That’s good in theory but there are also plenty of counter examples of companies forcing features and still making it by just sheer brand reputation or market share (HP still has DRM’d ink, Keurig is still going after “hacks”) or just money (OpenAI promised to open source their model).
I’m not saying we shouldn’t shame those companies for not abiding to their words, but there is more to it than outrage. Suing them (or the threat of) might also work here if they really went against the license.
My theory is either fudging the display (add/remove votes to prevent people from knowing the real count), or there’s a secret API that some bot found and exploited.
Or maybe it uses negative points when it detects bad actors voting/commenting on it? Just spitballing of course.
Maybe there is an API to downvote posts and some bots found it?
Edit: Actually I would wager that it’s some kind of automatic downvoting when the system detects bad votes or comments. Something like “friends shouldn’t vote”. Since we see voting on posts and not comments we could see it go negative.
I suspect there’s downweights the system can apply - and if five people voted it up, the -5 downweight was applied, and then four of the upvotes were removed, you’d end up at -4.
