RE: Call for implementation ideas

on Wednesday, May 18, 2011 5:05 PM  Boris Zbarsky wrote: 

>For what it's woth, the whenCanIUse tests are reasonably "shallow" (in
>the sense that they don't test various edge cases, complicated
>situations, etc, etc)...

I get that sense as well (at least for the SVG tests) but one thing that they might be able to provide, if the owners of the site might offer the data, would be data on how often features are queried. That might help drive the implementation agenda for features not yet implemented. Though clearly if the query data were used in such a way, folks might start trying to "rig" them.


Received on Wednesday, 18 May 2011 21:50:12 UTC