may need some qualification as it
requires the inclusion of pages using frames and tables, and the
accessibility information page which are perhaps inappropriate. On the
other hand, if frames are used they should be included, leading to a
fail.

Here is a draft text: "The scope of the evaluation must be defined and
included with the conformance claim. All content within the scope must
pass the MobileOK Basic Tests. For the manual tests, a sampling scheme
must be used and documented with any conformance claim. The sample may
include all content or only a representative sample. The sampling
scheme specified by UWEM 1.2 (http://www.wabcluster.org/uwem1_2/)
should be used."

If we can't reference a document outside of W3C space, we can adapt
the UWEM scheme and include it.






On 26/03/2008, Alan Chuter <achuter@technosite.es> wrote:
> This is a draft text for this as I took the action to write it. It is
>  currently under "Sampling Scheme" [2] but should be separated off,
>  with a title like "Test Scope". "Sampling Scheme" is still necessary I
>  think, but is different. Here's the text:
>
>  "While most tests apply to single pages or delivery units [1], some
>  tests, such as ACCESS_KEYS, NAVIGATION and URIS are tested across
>  multiple pages. To perform the tests correctly it is necessary to
>  define the scope of the evaluation. This may be expressed using URI
>  patterns [or groupings using POWDER]."
>
>  [1] http://www.w3.org/TR/di-gloss/#def-delivery-unit
>  [2] http://www.w3.org/2005/MWI/BPWG/Group/TaskForces/mobileOKPro/drafts/ED-mobileOK-pro10-tests-20080228#sampling
>
>
>
>  --
>  Alan Chuter,
>  Senior Web Accessibility Consultant, Technosite (www.technosite.es)
>  Researcher, Inredis Project (www.inredis.es/)
>  Email: achuter@technosite.es
>  Alternative email: achuter.technosite@yahoo.com
>  Blogs: www.blogger.com/profile/09119760634682340619
>


-- 
Alan Chuter,
Senior Web Accessibility Consultant, Technosite (www.technosite.es)
Researcher, Inredis Project (www.inredis.es/)
Email: achuter@technosite.es
Alternative email: achuter.technosite@yahoo.com
Blogs: www.blogger.com/profile/09119760634682340619
Received on Wednesday, 26 March 2008 13:13:54 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 26 March 2008 13:13:54 GMT