Now that user interface components are being created via scripted HTML elements such as div and span
Why should they be relying on such poor elements such as div and span is what I’d like to know.
Most of the time we happen to be able to achieve a reasonable level of accessibility through sensible use of a elements. I’m not saying it solves every problem, splits the sea and saves the chosen people, mind you, but (repeat) most of the time, and until a time when ARIA generalises, some anchors provide er... anchors (duh!) into the interface so that AT users can manage to use it.
Many RIA interfaces I’ve seen use divs and span and never a single a or button element. That’s poor.
By the way I’m not throwing the stone at Dojo in particular, especially considering the sheer volume of ARIA-oriented code they’ve injected in version 0.9.