Dalton Messmer

We recently [updated our Wine dependency](https://github.com/LMMS/lmms-ci-docker/pull/17) to use WineHQ which is currently at version 9.9. Closing this as completed.

It looks like this issue is present upstream, so I just contacted Tim at http://quitte.de/dsp/caps.html#Contact.

Rather than updating our MinGW image again, I think I'd like to try a proper solution, which would be using a plain GitHub build runner and getting our MinGW dependencies...

I've been experimenting with MinGW-w64 (with POSIX threads) + vcpkg on GitHub build runners here: https://github.com/messmerd/lmms/tree/mingw-vcpkg-testing It's been rough and I haven't been able to build all of the dependencies...

I think we should spend a little more time and refactor the `NotePlayHandleManager` class. Maybe use `std::vector` for storage.

I believe the data structure we'd want would be a "hive" or "colony" like the one [proposed for C++26](https://wg21.link/P0447). It's an unordered container which ensures stable memory addresses for elements...

@DomClark Do you have any ideas why the Linux build is failing on the Configure step now? It probably has to do with caching but I'm unfamiliar with how that...

@DomClark Thanks, that did the trick

> Someone recently said they like that we use containers in the CI because it makes it easier for them to just pull the container and directly compile LMMS I...

I removed caching of the system packages since they were contributing the most to the high cache usage. Fortunately this didn't negatively affect the build times much and the cache...