@dalias

That sounds a lot like "Forward Secrecy" would cover that. Found https://robertheaton.com/otr2 but don't know if that site is good

@regehr

If it should be compilable for obscure systems like SuperH or PDP-11. Anythink based on llvm probably won't do that https://github.com/llvm/llvm-project/issues/51562

@tuananh

Yes, but except for using the repository name instead of the url the generation seems to be deterministic.

@domi

I have seen the font in my browser recently change as well. Have you set monospace through font config like I did?

https://codeberg.org/sertonix/aports/src/branch/main/custom/0+desktop/font.conf

@domi @filmroellchen

That's one reason why I sometimes package glibc expecting software as an exercise when I don't need to.

@ariadne @dalias

You can assume good faith and still need to check if a downstream service/config/patch file needs to change. Stuff like that is not always in release notes (if they exist at all).

@amonakov @wolf480pl

Hm, upon reading the POSIX spec again it looks like it only guarantees the exit code to be >128 in that when a trap inerrupted wait. I think in practice one can assume that r=$?; [ "$r" -gt 128 ] && [ "$(kill -l "$r")" != CHLD ] returns is true if the not waited on process terminated "first".

But thinking about this more. You can't really deal with pid reuse in shellscripts. Assuming that you can just send a signal to both pids and ignore failure.

@wolf480pl

(If it's not clear what I mean I could try writing a PoC script)

@wolf480pl

You can use wait on one pid + SIGCHLD trap (which will make wait stop on any child exit). If you only have 2 processes it should be easy to determine which finished by checking the exit code of wait.