Benjamin Saunders

Note that LFS is widely disliked due to half-baked support from GitHub (small quotas, made worse by being consumed by [both third-party forks and CI activity](https://docs.github.com/en/repositories/working-with-files/managing-large-files/about-storage-and-bandwidth-usage)) and even less support...

> it does not create a journal entry each time a span is entered This is already the only supported behavior.

Currently the `MESSAGE` field is always exactly the tracing message string. A change to allow customizable formatting would be welcome, but hasn't been drafted.

Oops; I had read the email notification which did not contain your edits. The strategy you outline for testing window growth makes sense to me, though for completeness I would...

> the stream is deliberately made `!Send` for future compatibility with AAudio What does this have to do with giving your callbacks ownership of an `Arc`?

There are no platform-specific concerns. You communicate information from the stream callback the same way you communicate information *to* it: by passing it through shared memory. Stick an `AtomicBool` in...

You can use whatever inter-thread notification mechanism you like, so long as you don't block your stream callback for too long. That's what CPAL would have to do to implement...

+1 to @alexlian's comments on the need to walk a careful line here. A meaningful part of the appeal of buck2 for me is working around cargo's tendency to let...

> Secondly, feature flags may not technically exempt us from semver compatibility in releases, even if the feature flag is documented as such I don't think we need to overthink...

Can someone summarize what exactly the allocation patterns we need are? My usual complaint is that stuff like `gpu-allocator` is a general purpose allocator, which is *much* more complex, slow,...