3 ms·

Amazon also changed the way their Kindle Cloud Reader (the Kindle web app) works so that it now displays an image of text instead of HTML text. This appears to

Amazon also changed the way their Kindle Cloud Reader (the Kindle web app) works so that it now displays an image of text instead of HTML text. This appears to have been done for anti-piracy reasons, but it has the knock-on effect of breaking accessibility browser plugins that people with disabilities rely on to make Kindle books accessible. In case anyone is not aware, image-of-text is considered a newbie mistake from an accessibility perspective, so it's odd for Amazon to take this step backward. Amazon rolled out this update over a year ago, and my startup still gets emails from people who are upset that they can no longer read books they purchased with the BeeLine Reader plugin, as they expected they would when they bought the books. There's nothing we can do to get Amazon to fix this, but we are now launching partnerships with other ebook platforms that are more accessibility-friendly.

criddell 4y ago
> There’s nothing we can do to get Amazon to fix this It might be easier to get publishers to fix this. Not all publishers use DRM on books sold by Amazon (Tor is a notable example). Books not encumbered with DRM can be used by alternative apps.
- iggldiggl 4y ago
  > Not all publishers use DRM on books sold by Amazon Is there some way of finding that out in advance of buying the book? (Other than individually asking each publisher/author?)
  - gnicholas 4y ago
    Apparently the book's details will say "simultaneous device usage: unlimited" if it doesn't have DRM.
- gnicholas 4y ago
  Interesting point! I'm not sure that Amazon would change their web app's software stack for a subset of their publishers though. My guess is we'd need to find a big publisher and get them to ask Amazon not to use this new image-of-text method. But I assume that any big publisher would be unlikely to not use DRM.
dannyw 4y ago
You could consider contacting lawyers to explore accommodations under the ADA.
- gnicholas 4y ago
  To try to get them to change it back to the way it was, with HTML text? I'm hesitant to involve lawyers at all, since I don't want to get off on the wrong foot with Amazon. I view them as a potential licensee (we license to one of their smaller competitors already), and I'm sure their attitude toward us would take a turn for the worse if they heard about us via a lawyer/implied legal threat. Or perhaps I'm misunderstanding the suggestion?
dunham 4y ago
For what it's worth, last I checked (Nov 2021) the image was rendered client side. They send a collection of glyphs as SVG strings like "M31.9824,43.9453 L364.014,-719.971 L495.117,-719.971 L163.086,43.9453 L31.9824,43.9453 Z " And then each line is a sequence of indices into the glyph array. There isn't a mapping back to letters, that I know of, but one could prepare a table of glyph to character for the fonts they commonly use. That could be used to extract books, but might be a bit complicated/fragile for your use case.

criddell 4y ago

> There’s nothing we can do to get Amazon to fix this It might be easier to get publishers to fix this. Not all publishers use DRM on books sold by Amazon (Tor is a notable example). Books not encumbered with DRM can be used by alternative apps.

iggldiggl 4y ago
> Not all publishers use DRM on books sold by Amazon Is there some way of finding that out in advance of buying the book? (Other than individually asking each publisher/author?)
- gnicholas 4y ago
  Apparently the book's details will say "simultaneous device usage: unlimited" if it doesn't have DRM.
gnicholas 4y ago
Interesting point! I'm not sure that Amazon would change their web app's software stack for a subset of their publishers though. My guess is we'd need to find a big publisher and get them to ask Amazon not to use this new image-of-text method. But I assume that any big publisher would be unlikely to not use DRM.

dannyw 4y ago

You could consider contacting lawyers to explore accommodations under the ADA.

gnicholas 4y ago
To try to get them to change it back to the way it was, with HTML text? I'm hesitant to involve lawyers at all, since I don't want to get off on the wrong foot with Amazon. I view them as a potential licensee (we license to one of their smaller competitors already), and I'm sure their attitude toward us would take a turn for the worse if they heard about us via a lawyer/implied legal threat. Or perhaps I'm misunderstanding the suggestion?

dunham 4y ago

For what it's worth, last I checked (Nov 2021) the image was rendered client side. They send a collection of glyphs as SVG strings like "M31.9824,43.9453 L364.014,-719.971 L495.117,-719.971 L163.086,43.9453 L31.9824,43.9453 Z " And then each line is a sequence of indices into the glyph array. There isn't a mapping back to letters, that I know of, but one could prepare a table of glyph to character for the fonts they commonly use. That could be used to extract books, but might be a bit complicated/fragile for your use case.