Comment by ninkendo

> Otherwise you would have known that the defaults you're complaining about are clearly identified in the README files you didn't read.

None of what you’re saying about my scenario is true at all. You’re hallucinating shit worse than LLM’s. There was no README that outlined the behavior. There was no change log entry. The sqlx crate did not document a single thing related to this at all. You’re missing critical information and making enormous assumptions, and jumping to attacking my character as the very first step.

Maybe you googled just enough about the sqlx and whoami crates to come to a wrong conclusion. Because there was a changelog entry related to whoami in the sqlx release notes. And yes I read it. But guess what: that wasn’t the issue. The issue is a legitimate bug in sqlx related to how they use the whoami crate. There’s no clearly identified README entry. There’s just a sqlx bug in 0.9.

I’d explain in detail exactly what the bug is but you’re not interested because you’ve already made your conclusion about me before you have any information about what happened. (Hint: the whoami crate decided to move the functionality that actually grabs your username behind the `std` crate feature, and made the std feature enabled by default. But the sqlx crate has always included whoami with `default_features = false`. I’ll leave you to guess what happened as a result. Before you answer, you should also learn how cargo treats crate features in a build where the whoami crate is linked multiple times with different features. Once you think you have it figured out, go and find this README entry I should have read.)

Edit: actually, fuck it, I’m sick of being coy. Here’s the sqlx regression: https://github.com/transact-rs/sqlx/pull/4275

Note how it’s still open as of this writing.

Note how it’s not “clearly stated in the README”. It’s a bug. A bug that’s still not fixed.