Re: [selectors] Finding a way to run "complete" profile selectors in CSS

On Thu, Jul 11, 2013 at 11:08 PM, Sylvain Galineau <galineau@adobe.com> wrote:
> On 7/11/13 8:24 PM, "Tab Atkins Jr." <jackalmage@gmail.com> wrote:
>>On Thu, Jul 11, 2013 at 5:40 PM, Sylvain Galineau <galineau@adobe.com>
>>wrote:
>>> On 7/11/13 5:09 PM, "Tab Atkins Jr." <jackalmage@gmail.com> wrote:
>>>>On Thu, Jul 11, 2013 at 4:41 PM, Sylvain Galineau <galineau@adobe.com>
>>>>wrote:
>>>>> So we have two solutions and we're asking people to think of problems
>>>>> they could solve?
>>>>
>>>>...what?  No.  I'm asking for what the usage patterns of "slow
>>>>selectors" might be, so we can decide what solution best addresses
>>>>them.
>>>>
>>>>The problem in question is stated right at the very beginning of my
>>>>email.
>>>
>>> Solutions to the 'problem' should be derived from use-cases. Instead
>>>we're
>>> making up solutions and then asking what problems they'd solve to
>>>validate
>>> them. You really think nothing can go wrong here?
>>
>>I'm not sure exactly what you're having trouble with.  Or maybe I just
>>don't understand exactly what you're asking for use-cases for?  I
>>explained the basic issue - there's a set of selectors that are, in
>>general, too slow for normal CSS matching.  Are you arguing against
>>that, or asking for examples of those selectors in use?  (If you are,
>>that's a completely different question - I'd be happy to justify the
>>addition of each of the "slow" selectors in a separate thread.)
>>
>>If that's not what you're asking, then you'll have to be more
>>specific.  The use-case is "these selectors are useful, but I want to
>>use them in normal CSS rather than just qSA", and should be pretty
>>much self-explanatory.  Is this not sufficient? What else are you
>>asking for?
>
> Ah. So you provided the use-case and concluded by asking 'what are the
> use-cases?'. That clarifies it!

I'm not willing to respond to you if all you do is continuously snark.
 Please be more helpful in your responses.  I asked specific questions
to see exactly what you wanted, and what part of what I'm providing is
insufficient.

> Fwiw Francois provided input that was much closer to use-cases. And
> from some of these we can determine there are other solutions.

He provided a limited set of simple things off the top of his head,
and we were able to determine that some of them, as written, are
simple enough that we may be able to introduce specialized selectors
for them in the fast profile.  That doesn't help us with the general
cases for those selectors, though, which are still apparently slow.

~TJ

Received on Friday, 12 July 2013 06:57:53 UTC