- From: Patrick H. Lauke <redux@splintered.co.uk>
- Date: Thu, 17 Dec 2015 22:32:13 +0000
- To: "public-mobile-a11y-tf@w3.org" <public-mobile-a11y-tf@w3.org>
On 17/12/2015 22:06, Richards, Jan wrote: > With the huge variety of screen sizes and resolutions, it does feel odd to use "mm" (and measuring from the screen does depart from usual WCAG testing practice). > > On the other hand, I don't think a minimum physical size is outside of developer thinking. Mobile developers are constantly grappling with this because even if they don't know about accessibility, they don't want 1mm buttons on small devices. It's not outside of developer *thinking*, but simply outside of developer *control*. There is no way for a dev to guarantee a certain physical rendering size. All it would take is a device to come around with some crazy mapping (and I now remember the uproar that went around the responsive web design community when the iPad Mini with its "smaller screen/same resolution in CSS pixels" came out) for any previously safe and conformant content to be immediately non-conformant when tested on that device. Sure, devs will likely find some hacky loophole to try and make their sites/content work ok even in those scenarios, but it's basically a fundamentally shaky foundation to build a hard pass/fail requirement on. > As Jon points out, maybe CSS pixels are the way to go. Indeed - for web content, they're the most natural fit for any kind of measurement (he said, still wondering why anybody would ever talk about "points" in the context of web content...looking at you, "large scale (text)" http://www.w3.org/TR/UNDERSTANDING-WCAG20/visual-audio-contrast-contrast.html#larger-scaledef) > They are actually an angular measurement (http://inamidst.com/stuff/notes/csspx), so they get larger on devices intended to be viewed from further away. (provided that device manufacturers, OS vendors, and UA vendors implement them correctly - but for the majority of smartphone/tablet devices, it's often correct - barring examples like the iPad Mini, and a myriad of other devices in the Android world - though certainly not implemented with the mathematical rigor implied by https://drafts.csswg.org/css-values/#reference-pixel) > So 44px was 11 mm on my desktop monitor but 8.5mm on my Samsung Note phone. > > Maybe the SC wording needs to be more vague and then CSS px and mm measurements can be offered as sufficient techniques? I'd be in favor of having the *normative* text be more vague. I fear though that even if it was softened to something like "must be at least as large as an average person's fingertip" (see for instance the note on "Size and positioning of UI" in https://msdn.microsoft.com/en-us/library/windows/apps/hh465370.aspx), the end result is still that there's now a hard requirement to achieve a minimum physical size, but that there's no way for a developer to actually *guarantee* this as they can't check for physical device size, whether or not that device has implemented a sensible device pixel ratio, and they can't test on every possible device out in the wild and coming out in the future. Unless a strong note/exemption can be added somehow to the normative text which puts the onus on device/OS/UA having implemented a sensible mapping? In a further *informative* note we could then provide commonly used/suggested sizes such as 44x44px, and clarify that it may not always be possible for a developer to explicitly know the exact physical size that this will map to, as it depends on multiple factors outside of their control. P -- Patrick H. Lauke www.splintered.co.uk | https://github.com/patrickhlauke http://flickr.com/photos/redux/ | http://redux.deviantart.com twitter: @patrick_h_lauke | skype: patrick_h_lauke
Received on Thursday, 17 December 2015 22:32:37 UTC