Ron Yorston

The issue with alternate screen buffers appearing entirely black has been fixed in Wine 8.0-rc4 (January 2023). See Wine bug [54287](https://bugs.winehq.org/show_bug.cgi?id=54287) for details.

The problem is in the upstream TLS code. This is less of an issue there as the default build uses OpenSSL. When upstream BusyBox is built to use the non-default...

My understanding of the TLS code remains insufficient, especially since I've now determined that the problem seems to be in some x86_64 assembly language code, in which I'm less than...

I've figured out the problem with `files.catbox.moe`: its server certificate uses an elliptic curve (P384) which isn't supported by upstream BusyBox. Upstream's TLS implementation only handles P256 and X25519.

Here's another (closed) [OpenSSH issue](https://github.com/PowerShell/Win32-OpenSSH/issues/1652) on the matter. (EDIT: and here's [another one](https://github.com/PowerShell/Win32-OpenSSH/issues/1713) closed with "Resolution - By Design".) Microsoft's point of view appears to be: - It isn't possible...

The most recent [prerelease binaries](https://frippery.org/files/busybox/prerelease) include the new `runuser` applet. This is based on the application of the same name from util-linux, though it has somewhat reduced functionality. Basically it...

Being compatible with util-linux `runuser` comes with some limitations. The latest [prerelease binaries](https://frippery.org/files/busybox/prerelease) introduce `drop` as an alternative to `runuser`. - `drop` works even if the current process doesn't have...

More churn: - `runuser` has been removed. It may be the closest existing utility to the required functionality but it isn't a very good fit. - When `drop` is used...

I think we now understand what's going on. It's clearly a Microsoft issue but the latest busybox-w32 release has a mitigation for the problem in the `drop` applet.