I’m finally learning to use PackageCompiler, things seem to be working great so far. I just noticed that at the “creating system image object file” step building an app for my project there seems to be no parallelism. I assume there may be chances to parallelize stuff there, not unlike a “make -j6”… Is it actually not simple to parallelize this, am I missing some configuration, or is it just a feature request that nobody has worked on it yet?
It is not simple. It uses the same mechanism as Julia does to compile itself and that is not multithreaded either.