Paper is here: https://arxiv.org/pdf/2501.06623

Why? Can you explain a little bit?

Let’s be honest: That totally is a monopolistic tactic of Microsoft and regulators would do great to force this open. Force them to release a Linux Office. Force them to really make their file format transparent so that everybody can write working parsers

Most users could work with Linux without problems. They are using their browser, maybe a mail client and maybe some office suite to write stuff. If you mention this on the internet, there will always be someone who shows up and complains that his personal workflow with some obscure software that powers the nuclear reactor that he is running as hobby in his home won’t work with Linux.

It’s kind of the same with discussions about commuting/bicycles/cars. If you’re discussing that maybe more people could commute by bike, the same people will show up and complain that this would not be possible for them because they live on a remote mountain top in the scorching desert far away from civilization and it’s raining every day where they live.

Let me guess: Valve was not getting their cut from the ad money?

You can also do a simple content filter. Using feddit.org as instance:

feddit.org##.d-sm-block.d-none > .post-container.row:has(.post-title:has-text(/trump|musk|puppies/i))

This will hide every post which has the words “Trump” OR “Musk” OR “puppies” in the title. Or if you don’t use Apple devices and don’t want to read about their new devices in every tech community:

feddit.org##.d-sm-block.d-none > .post-container.row:has(.post-title:has-text(/apple|iphone|macos|icloud|iOS|ipad/i))

They downloaded the torrents from Annas Archive, which are standing at ~500TB currently. Keep in mind that you’re dealing not only with text, but also with books scanned as images, books with lots of illustrations, scientific articles with illustrations and also comic books.

Be careful there: The big publishers are watermarking the articles you’re downloading. So if you’re uploading them to SciHub, you can get into trouble. Inform yourself and remove the watermarks