[Webkit-unassigned] [Bug 22261] Clicking on a non-text input element does not give it focus

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=22261

--- Comment #77 from David Scourfield <bugzilla at dai.codes> ---
(In reply to Gérard Talbot from comment #76)

> > User agents *should consider* focusable areas with non-null tabindex values to be click focusable.
> (...)
> > so this is a bug as it's a violation of the HTML spec.
> 
> Generally speaking, when reading a spec, "SHOULD" means recommended but not
> required. 
> https://www.ietf.org/rfc/rfc2119.txt 
> Along with "consider" verb, this means very little. So, this is not a
> violation of the HTML5 spec. Safari could still claim compliance with the
> HTML5 spec.

Yes, agreed that 'should' is a recommendation rather than a requirement, but nonetheless recommendations should not be so easily dismissed.  It's reasonable for content authors to expect user agents to adhering to the non-null tabindex recommendation from the spec, just as it's reasonable for Safari to choose not to focus on click by default.

I think you're misunderstanding the meaning of the verb 'consider' in this context.  'Consider X to be Y' means that you should think of X as Y.  E.g., 'you should consider the case closed' does not mean you should think about whether or not the case is closed, it means you should think _that_ the case is closed.

Therefore this sentence in the spec is not an instruction for UA authors _to think about whether they want to_ make such elements click focusable, it's an instruction for UA authors to _explicitly treat_ such elements as click focusable.  This is quite clear when reading the sentence in the context of the whole section of text in which it appears in the spec.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20210613/f9352b03/attachment-0001.htm>