[thelist] Is this a list?

[Matt Warden]

Ian,

On 9/26/05, Ian Anderson <ian at zstudio.co.uk> wrote:
> The fieldset markup for checkboxes should read something like (forgive
> the shorthand for actual form item markup):
>
> <fieldset>
> <legend>Choose gift:</legend>
> <checkbox 1 id="cb1"><label for="c1">Cheese sandwich...</label>
> <checkbox 2 id="cb2"><label for="c2">Privacy...</label>
> ...
> </fieldset>

Actually, accessibility guidelines suggest that the form element
should be within the label element. i.e..:

<fieldset>
<legend>Choose gift:</legend>
<label for="c1"><checkbox 1 id="cb1">Cheese sandwich...</label>
<label for="c2"><checkbox 2 id="cb2">Privacy...</label>
...
</fieldset>

see e.g. http://www.w3.org/TR/WCAG10-HTML-TECHS/#forms-labels

This of course annoys me to no end, as I typically use <label> to push
my <input> elements over to the same horizontal column on the page,
like in the old table-based form layouts. (Although this is not really
an issue with checkboxes.)

> > Also, what's the general list opinion on JavaScript message boxes? I
> >  thought perhaps some sort of "Are you REALLY SURE you want to commit
> >  to The Fridge? Yes/No" alert would be quite handy, but if it's bad
> > practice, it'd be just as easy to have that as the next page rather
> > than as an JavaScript popup.
>
> Alerts, with their modal dialog box and system beep, are designed to
> alert the user to an error condition. They should not be used as you
> describe - I would come firmly off the fence and suggest this would be
> poor practice. The confirmation is part of the normal flow of the
> transaction and it should be a secondary page using form controls, not
> links.

That seems like a lot of bandwidth just for a confirmation.

> This prevents browsers and caching solutions that pre-fetch subsequent
> pages from inadvertently activating functionality, prevents search
> engines from spidering the next page, and allow you to
> support users who have disabled JavaScript in their browsers.

I don't see how it would have any effect on search engines or non-JS
users, except that they would not get the final opportunity to cancel
the operation.

--
Matt Warden
Miami University
Oxford, OH, USA
http://mattwarden.com


This email proudly and graciously contributes to entropy.