Live data from Hacker News

Web accessibility for 2019

blog.sourcerer.io

11–13 of 13 posts

Re: Web accessibility for 2019

#11

The author describes what he purports to be a simple, universal language for literally all content . Everything, he says, can be described and explained in terms of "role, states, attributes, relations, and actions." > Does your web app use chemistry ring formulas or math equations or anything else from any other subject you can think of? So, you can definitely describe all those symbols and formulas to your friend,…

Those five concepts are not random, they go from the assistive technologies world: most if not all desktop APIs use these terms. Agreed, the state concept (or even role concept) can be technically mapped into attributes, but I keep using them just to stick closer to existing terminology. When it comes to ARIA, all browsers map ARIA taxonomies into desktop APIs, and thus I think it's fair to say ARIA is mapped into th…

No, those problems can’t be solved this way. Your taxonomy may seem more “expressive” to you, but your five concepts are too vague to build a useful screen-reader around. Web developers won’t use the concepts consistently; screen readers won’t interpret them consistently.

None of the relevant vendors are on board, because they’ve all considered a simpler taxonomy and rejected the idea, in favor of a richer, more complex system that can actually be implemented.

Re: Web accessibility for 2019

#12

Earlier quoted context omitted.

Those five concepts are not random, they go from the assistive technologies world: most if not all desktop APIs use these terms. Agreed, the state concept (or even role concept) can be technically mapped into attributes, but I keep using them just to stick closer to existing terminology. When it comes to ARIA, all browsers map ARIA taxonomies into desktop APIs, and thus I think it's fair to say ARIA is mapped into th…

No, those problems can’t be solved this way. Your taxonomy may seem more “expressive” to you, but your five concepts are too vague to build a useful screen-reader around. Web developers won’t use the concepts consistently; screen readers won’t interpret them consistently. None of the relevant vendors are on board, because they’ve all considered a simpler taxonomy and rejected the idea, in favor of a richer, more comp…

Having these concepts exposed to the web doesn't repeal the necessary of standardization, i.e. both web developers and AT vendors has to stay in agreement on content semantics they operate with, but it will make the agreement process simpler than it's now, because it doesn't require implementation from browsers for each and every new feature. Also it will enable experimenting on the web, where web developers and AT vendors can try new things before stuffing them into standards, which should boost the accessible web.

Re: Web accessibility for 2019

#13

Earlier quoted context omitted.

No, those problems can’t be solved this way. Your taxonomy may seem more “expressive” to you, but your five concepts are too vague to build a useful screen-reader around. Web developers won’t use the concepts consistently; screen readers won’t interpret them consistently. None of the relevant vendors are on board, because they’ve all considered a simpler taxonomy and rejected the idea, in favor of a richer, more comp…

Having these concepts exposed to the web doesn't repeal the necessary of standardization, i.e. both web developers and AT vendors has to stay in agreement on content semantics they operate with, but it will make the agreement process simpler than it's now, because it doesn't require implementation from browsers for each and every new feature. Also it will enable experimenting on the web, where web developers and AT v…

I think people have been moving away from the semantic organization because trying to properly dimensionalize the web has been so fraught.
Post reply on HN