Thanks Mike. I guess I would prefer verbose and have them fill the form out once than have them have them misinterpret and have to fix errors, which I imagine can be tedious using a screen reader. Is this the case?
It would be great if you could keep us posted about any feedback you get in March when the site goes live. For the average user however, what I think I will do is run a few simple A\B tests on some of my sites and log the amount of JavaScript errors for each of the different methods described (there seems to be at least three plausible options). It will take some time to get statistical significance however so it might be a while before I have something useful. Cheers, Matt On 2/26/08, Mike at Green-Beast.com <[EMAIL PROTECTED]> wrote: > Hi Matt, > > > > that the following legend is > > superflous and prevents logical grouping. > > > > >> <fieldset> > >> <legend>Required</legend> > >> <label for="name">Name (required) <input name="name"> > >> <label for="email">Email (required) <input name="email"> > >> </fieldset> > > > I agree, actually. With that example (and the image one I gave) using the > word required, in the case of a user listening with a setting that reads the > legends (default), would make it too verbose. It'd read: > > Required Name Required > Required Email Required > > Though I guess there'd be no missing it. ;-) > > The use of the Required legend seems to work well with the asterisk, with > its meaning defined in a non-associated label (one with no for attribute). > It's a compromise method. I do have one form on a real-deal AAA WCAG 2.0 > site I made (to be officially announced Mar. 11-12th) with this specific > configuration. It's open now by invite to a few disabled users/testers and a > couple of key WCAG 2.0 Editors, and I got more very positive comments about > that particular set-up tonight... a few minutes ago actually. > > > <fieldset> > > <legend>Personal Details</legend> > > > > > <label for="name">Name <span class="required">(required)</span></label> > > <input name="name"> > > > That is a solid method for sure, but there's only one problem and that is to > *some* users (default settings) it might sound too verbose. > > Personal Details Name Required > Personal Details Email Required > Personal Details Phone > > The problem is not the technique, yours or mine, or any of the other > accessible methods. It's the myriad configurations possible that really > challenge us. There are so many variables (not even including those of > sighted users) that while there are a number of feasible methods, there > seems no perfect one-size fits-all answer. It's all a compromise. > > > Mike > > > > > ******************************************************************* > List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm > Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm > Help: [EMAIL PROTECTED] > ******************************************************************* > > ******************************************************************* List Guidelines: http://webstandardsgroup.org/mail/guidelines.cfm Unsubscribe: http://webstandardsgroup.org/join/unsubscribe.cfm Help: [EMAIL PROTECTED] *******************************************************************
