The military censor in Israel does the exact same thing: ostensibly to prevent enemies from using the data to improve their systems, in reality as an attempt to keep domestic morale high (it only ever manages to slow down the inevitable fall, though).

There’s a morpheme boundary here, probably has something to do with it. The examples in the post have no morpheme boundary before the main stress, or at least not one that’s transparent to English speakers (ab/solu/te/ly might hypothetically have been more transparent to a Latin speaker though)

Not really, it’s still fully FOSS, they were just terrible at communicating what they actually wanted to do and people got spooked.

Yes, but it’s usually very subtle (e.g. in realizations of single phonemes or in intonation). There are also more extreme cases which other commenters have pointed out.

I recommend you look up sociolects and sociolinguistics.

It’s the other way around: RHEL is a corporate fork of Fedora.

Also, for people using some Readarr derivative with Hardcover metadata, how much of a pain is it to migrate from Goodreads to Hardcover (and is it worth it)?

Calibre-Web has always been interesting to me. Can it be deployed in such a way as to keep a Calibre content server also accessible? (e.g. for sync with the desktop app/Koreader/etc.)

Yeah, unfortunately. Apparently it was hell to maintain, especially the metadata server and all.

It was never about skin colour though, was it?

Was reminded of this article if you’ve got some time to read