18:11 <satacker[m]> K-ballo: can you please explain why clang cannot recognize this overload while gcc does https://godbolt.org/z/q1hMrfKjv ? i think this might be a clang issue

21:08 <gnikunj[m]> satacker: remove the `::` in `::delete[] p2` and it will work. Funny how gcc can understand it but clang can't. Clang defaults to system implemented delete when trying to access delete through global namespace while gcc do not. I'm sure K-ballo will have interesting remarks.

21:08 <gnikunj[m]> In other news, TIL: https://stackoverflow.com/questions/51052610/error-with-c-syntax-compiler-doesnt-warn-or-error-for-int-v-funcv

21:42 <hkaiser_> gnikunj[m]: this is not the issue, it's because of the missing overload for the two parameter version

21:43 <hkaiser_> and we can't use the non-qualified delete() as this would possibly invoke the wrong overload (the type specific one)

21:44 <hkaiser_> gnikunj[m]: https://github.com/STEllAR-GROUP/hpx/blob/master/libs/core/type_support/include/hpx/type_support/generator.hpp#L327

21:55 <hkaiser_> pansysk75[m]: https://app.circleci.com/pipelines/github/STEllAR-GROUP/hpx/14839/workflows/73b98320-e38c-4371-b50e-057e6604e39b/jobs/346563 :/

21:56 <hkaiser_> could be a genuine timeout, though

21:57 <hkaiser_> let me increase the timeout threshold

22:14 <pansysk75[m]> hkaiser: yep, it's just a threshold thing, no worries

22:15 <pansysk75[m]> I reduced the workload on that test on some PR that hasn't been merged yet