Re: Comments on the discussion about target size and accidental activation (was Re: Minutes: AGWG meeting April 4, 2017)

Gregg wrote:

> sounds very HTML Specific……



It does sound that way, but I’d say it is browser-specific.



Just to follow the logic:

·         Any content shown in a “web page” (our unit of testing) must be shown in a browser, or be capable of being shown in a browser, therefore you can test it in a browser.


·         Anything shown in a browser has to be rendered by a browser.


·         Anything rendered in a browser is translated into CSS pixels, or can be compared to CSS pixels.



So given the variety of hardware pixel densities, using CSS pixels as a unit of size measurement is the most robust way across devices, even for plugins like Flash/Silverlight. The worst-case scenario is that you compare the size of something in a plugin black-box to HTML content next to it, or in another browser window.



On a related note, are we worried about plugins for 2.1? Given that browsers are deprecating them [1], by mid-2018 they might be a non-issue. It is getting to the point where companies are having to pay people to enable plugins [2].



Cheers,



-Alastair



1] https://arstechnica.co.uk/information-technology/2016/08/chrome-starts-retiring-flash-in-favour-of-html5/


“with Chrome 55 in December [2016], Flash will be deprecated entirely, with exceptions for "sites which only support Flash." In both cases HTML5 is expected to take up the reins”



2] http://www.theverge.com/2017/3/24/15052286/fedex-adobe-flash-five-dollar-discount-print-orders


“The world’s most hated plugin is clinging on thanks to a generous $5 discount offer from FedEx, which apparently desperately needs customers to reenable Adobe Flash to be able to print things from its website”