4 ms·

In the end, software accessibility represents an edge case affecting a tiny fraction of users. You can argue it’s morally wrong to not support that edge case,

In the end, software accessibility represents an edge case affecting a tiny fraction of users. You can argue it’s morally wrong to not support that edge case, but I do understand why it wouldn’t be a priority for most without legal prodding.

acdha 3y ago
It’s more than a tiny fraction, but yes, that’s why the laws are so important: the cost of not having them is high to specific people, and businesses may not see that directly because someone who finds your product hard to use is likely to just stop without telling you so it’s hard to tell how many affected people there actually are. Having a simple legal requirement short circuits all of those discussions.
squigz 3y ago
> In the end, software accessibility represents an edge case affecting a tiny fraction of users. I'm curious: What fraction do you think it is? And what fraction do you think it should be for it to be taken more seriously?
- infotainment 3y ago
  To get into exact numbers, it starts to depend on what exactly is meant by "accessibility". For simplicity's sake, let's take supporting screen reading software. What % of the population uses a screenreader? If we assume that every blind user uses one, then the slice is about 2.4%. According to CanIUse, that's a similar percentage to the share of users who use browsers that don't support Flexbox. Have you used Flexbox recently? If so, you've excluded a similar slice of your potential user base. And, just like with ignoring flexbox, you can of course code around this. I've actually written SPAs with screen reader support, and it's a pain. Advocates will tell you that if you just follow the WAI-ARIA standard, it'll work seamlessly! But no -- in practice, each screen reader is like a mini-IE6; they all deviate from the standards in strange and unpredictable ways.
  - squigz 3y ago
    > To get into exact numbers, it starts to depend on what exactly is meant by "accessibility". For simplicity's sake, let's take supporting screen reading software. What % of the population uses a screenreader? If we assume that every blind user uses one, then the slice is about 2.4%. Ignoring the fact that this alone represents a small slice of a larger pie, 2.4% still doesn't seem insignificant to me - that's nearly 8 million people in America alone. In any case, that answers my first question - now how much higher would that number need to be for it to be significant to you?
    - infotainment 3y ago
      What percentage was it when everyone collectively decided to stop supporting IE6? 20%? To me it’s less of an absolute percentage and more of a function of how much extra effort needs to go into it, and whether the usability of the rest of the product needs to suffer. I’ve seen potentially useful features nixed because they can’t be made fully compatible with JAWS, which seems silly to me.

acdha 3y ago

It’s more than a tiny fraction, but yes, that’s why the laws are so important: the cost of not having them is high to specific people, and businesses may not see that directly because someone who finds your product hard to use is likely to just stop without telling you so it’s hard to tell how many affected people there actually are. Having a simple legal requirement short circuits all of those discussions.

squigz 3y ago

> In the end, software accessibility represents an edge case affecting a tiny fraction of users. I'm curious: What fraction do you think it is? And what fraction do you think it should be for it to be taken more seriously?

infotainment 3y ago
To get into exact numbers, it starts to depend on what exactly is meant by "accessibility". For simplicity's sake, let's take supporting screen reading software. What % of the population uses a screenreader? If we assume that every blind user uses one, then the slice is about 2.4%. According to CanIUse, that's a similar percentage to the share of users who use browsers that don't support Flexbox. Have you used Flexbox recently? If so, you've excluded a similar slice of your potential user base. And, just like with ignoring flexbox, you can of course code around this. I've actually written SPAs with screen reader support, and it's a pain. Advocates will tell you that if you just follow the WAI-ARIA standard, it'll work seamlessly! But no -- in practice, each screen reader is like a mini-IE6; they all deviate from the standards in strange and unpredictable ways.
- squigz 3y ago
  > To get into exact numbers, it starts to depend on what exactly is meant by "accessibility". For simplicity's sake, let's take supporting screen reading software. What % of the population uses a screenreader? If we assume that every blind user uses one, then the slice is about 2.4%. Ignoring the fact that this alone represents a small slice of a larger pie, 2.4% still doesn't seem insignificant to me - that's nearly 8 million people in America alone. In any case, that answers my first question - now how much higher would that number need to be for it to be significant to you?
  - infotainment 3y ago
    What percentage was it when everyone collectively decided to stop supporting IE6? 20%? To me it’s less of an absolute percentage and more of a function of how much extra effort needs to go into it, and whether the usability of the rest of the product needs to suffer. I’ve seen potentially useful features nixed because they can’t be made fully compatible with JAWS, which seems silly to me.