alt Hacker News> So your best bets is to keep backups of your data soemwhere else. Which you should always be doing.
Generally backing up a git repo is easy; it's all of the other stuff that tends to get hosted together with the code (issue tracking, pull requests, releases, potentially even CI) that's a lot worse to have to migrate over. Plus, anywhere a link was published to the repo is going to point to something that no longer exists.
It actually sounds like we're in agreement for the most part about the relative levels of risk here. I misunderstood your earlier reply to me as a rebuttal to what I was saying to the parent commenter, but from re-reading it I'm realizing I might have inferred incorrectly.
We are. My main point was that it's not worth it to disparage others because of their rule. Unless there's a contract or other legal clause, everyone is free to do whatever with their stuff. It's wholly a matter of trust. And I never trust any online service to be permanent in their behavior. I just enjoy it while it lasts.
> it's all of the other stuff that tends to get hosted together with the code (issue tracking, pull requests, releases, potentially even CI) that's a lot worse to have to migrate over. Plus, anywhere a link was published to the repo is going to point to something that no longer exists.
I think most of them provides API endpoint to access those data or some export tools. And as far as links go, I would prefer to have a webpage for any relevant projects, something like the one for msmtp [0], tied to my own domain.
[0]: https://marlam.de/msmtp/