I do want to replace these files with real lock files entirely. Given that they can end up with very different resolutions I'm not sure if I want to attempt to have a single requirements based lock file at the moment.

I think there is power in the ability to install the lock files using pip install -r requirements.lock for people cloning a repo and not already having rye installed so replacing these with a different solution might be worse in that sense.

replacing these with a different solution might be worse in that sense

We could standardize pyproject.lock (similar to Cargo.lock), then pip could easily handle it.

@shroominic misusing lockfiles as user-requests (requirements) is problematic when you work with more than one environment (a combination of Python/OS/arch) since they can easily contain transitive dependencies that are incompatible with other envs.

Here I tried to explain how the concept of lockfiles and the idea of what they do is misunderstood in the context of Python specifically: jazzband/pip-tools#1326 (comment)