Mobile Web Test Suites Working Group Blog

As the inputmode attribute has been dropped from the current draft of the HTML5 spec, we have now replaced the inputmode part of the Web Compatibility Test for Mobile browsers. In its place you will now find a new JavaScript framework test.

Modern web sites and web applications are becoming increasingly complex, often relying on large client-side scripts. A few different JavaScript frameworks aiming to ease the life of script authors have become very popular in recent years, and is today deployed on countless sites.

Any mobile browser that intends to make full use of said sites and applications must be able to load and use these libraries. Our new tests consists of loading the library of jQuery and the executing the simplest possible test using this library:

$(document).ready(function() { $("#jquery").addClass("green"); });

While the Mobile Web Test Suites Working Group has documented the technologies used in the Web Compatibility Test for Mobile Browsers for quite some time, I have always felt we could give more details on the said technologies and their state of development in a more graphical overview of the test.

To that end, I have developed the following graphic, that relates each of the square of the test to a given technology, with its dates of development and its state of standardization:

(also available as SVG, with links to the relevant specifications)

In the last few months the Mobile Web Test Suites Working Group have been not secretly meeting every Tuesday, discussing, among other things, an update to the Web Compatibility Test for Mobile Browsers. As the inputmode attribute is no longer part of the HTML5 specification, we need to replace our test on this with something else..

The last couple of months we have been investigating a replacement test around HTTPS/SSL and perhaps the Online Certificate Status Protocol. Different mobile devices can have different SSL trust roots installed (which often aren't viewable) and this security fragmentation could lead to a poor user experience. Or worse, Web developers fear to use SSL at all for this very reason.

SSL has always been a complex area and the current participants of this workgroup are not necessarily experts in the area. Perhaps you dear reader are? That didn't stop Opera's Wilhelm shooting SSL a parting shot with:

SSL is not establishing a hierarchy of trust, but a hierarchy of whom to blame.

So we are still looking for a new test for the famous Web Compatibility Test for Mobile Browsers. Currently we have proposals to look into:

1. Timers
2. MathML via CSS
3. A Javascript test from a well known framework

You could participate by either reading the meeting minutes, joining the call, hanging out on the #mwts channel on W3C IRC (irc.w3.org:6665), mailing the mwts mailing list or commenting on the blog. Or all of the above! Your feedback, ideas and participation are very welcome.