W3C home > Mailing lists > Public > www-style@w3.org > February 2014

Re: Selector Syntax Survey - subject indicator vs :has()

From: Tab Atkins Jr. <jackalmage@gmail.com>
Date: Wed, 12 Feb 2014 10:12:26 -0800
Message-ID: <CAAWBYDAndCt575SKc4324e=nUtfSjgsF7TkJqJd_BBFCUnPuTA@mail.gmail.com>
To: Daniel Glazman <daniel.glazman@disruptive-innovations.com>
Cc: www-style list <www-style@w3.org>
On Wed, Feb 12, 2014 at 8:52 AM, Daniel Glazman
<daniel.glazman@disruptive-innovations.com> wrote:
> On 12/02/2014 17:49, Tab Atkins Jr. wrote:
>> We know exactly how many people voted before the change (130) and will
>> keep that in mind when evaluating the results.
>> (The results are sorted by time, so we know *which* votes were cast
>> before the change as well.)
> I am saying asking between ! and :has() on one hand and ^ and :has() on
> the other is *NOT* the same question because of the negative (pun
> intended) aspect of !. I disagree you can infer any sort of result from
> the poll since 130 answers will be biaised.

Let me quote myself from immediately above:

>> (The results are sorted by time, so we know *which* votes were cast
>> before the change as well.)

We know which results were for which version of the poll.  We can
differentiate the results if necessary.

Note that we're not trying to do a complete poll of a small set of
people, like we do when straw-polling the group - if we were, we might
have to care about people who answered the first variant not answering
a second time for the other variant.

Instead, we're doing sampling of a large population.  The first
several hours of sampling are not different from the next several
hours.  If people answered the first version and not the second, that
doesn't matter, because the people who *did* answer the second version
are functionally identical for our purposes.

Now, some quick analysis.

Among the first 135 (those who saw the ! version of the poll), 26
voted for ! and 109 voted for :has().  In percentages, that's 19% and

Among the remaining 946 (those who saw the ^ version of the poll), 173
voted for ^ and 773 voted for :has().  In percentages, that's 18% and

As we can see, the actual character used doesn't seem to make any
difference.  People prefer :has() by a 4:1 ratio either way.

So I think this is pretty cut and dry.  We should use :has().

(I'm reading through the additional comments now.  I'll let y'all know
if I find anything interesting.)

Received on Wednesday, 12 February 2014 18:13:15 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:39:19 UTC