Daniele Nicolodi

This problem arises from the Python packaging ecosystem decision of treating binary or source installs as equivalent and transparently fall back to compiling from source when a binary package is...

> it doesn't look at memory usage and that is the problem here This is exactly the reason why I don't think that an heuristic based on the number of...

Another possible solution is to allow variable substitutions in the arguments passed to meson. This way it would be possible to write something like: ```toml [tool.meson-python.args] install = ['-j', '%(ncores)s']...

> Interesting idea. I think I like it. Maybe it's simple enough if we only allow `*`, `/`, `+`, `-` and Python scalars (or integers)? This is another demonstration that...

> Interesting idea. I think I like it. Maybe it's simple enough if we only allow `*`, `/`, `+`, `-` and Python scalars (or integers)? I've implemented the parsing and...

I actually just realized that supporting the `%(ncores)d` format is even easier. I'm opening a PR with a draft implementation.

Does it work if you build a regular wheel? I don't think there is a way for meson-python to support cmake or autotools subprojects: there is no way to map...

By the way, what's the purpose of building an editable wheel for running the tests in CI? The only advantage of editable build is to have the package automatic rebuilt...

To clarify: meson-python assembles a Python wheel from the files installed by the Meson project. To do so it uses introspection data exported by Meson and maps installation locations into...

I must take back some of the considerations above. I just put together a simple test package that installs an extension module that links to a shared library compiled as...