Luca Bruno

Speaking of `async`, I do wonder whether it makes sense to have a dedicated `async_fail_point!()` instead. I currently see a mismatch with actions like `sleep` and `yield` in an async...

@BusyJay nice! A bunch of high-level questions: * can the thread-group logic be folded into [`FailScenario`](https://docs.rs/fail/0.3.0/fail/struct.FailScenario.html)? `register_current` doesn't seem to be very different from threadgroup-bound `setup`. * would it be...

@seanmonstar in case you were looking for feedback, I definitely prefer the first option because it allows to leave the scheme and host-fqdn part untouched. This is important if at...

Gentle ping to @mitsuhiko and @rust-lang-nursery/failure. I'm also again in need of this and it would be better if re-opened.

@mitsuhiko I am not sure there is a way to do that, thus this was mostly so that it doesn't get forgotten for next breaking release. Speaking of that, given...

I think I've hit this on GH (`Tests / Debian Stable with curl, sign-ed25519 and no gpgme` in https://github.com/ostreedev/ostree/pull/2421) with a slightly different output (a refcount assertion): ``` ERROR: tests/test-concurrency.py...

I'm not intimately familiar with `eos-updater` flow, so please bear with me if my questions below are imprecise. >I finally got around to changing our updater to use staged deployments...

Regarding this logic, https://github.com/ostreedev/ostree-rs/pull/33/files sketches an alternative/different implementation (which I think I'm liking more than status quo) along the following lines: * a dedicated internal `repo` reference, nullable * explicit...

I made a pass on this and ported back some of the enhancements from the Rust side, PR at https://github.com/ostreedev/ostree/pull/2454. I also looked into moving this guard to a gobject...

From an out-of-band chat, it would still be preferable to have auto-generated bindings for this. Going back to a boxed type for this, we can add internal reference counting to...