For the benefit of reviewers - #8673 implements the @rpath removal approach mentioned in the PR description.

Gentle nudge on this (and/or #8673) - is there something preventing a merge on one of these two fixes for #8671?

@freakboy3742 Presumably either this or #8673 can go in the next release for which the deadline to merge would be end of March.

Thanks for the PRs! Yeah, we'll definitely include one or the other in the next release. I don't see too much difference so will flip a coin unless someone would like to argue for one over the other :)

Thanks for the PRs! Yeah, we'll definitely include one or the other in the next release. I don't see too much difference so will flip a coin unless someone would like to argue for one over the other :)

I'd argue the #8673 is better on the basis that it fixes the problem right next to the place where the problem is created, rather than relying on remembering later that a fix is needed for one specific library case. However, that's a strong opinion, weakly held; both solutions work.

In terms of timing - while it's great this will be in the next release, the bug prevents testing of macOS builds on a local machine, so I'd argue it would be beneficial to merge it sooner rather than later. In order to test my iOS patch, I need to have include a version of this patch.

I've just submitted #8743, which I believe is an even better solution, as it fixes the problem at the source; and, with any luck, eventually the patch used by that PR won't be needed.

#8673 has been merged.