Re: [mediaqueries] CSS "prefers-" media features (from TPAC discussion)

> On Feb 6, 2016, at 11:19, fantasai <fantasai.lists@inkedblade.net> wrote:
> 
> On 02/05/2016 11:11 AM, Dean Jackson wrote:
>> 
>>> On 5 Feb 2016, at 6:43 PM, James Craig <jcraig@apple.com> wrote:
>>> 
>>> Tab et al,
>>> 
>>> Do you have an updated time estimate? We've been shipping these features for multiple years on our naive platforms and we'd love to make the same accommodations  possible in the Web. We received another request for prefers-* Media Features today; engineers from one of our major web sites have expressed an interest in being able to use these soon.

I haven't been able to prioritize this so far, and since Tab said he would, I haven't really tried. But we are officially stalled, so let's try again.

>>> How can I move this along? Would it help if I proposed an edit to give you some "straw man" text to rewrite?

I do think it would help, thank you for the offer.

That said, what I'd want to do with that is not so much re-write it (as I have no reason to believe that my writing skills are any better than yours), but run it by other browser vendors and other OS vendors, to see how your list matches their view of the world.

We have to figure out if the various OSes agree on which topics we should be able to respond to, how fine grained to decompose this into, and what to do along each axis. I think "what to do" is typically one of 3 things:

a - The UA/OS would prefer you to foobar, but has done nothing to you. Please foobar yourself. (e.g. prefer-reduced-motion)

b - The UA/OS has foobared you, and there's nothing you can do about it. You may respond to it by changing the bits of the page you know are bad when foobared. (e.g. pixel level color inversion)

c - The UA/OS has foobared you, and here's a property that can undo the effect, which you should use if you know of a better way of foobaring yourself. (e.g. ink saving for printing)

However, on each topic, there's no reason a priory to assume different OSes take the same approach.

I don't know whether it will be more effective to start with a concrete proposal that matches the apple view of the world (since you're offering to start), and work on it until it works for everyone else,
or if we should start by a big table of topics and behaviors and see where each OS falls.

                         a    b    c
contrast
color-invert
reduced-motion
reduced-transparency
ink-saving
...

>> Maybe you could send a pull request on https://github.com/w3c/csswg-drafts/tree/master/mediaqueries-3 ?
> 
> It should be against MQ4; L3 is REC and thus frozen.
> 
>> (Note: even though I'm in the group, I'm not sure if CSS Core uses github this way - I know Houdini can)
> 
> We generally want to make sure things get discussed on the list first,
> rather than in pull requests, but since this has, that's not a problem.
> 
> We should probably get a more solid proposal for the WG to discuss in
> any case; that could be edits to the spec, a separate file somewhere,
> or--more simply--just more detail in the email: that is, the definite
> list of queries, grammars, and their meaning. The initial post was
> just "some examples".

If you do a pull request, please mention it on the list as well. I'd say
a pull request is fine, because it gives us something concrete to look at,
but since we need to the review I talked about, www-style should stay the
anchor point for this.

 - Florian

Received on Monday, 8 February 2016 03:54:50 UTC