People regret buying Amazon smart displays after being bombarded with ads

Post by Andrex » Wed, Oct 15, 2025, 6:51 PM UTC

pixelmelt wrote: post #63

Andrex wrote: post #57

Earlier quoted context omitted.

They got me by the balls with Kindle. Someone(/me if I had enough time) should make some kind of LLM scraper extension for Amazon's Web Reader. Run it on your own ebooks while you're signed in and you get the EPUBs. Maybe this already exists... I know there was some kind of cutoff earlier this year for unDRMing your Kindle books, but I definitely missed that by a mile.

I've been working on exactly this! After buying a book and having an awful exerience in their native reader I realized they no longer let you download your ebooks. Thus began the war on their web reader, they have some interesting defenses on how they encode the book, its basically a bunch of SVG paths that render to letters that are mapped to IDs that change on every request to fetch book data (the most you can fetc…

Post by pixelmelt » Thu, Oct 16, 2025, 8:28 PM UTC

boldlybold wrote: post #127

pixelmelt wrote: post #63

Earlier quoted context omitted.

I've been working on exactly this! After buying a book and having an awful exerience in their native reader I realized they no longer let you download your ebooks. Thus began the war on their web reader, they have some interesting defenses on how they encode the book, its basically a bunch of SVG paths that render to letters that are mapped to IDs that change on every request to fetch book data (the most you can fetc…

Damn, that's an impressive obfuscation and debugging. If you get around to releasing it I would be interested.

Post by Andrex » Fri, Oct 17, 2025, 6:27 AM UTC

pixelmelt wrote: post #192

Andrex wrote: post #191

Earlier quoted context omitted.

That is amazing!! I will check that out whenever it's done. Thank you for your work! Dumb question: why not flip through the pages, take screenshots, then use OCR? Maybe some kind of jank Amazon DRMalike?

OCR would likely work, however I would need to render everything out with a browser and do (probably) more processing. I also wanted a way to preserve text styling exactly as the web viewer shows it, not sure if OCR supports detecting text alignment. Also I've released the tool and a post about it here: https://news.ycombinator.com/item?id=45610226