On ios, we build statically linked executables. This means that if lime uses mbedtls 3, hxcpp is also forced to use it which can cause crashes and other problems, as hxcpp is currently written for 2.28
It was hardcoding VS 2017, but now uses vswhere.exe to find newer Visual Studio versions too.
Newer versions of VS don't actually support UWP JavaScript, though. However, if users have UWP installed in VS, the error message will be clearer about that lack of support.
The `Asset` constructor sets `asset.id` to the asset's path, a sensible default. Of course, if the asset has a name instead of a path, we have to use that instead. Or if the user specified an ID, that's even better.
We just don't want to default to `""`, since then a bunch of assets could have the same ID.
Resolves#1758.
Casting `const char*` → `char*` is bad form, and we'd rather copy the string than do that. However, `GetGlyphIndex()` now takes a const, so neither workaround is needed.
Originally, we forced compilation on x86, presumably because at the time HashLink lacked good 64-bit support. When this support improved, the line was changed to force x64 compilation rather than being removed, which may have been a mistake. Now that there are even more valid architectures, it just doesn't make sense to force one.
The old logic could produce inappropriate results, such as attempting to compile an x86 binary on ARM64, or compiling no binary when "64" is specified on a 32-bit machine. I'd argue that it makes sense to only check the flags when they're supported, and not to bother otherwise.
If you check `paths.exists()` first, then there's no benefit to calling `getPath(id)`: it always just returns `paths.get(id)`. It's more effective to let `getPath()` perform all of its checks.
See pull request for details. Summary:
* Reduce dependencies between CI jobs, allowing more of them to run in parallel.
* Use Linux for as many jobs as possible.
* Merge matching "-ndll" and "-samples" jobs. This way, the latter doesn't have to waste time recreating the former's environment, and the workflow summary is simplified. Downside: this delays package-haxelib.
* Update to a newer macOS version, as the old version massively slowed down Homebrew.
* Skip the HelloWorld test for Mac and iOS, as those two can build slowly under some circumstances, delaying package-haxelib. These could be restored later, if performance seems good enough.
* Add `HAXE_VERSION` environment variable, so future updates will only have to make one change. Unfortunately, doing the same for the Windows/Mac/Linux versions would severely hurt performance, so those remain hard-coded.
* Split up some steps within jobs for better organization.
