3 ms·

>Second, putting out unaccessible PDF content is 100% certain to be problematic for people with accessibility needs. That's where you're wrong. A plain PDF/A i

by konjin 6y ago

>Second, putting out unaccessible PDF content is 100% certain to be problematic for people with accessibility needs. That's where you're wrong. A plain PDF/A is the simplest thing to scrape for text. You know what is a nightmare to deal with? Html with javascript: https://www.dailymail.co.uk/home/index.html https://www.dailymail.co.uk/home/index.html Try a screen reader on the above nightmare and tell me what you hear. For me it's: iframe-iframe-iframe-iframe-iframe-iframe-iframe-iframe-iframe... turning it into a png and trying to scrape text doesn't help because there is literally zero predictable structure to the text, is it one, two, three or 8 columns wide? Depends on where in the page you look. So again, thanks for the outrage, but we don't need people like you who don't understand the first thing about what we go through defending us and shitting on people trying to make the world better.

jl6 6y ago
Since you’re a screen reader user I’d appreciate feedback on the aforementioned PDF website: https://www.lab6.com/0 https://www.lab6.com/0 I did very brief testing using an iOS screen reader and it worked OK but didn’t really expose any of the semantic tags.
- konjin 6y ago
  s/not because it easy/not beucase it's easy/g s/that ISO want 198/that ISO wants 198/g Other than that 10/10. As readable as hacker news on my setup. The only way that can be better if is you wrote it as pure text. If you follow basic formatting like double new lines after headings and punctuation most screen readers do a better job of giving you the semantic information by breaks and tones than the always broken 'semantic tags'. Put it another way, would you want to read raw html to make sure that what you thought was a <p> was actually a <pre>? That's what I get whenever I try and use those. Again a solution by people who don't need to eat their dog food.

by konjin 6y ago

jl6 6y ago

Since you’re a screen reader user I’d appreciate feedback on the aforementioned PDF website: https://www.lab6.com/0 https://www.lab6.com/0 I did very brief testing using an iOS screen reader and it worked OK but didn’t really expose any of the semantic tags.

konjin 6y ago
s/not because it easy/not beucase it's easy/g s/that ISO want 198/that ISO wants 198/g Other than that 10/10. As readable as hacker news on my setup. The only way that can be better if is you wrote it as pure text. If you follow basic formatting like double new lines after headings and punctuation most screen readers do a better job of giving you the semantic information by breaks and tones than the always broken 'semantic tags'. Put it another way, would you want to read raw html to make sure that what you thought was a <p> was actually a <pre>? That's what I get whenever I try and use those. Again a solution by people who don't need to eat their dog food.