Thomas Eizinger

I thought about this some more and what we currently have starts making more sense now. Despite it being called `Array`, I didn't really think of it is having regular...

> this is a compelling argument against this change. now to retain ergonomics we're having to do this error-discarding nonsense. is it better? I'll let @alessandrod adjudicate. In my eBPF...

> My beef here isn't ergonomics, it's that the error is being discarded. I am not sure I am following. You say your beef isn't ergonomics but then previously you...

> The use of the private syscalls is now behind a non-default feature. Thank you for making it optional! It can be an issue to toggle code-paths like these with...

Found something that got me further: https://github.com/mozilla/rust-android-gradle?tab=readme-ov-file#specifying-paths-to-sub-commands-python-cargo-and-rustc. We can specify: `RUST_ANDROID_GRADLE_CARGO_COMMAND` and `RUST_ANDROID_GRADLE_RUSTC_COMMAND`.

Yep, it is one system-call away! https://learn.microsoft.com/en-us/windows/win32/api/sysinfoapi/nf-sysinfoapi-getproductinfo I'll see if I can draft a PR for this.

The version is covered already by the build number! Windows 10.0.26100 is for example (confusingly) Windows 11. The only missing bit is Home, Pro, Ultimate etc

I am opening this as a starting point for a discussion. The implementation is purposely very minimal for now. I am happy to extend with comments, consts etc. I wanted...

I checked all of these threads on a high-level and none of them looked like they would consume a lot of stack-size. ~So I set it to 100KB for all...

> I think this is reasonable. As you commented, running a full tokio runtime for usage with `reqwest` is still quite heavy on stack usage though. 500KB seems to allow...