- From: Charles McCathieNevile <charles@sidar.org>
- Date: Tue, 08 Feb 2005 01:43:49 +0100
- To: Kurt_Mattes@bankone.com, w3c-wai-ig@w3.org
I think that the cost of testing here is overestimated - without going into specifics (which can obviously vary by orders of magnitude) I would suggest that with reasonable expertise tens of hours would be mor like it. (Note that developing expertise takes time and money, too). An important point that Kurt hasn't noted explicitly is that what is important is to examine how sites are actually created in practice, looking for ways that errors may creep in unnoticed, or may be picked up by the "background QA" that everyone in an organisation does as they work... Testing is difficult. But expertise allows testers to do it relatively rapidly, and effectively, where people who have just started learning to do it will be slower and pick up less - although this is a gross generalisation. *Good* testing tools can also help substantially, although they cannot yet (in my opinion) allow a total novice to do the job of an expert. They can also be very valuable in turning a novice into an expert as fast and cheaply as possible. Cheers Chaals On Mon, 7 Feb 2005 11:47:45 -0500, <Kurt_Mattes@bankone.com> wrote: > > Perhaps there is a lack of large site experience on this list. > Testing a 'template based' bank site on two platforms with two > browsers for each platform requires thousands of QA hours. > Each additional browser/platform combination adds hundreds of > additional testing hours. It is nice to think the cost of > accessibility is low, unfortunately reality proves that for large > dynamic sites it is considerable. Without testing, how do I > know it degrades gracefully? When it can be said that a site is > accessible after successfully testing one or only a few pages built > with each template in the sites design, it can be said that testing > isn't that difficult. -- Charles McCathieNevile - Vice Presidente - Fundacion Sidar charles@sidar.org http://www.sidar.org (chaals is available for consulting at the moment)
Received on Tuesday, 8 February 2005 00:51:11 UTC