[whatwg] LABEL and radio/checkbox onclick

[Matthew Raymond]

Chris Kaminski wrote:
>>    This it just the nervous click plus enter key scenario 
>>again, and you have yet to produce a single individual this 
>>has actually happened to.
> 
> My S.O. She's done this. More than once.
> 
> She's a bit high-strung, and releases nervous energy by swirling the mouse
> around in circles on the desk while she thinks. From time to time, she
> clicks the button accidentally. It doesn't happen often, but it happens, and
> occasionally I'm called over to 'fix' her 'broken' computer because she
> clicked somewhere without realising it, sometimes with disastrous effect.
> 
> Sadly, she's not terribly anxious to have her name, age, social security
> number, national insurance number, passport number, weight, height, address,
> handy number and land line posted to a public list. So you're just going to
> have to trust me on this one. Or not.

    Trust had little to do with it. I was being asked to believe that a
certain hypothetical person (which I'm not entirely sure matches your
S.O.) existed without being given so much as an example like you have.

> As well, that's only one user. Even if there are millions like her, I don't
> know as focus-passing labels are particularly detrimental to her (she clicks
> buttons, fields, sliders, etc. accidentally from time to time too). 

    Okay, so what's the point then?

> But bottom-line, this 'hypothetical user' you keep dismissing really does
> exist, so you can take one point of contention off the table, at least.

    First of all, you haven't established that the person specified in 
previous messages does exist. That person is someone who deliberately 
clicks in the empty spaces between controls AND on labels to relieve 
nervous tension, but not on controls themselves. Your significant other 
does not fit this behavior because she clicks on everything.

    Secondly, my point wasn't that such a person didn't exist at all. My 
point was that the situation in which label focus passing caused a 
problem was so limited and improbably that the situation failed to 
justify a change to the specification. This does not mean that I oppose 
corrections to this part of the spec that elaborate how <label> should 
be handled in various user agents. I simply oppose the elimination of 
functionality and backwards compatibility with HTML 4.01 due to "a 
foolish consistency".