RE: https://social.tchncs.de/@lizzard/116300642945406280

This right here is essentially the reality many people face. We're a problem, just ... not any particular specific developer's problem, which means they don't need to solve the problem. Let somebody else solve the problem, or don't, see if they care.
Is it any wonder that people turn to the dreaded #AI given humans have kicked them to the curb more often than I can count?
Is this where I say I have a #Patreon now and put time into making #floss tools, #games and other digital things #accessible because people are done waiting for the heat death of the universe for it to become a priority? I mean it fits in the character count, might as well :P #accessibility

@zersiax I most often use the following analogy when talking about PDFs but it unfortunately applies to virtually every other accessibility scenario as well.

Imagine if you showed up every day to your job as a truck driver only to discover that the truck you are supposed to use is parked in a lake. Every damn day you arrive to the same thing. Of course you'll do your best to accomplish your deliveries but you'll struggle and the result won't be as awesome as you might want.

@zersiax

Someday you have ask; why is the truck in the lake?

Why do even Abode products routinely discard the document semantic structure when generating PDFs requiring remediation?

Why is it even possible to express a field and label in HTML without them being semantically related?

Why are the React/Qt/Android/etc. accessibility APIs such incomplete and frustrating piles of shit?

etc.

@zersiax My argument is that #a11y is no different than #i18n, security, or text layout. If we put any or all of these as burden on individual developers we should totally expect spectacularly inconsistent, incomplete and frequently terrible outcomes. Each of these things is just one of 5000 competing concerns that they have building a product.
@zersiax Assuming that at the product level, what reaches users, we won't ever escape the "ship early, ship often", "worse is better", "non-shipping software has no quality" mentality of "meh, good enough" then the only rationale solution is to build platforms and infrastructure that either solve the problem for the developer or make the problem impossible. eg. Buttons cannot be unlabelled as all buttons must include a non-empty label.