Re: [selectors4] Open issues

On Fri, Apr 19, 2013 at 7:17 PM, Ian Hickson <ian@hixie.ch> wrote:
> On Wed, 17 Apr 2013, Tab Atkins Jr. wrote:
>> > That's even more lax than HASH.  I would normally think it appropriate
>> > for CSS to defer to HTML5 on this sort of thing (and I do think the
>> > principle of least surprise says that whatever can go in id="...", CSS
>> > should be able to match it with #...; and I'm always in favor of
>> > getting rid of quirks, ceteris paribus) but there's no way we can be
>> > *that* lax.  Maybe both specs need to change here.
>>
>> Right, it's looser, but I don't think we want to special-case it
>> sufficiently that we go past the confines of the hash token.  I'm
>> satisfied with using escapes at that point.
>
> The intent on the HTML side was to rely on CSS escaping, FWIW.

I think what would make me personally happy, here, is: CSS relaxes
#xxx in a selector to accept any HASH token.  HTML adds a
non-normative note for authors discouraging use of id values
containing ASCII characters other than A-Za-z0-9_- .

zw

Received on Friday, 19 April 2013 23:40:44 UTC