It's definitely simpler to completely ignore accessibility. It's also a bad thing.
It's easier to not put a ramp at my shop, after all there aren't that many people in wheelchairs. It's easier to not add induction loops to help those with hearing aids, after all there aren't that many people with them.
We can keep going with this logic. Facial recognition only works with white faces? But most of my users are white, let's just keep it.
It's easier to exclude people. Across society, that ends up pretty poorly, particularly when it's the same groups that are excluded time and time again.
> What's the incentive?
It would be nice if making the content work for people with accessibility requirements was a reward in itself, but this is why we end up with laws, to make not doing these things a punishable offence so that in the end it's easier to actually do them.
Weird how that expectation only seems to apply to tech folk using assistive technologies. I'm sighted and while I think HN is needlessly ugly and user hostile in some places the only thing I need to do to use it is zoom in a bit. But I guess that just means I'm more technically adept than someone who has to mess around with their screen reader or crawl through the pile of poorly maintained "UIs" to make the site work.
> It works and is lightweight.
That's quite the response to someone demonstrating that it doesn't work.
The only way I can see to fix this is to enable something like OAuth2, which would be a fairly significant departure from HN's brand of minimalism.
My understanding is that dang both moderates the site and maintains the codebase. I've seen new functionality slowly pop up over time; the site is definitely in a slow-but-steady state of continuous development. You may have a point in terms of noone wanting to touch the codebase in specific places (like the voting logic), but I can't see semanticity improvements being too difficult - at least, my surface expectation is that it shouldn't be too hard. It's possible the Arc code specifically powering HN is a giant spaghetti mess.
The incentive perspective is also a very good point. IMHO this is one of those things that's high-effort, high-QoL-improvement, yet low-*perceived*-impact. I'm still trying to figure out what to call it. Whatever it is, it's really hard to straightforwardly make the point that this functionality is absolutely worth investing in, very sadly. It's just one of those things that just doesn't pop out in an... obvious way, for want of a better way to put it. And it's an absolute shame.