Test that :read-only matches <input>s to which [readonly] doesn't apply

[cvrebert]

This tests the other side of #2839.
As a reminder, per https://html.spec.whatwg.org/multipage/scripting.html#selector-read-only

The :read-write pseudo-class must match any element falling into one of the following categories [...]
The :read-only pseudo-class must match all other HTML elements.

Refs https://bugs.chromium.org/p/chromium/issues/detail?id=604154
Refs https://developer.microsoft.com/en-us/microsoft-edge/platform/issues/7229941/

[hoppipolla-critic-bot]

Critic review: https://critic.hoppipolla.co.uk/r/6400

This is an external review system which you may optionally use for the code review of your pull request.

In order to help critic track your changes, please do not make in-place history rewrites (e.g. via git rebase -i or git commit --amend) when updating this pull request.

[wpt-pr-bot]

Reviewers for this pull request are: @Ms2ger, @foolip, @gsnedders, @jdm, @jgraham, @plehegar, @sideshowbarker, @zcorpan, and @zqzhang.

[zcorpan]

Thanks for this! The tests match the spec AFAICT, but it's not clear to me that implementors want to match the spec. It's also pretty useless to have a selector for :read-only if it just means :not(:read-write). But maybe it makes sense somehow? cc @Hixie @tabatkins

[cvrebert]

Yes, I agree that the spec isn't particularly reasonable here, but didn't dig in to evaluate whether the weirdness is justified or not (perhaps there's a history here).
Note that Safari currently implements this as-currently-spec'd, unprefixed.

it's not clear to me that implementors want to match the spec.

Given their response to https://developer.microsoft.com/en-us/microsoft-edge/platform/issues/7229941/ , the Edge folks currently seem to be in the "the spec is bad" camp.

[zcorpan]

This changed in WebKit in WebKit/WebKit@c097867

@BenjaminPoulain what are your thoughts? Also see @FremyCompany's comment in the Edge issue

Maybe we should take this to www-style?

[BenjaminPoulain]

It's also pretty useless to have a selector for :read-only if it just means :not(:read-write).

That's pretty common. Take ":link" and ":visited" for example.

I would expect :read-only and :read-write to be opposite, as it is for any permission system. That what follows the principle of least astonishment IMHO.

Maybe we should take this to www-style?

Yep. I do not intend to change WebKit to go against the spec.

[zcorpan]

:link and :visited are mutually exclusive but are not the negation of each other; neither of them mathes a div, for instance. They only match hyperlinks. Similarly it seems reasonable to assume that :read-only would only math elements that are potentially writeable but are currently in a readonly state.

With contenteditable I suppose it's possible to argue that any element can become writeable, though.

[zcorpan]

I'll merge this since it's correct per the current spec. I can start a new thread in www-style.

[cvrebert]

The new thread: https://lists.w3.org/Archives/Public/www-style/2016Apr/0294.html