borrowck changes

[Mark-Simulacrum]

This issue may need splitting up into smaller issues:

https://crater-reports.s3.amazonaws.com/beta-1.39-1/beta-2019-09-28/gh/eggmund.Prey/log.txt
https://crater-reports.s3.amazonaws.com/beta-1.39-rustdoc-1/beta-2019-09-28/gh/PEshwar.rust_blockchain/log.txt (unrelated to rustdoc)
https://crater-reports.s3.amazonaws.com/beta-1.39-rustdoc-1/beta-2019-09-28/gh/djcsdy.gfx-playground/log.txt (unrelated to rustdoc)
https://crater-reports.s3.amazonaws.com/beta-1.39-rustdoc-1/beta-2019-09-28/gh/efb9-860a-e752-0dac.rustchal/log.txt

cc #64962

[ecstatic-morse]

The errors in eggmund.Prey are a result of NLL errors becoming fatal. They existed as warnings on 1.38. This is expected breakage.

[ecstatic-morse]

Regressions 2 and 4 are the same, occurring in soketto-0.2. I partially minimized the problem. The trait object is necessary. This is a true regression, although I think this code should never have been accepted on stable?

use std::any::Any;

fn drain_dyn(v: &mut Vec<Box<dyn Any>>) -> impl Iterator<Item = Box<dyn Any>> {    
    v.drain()
}

See #65582.

[ecstatic-morse]

Regression 3, in gfx-rs, is also a result of NLL errors becoming fatal. This is expected breakage.

[ecstatic-morse]

@Mark-Simulacrum I've triaged all four regressions and opened an issue for the only unexpected one. I'll close this issue in favor the new one.

[imclint21]

Hi @Mark-Simulacrum,

We own the repo efb9-860a-e752-0dac do you know why this log file contains our repository rustchal?

=>https://crater-reports.s3.amazonaws.com/beta-1.39-rustdoc-1/beta-2019-09-28/gh/efb9-860a-e752-0dac.rustchal/log.txt

[Mark-Simulacrum]

Crater queries GitHub for all Rust repositories (I'm not sure on specifics of how GitHub determines that). We then build all of those repositories on our infrastructure.

[imclint21]

Oh very interesting, thank you 👌