Re: [css-colors] Specify the System Colors colors


On Aug 31, 2013, at 2:04 AM, "L. David Baron" <dbaron@dbaron.org> wrote:

> On Friday 2013-08-30 12:48 -0700, Dirk Schulze wrote:
>> So according to the tests from Simon, all browsers support some kind of color sets that are not interoperable at all.
> 
> Um, Simon's tests don't show lack of interop.  The purpose of the
> system fonts is to reflect system settings; showing lack of interop
> requires showing lack of interop between different browsers on a
> system with the same settings, or showing failure to match the
> defined behavior.
> 
> That said, I suspect actual interop is pretty good on Windows, and
> less good on other desktop platforms.

Well, that is the point: you can still identify the system that is used… at least. If system colors are implemented correctly, then you have changes on each UI system theme.

And just a reminder to the group: The idea of system colors was not just to have pleasant and integrated UI designs. It can be important for accessibility as well. It allows the browser to use high contrast or inverted UI themes for users with special needs and preferences.

> 
>> There is just one exception with Android and iOS and we don't know if that changes with one of the next versions.
> 
> If we standardize on a set of colors, I think it would make more
> sense to use a set of colors from the default theme of a recent
> Windows version than to use the Android/iOS defaults.

Simons test do not show if the browsers actually support the purpose of system colors. So if you change your theme to hight contrast themes, does it affect system colors on browsers? (As it should?)

Personally I am not particular concerned to use system colors for particular pleasant integration. (You never got windows looking native in the past.) As noted above, the accessibility is the important factor. Do the systems above (Android and iOS) allow changing system colors to high contrast? And does that change the CSS system colors? If not, the whole purpose of system colors is not really relevant.

> 
>> Do UAs not see any privacy concerns? If so, why does the view between the WG and the UAs differ? Are the color profiles actively used as spoofing mechanism? And what does the attacker get for relevant information from the color settings (even if they would be OS theme dependent)?
> 
> So I think there are two separate privacy/security concerns:
> spoofing (presenting fake dialogs to the user that appear to be
> real) and fingerprinting (using data that differs between users to
> identify them).
> 
> In practice, I'm not that worried about spoofing.  Users seem to be
> spoofed just fine with screenshots of dialogs.  (Perhaps that's a
> sign that there's so much non-native-looking UI around that users
> have no expectation of native-looking UI.)  Though spoofing could
> become more of a risk in the future, I suppose.
> 
> The fingerprinting is perhaps more of a real concern, but I think
> this is far from the worst fingerprinting vector available in CSS.
> (I suspect that's fonts.)
> 
> -David
> 
> -- 
> 𝄞   L. David Baron                         http://dbaron.org/   𝄂
> 𝄢   Mozilla                          https://www.mozilla.org/   𝄂
>             Before I built a wall I'd ask to know
>             What I was walling in or walling out,
>             And to whom I was like to give offense.
>               - Robert Frost, Mending Wall (1914)

Received on Saturday, 31 August 2013 04:41:54 UTC