On Sun, 2019-10-13 at 16:37 +0800, Paul Wise wrote: > I managed to get a little bit further in the bisect using the attached > script but I hit a point where the commit isn't easily backportable, > I'll try to continue the backporting another day, spent too long today.
I thought about it a bit more logically and I went back to the first commit that added support for using IO::Pty::Easy to supervise the output of the command and added a new commit to make it just use read instead of read and is_active. The result was that I could reproduce the bug even with that commit. So when running on bullseye the issue appears to have been there from the beginning. I guess the next step would be to bisect the dependencies between stretch and buster, so I guess that is basically Perl only? -- bye, pabs https://wiki.debian.org/PaulWise
signature.asc
Description: This is a digitally signed message part
