Andreas Klöckner
Andreas Klöckner
Reproducible with 9cf3654 (using Debian's llvm/clang 1:18.1.8-13).
Thanks for investigating, and sorry about the red herring here!
Thanks for the detailed analysis! The real solution to this will be to stop relying on a hacked-up Boost that's a pain to upgrade, see #393 for work towards that...
Thanks! A release with this is forthcoming.
I would be happy to consider a PR along those lines.
Thanks for doing this. All this is definitely intended to "just work", at least eventually.
What made you decide to close this? I figure we'll want something like this in the medium term?
Potentially, we could `fork` before each test? This avoids paying for the interpreter startup cost repeatedly. I don't know that there is a reliable way to reset a monkeypatched module...
`fork` would also offer a way to parallelize runs.
I also just encountered this issue, on a self-hosted instance. https://github.com/huginn/huginn is a good example of a webhook receive that will, by default, return a code of 201.