Post

The link to the Meetup event requires a login to see it (because I was curious the presenter, topic, and possible slide links) but I burned my Meetup account a while ago. Also, FWIW, I consider `aria-label` to be so risky that I would not use it as an interim approach. I’m waiting for the interop.

Here is an alternative link to the event on LinkedIn, if that's better for you: www.linkedin.com/events/remot... I'll add it to the post as well. If you are interested in the name of the presenter, I only got to learn him as "Christian".

As for slides, there weren't any, except for an introduction by an agency that offers accessibility testing services with disabled testers, amongst which Christian. But the organizer, Johannes Lehner knows more: www.linkedin.com/in/jbeargrap...

One potential solution to rehide is to add an event listener for the blur (or focus out etc) event to the anchor and then re-add the attribute in it?

That sounds like an approach to explore. Although I had the browser console.log the document.activeElement to me during find-in-page and cycling through results, and focus did not move to the result until one taps the enter key.

If you used popovers for the labels you could rely on the only one being open at once behaviour along with a toggle event listener to readd hidden until found to the label? Could call showPopover from the beforematch event?

Dam this could legit replace sr-only/visually-hidden utilities

Ah, der Schepp teilt nen Link. Moment …! Der Schepp teilt einen eigenen Post! 🤯🔥🤩

It's ok to celebrate with cake right?

Given that I write a post once a year on the average, that sounds legit to me! 🍰

Eventuell bin ich noch unterkoffeiniert, aber ist das Codebeispiel hier schepp.dev/posts/rethin... dann nicht aus Screenreader-Sicht trotzdem ein leerer Link?

Ich will gar nicht wissen, was JAWS daraus macht ehrlich gesagt. Wo doch dieser SR so wunderbar und fehlerfrei funktioniert. NOOOOOT.

Ist bei den ATs nicht entscheidend, was im Accessibility Tree ankommt? Falls ja, müsste es denen ja dann egal sein, auf welchem Weg es dort ankommt.

Gemäß www.w3.org/TR/accname/#... müsste das mit `aria-label` funktionieren und ich hab das Konstrukt auch in der Accessibility Tree Ansicht von Firefox inspiziert und da war der `name` gesetzt (bzw. nach meinem Fix wieder gesetzt und vorher nicht). Es erschien mir demnach gut so.

Aber ich update den Artikel gerne, wenn es anders besser wäre.