Bayesian analysis vs. statistical significance, clearly explained.
This might just be the most nerdily specific book I’ve read and enjoyed. Even if you’re not planning to build a web browser any time soon, it’s kind of fascinating to see how HTML is parsed—and how much of an achievement the HTML spec is, for specifying consistent error-handling, if nothing else.
The last few chapters are still in progress, but you can read the whole thing online or buy an ePub version.
Rebuttals to the most oft-asked requests for browsers to change the way they handle CSS.
Instead of being prescriptive about error messaging, we use what the browser natively gives us.
Usability Testing of Inline Form Validation: 40% Don’t Have It, 20% Get It Wrong - Articles - Baymard Institute
I saw Christian speak on this topic at Smashing Conference in Barcelona. Here, he takes a long hard look at some of the little things that sites get wrong when doing validating forms on the fly. It’s all good sensible stuff, although it sounds a bit medical when he takes about “Premature Inline Validation.”
This is really, really clever. You can’t use generated content (
:after) on replaced content. The
img element is replaced content …but only when the image actually loads. So if the image fails to load, you can apply specific fallback styles (using
Friendlier HTTP errors.