I can confirm that the crash doesn't occur anymore.

Thanks!

In D15251#322168, @kossebau wrote:

...as well as request to finally enable BUILD_QCH on CI, so this is covered.

Ping?

Indeed! Thanks for the reminder.

Remove unrelated commit

Ping?

Btw, this is BUG: 393999

All changes look fine to me.

I would have clang-tidy & Co handy to do some simple common modernization, if you like (done already a few time elsewhere this year). Just tell me "Go" and I would run nullptr & override at least and commit to master (after manual check) :)

Cool idea and patch!

Ping?

Considering the feedback I'll probably merge this in a few days if no objections turn up in the meantime.

All there for Linux and FreeBSD. Windows is postponed until amarok doesn't need mysqle anymore.

Thanks!

In D12118#262926, @ltoscano wrote:

I'd argue that ki18n_install should be fixed instead. It does not make sense to duplicate the if check for every call to ki18n_install.

In D12118#262921, @ltoscano wrote:

This change should not be needed; ki18n_install is a noop operation if the directory does not exist - did you see any issue?

Ping?

Committed with 91102eee99a1e77f4671d3c1a19dcc203c64c7a3 but apparently I managed to remove the "Differential Revision" line, sorry.

Committed with 09501ab26c4ec6e5608ba2c57c5e3bcdf0f35ab4 but apparently I managed to remove the "Differential Revision" line, sorry.

In D12374#250388, @asturmlechner wrote:

I was able to reproduce the build failure now with Qt-5.9.5, but not on my other system with Qt-5.11.0_beta4...

In D12216#247225, @apol wrote:

I'm unsure if it's worth re-spinning ki18n.

Combined the two find_package(KF5...) calls as suggested.