2.090beta revealed two unittests, which assumed real != double. I fixed this with PRs [1] and [2]. Having had an outdated local copy I thought the original PRs were not yet in stable and therefore I based these two on master. But unfortunatley I should have based them on stable. Now I don't know if I can do anything about it anymore... It probably won't help to add two more PRs with the same content, but based on stable. Or am I wrong? What should I do?

[1] https://github.com/dlang/phobos/pull/7328
[2] https://github.com/dlang/phobos/pull/7329

On Thursday, 26 December 2019 at 10:39:59 UTC, berni44 wrote:
> 2.090beta revealed two unittests, which assumed real != double. I fixed this with PRs [1] and [2]. Having had an outdated local copy I thought the original PRs were not yet in stable and therefore I based these two on master. But unfortunatley I should have based them on stable. Now I don't know if I can do anything about it anymore... It probably won't help to add two more PRs with the same content, but based on stable. Or am I wrong? What should I do?
>
> [1] https://github.com/dlang/phobos/pull/7328
> [2] https://github.com/dlang/phobos/pull/7329

No worries. You can still target them for stable. Simply cherry-pick the merge commit, so that it doesn't lead to conflicts later.
Something like this:

git checkout upstream/stable
git cherry-pick -m1 <merge-commit>

And then open a PR as usual.

Thanks!

On Thursday, 26 December 2019 at 10:51:19 UTC, Seb wrote:
> No worries. You can still target them for stable. Simply cherry-pick the merge commit, so that it doesn't lead to conflicts later.
> Something like this:
>
> git checkout upstream/stable
> git cherry-pick -m1 <merge-commit>
>
> And then open a PR as usual.

I hope I got it right. :-) Thanks a lot!