[whatwg] Re: Web Forms 2: Altenative to <select editable>

[Malcolm Rowe]

Jim Ley writes:
>>> WF2 UA's must provide the ability for Users to submit forms without
>>> client-side validation.
>> Well, I'm not sure if I'd support it as a MUST/SHOULD recommendation - I
>> just rewrote what you'd put to indicate that a UA could ('MAY') provide
>> such a dialog. But it's worth considering.
> I believe it's at least SHOULD, as we noted there's high risk of bad
> implementations, or buggy implementations, and this would at least let
> people carry on and achieve their end.  Bugs in UA's shouldn't stop
> pages working.

I'm not sure I agree that there *is* a high risk of non-compliant 
implementations, and there is something of a contradiction in requiring 
('MUST') behaviour of a UA that is only really necessary if it fails to 
implement required ('MUST') behaviour in other parts of the specs. I do have 
some agreement with your position, though, and I'd be happy to see it 
included as a 'MAY', though whether it then belongs in the spec at all would 
be a valid question. 

On a slightly tangential point, are there any accessibility considerations 
that would make disabling client-side validation a necessary feature? I 
imagine not, because you can't disable server-side validation. 

>> but it's a *lot* easier to fix a broken UA than it is to
>> update scripts on all broken sites.
> Right, but I don't care about other sites (in fact it may be to my
> advantage that they're broken) I do care about my scripts, and it's
> much easier to fix those than it is to fix a broken UA, and get it
> deployed.

Okay, you've got a point, but while *you* might be diligent in updating your 
site, other people won't be. And while it's highly likely that initial 
implementations of WF2 support from Mozilla or Opera will be incomplete and 
buggy (to some extent), the buggy features simply will not get used on most 
sites. Plus, six months down the road, the quality of the browser's 
implementation will have improved significantly, but the quality of *most* 
site's scripts will not. 

To some extent, the people who will be initially using a Web Forms 
2-compliant UA are the same people who are capable of upgrading to a newer 
version. I do agree that this doesn't apply as much to embedded browsers 
(Mozilla/Minimo or Opera/Symbian), but I'd imagine they're more likely to 
have a restricted or more conservative feature set. 

Your argument would be a lot more applicable if we assumed that a common 
implementation could harbour a bug that significant numbers of designers 
depended upon, similar to IE's box-sizing incompatibility. I'm not sure how 
likely this is, as given that the market shares of Mozilla and Opera are 
roughly the same, a designer using WF2 would likely test on both. 

Regards,
Malcolm