Use Hatch for build backend

[facelessuser]

Supersedes #1640

[facelessuser]

I was holding out on this as my Python 3.10 would not work with this change. Turned out that the brew install on my mac was just busted in a lot of ways (I hadn't switched over to 3.10 for this and many reasons). So I finally installed the official distribution, and everything worked...go figure.

I may be merging this soon. I realize this is a new build backend, and there are probably not a lot of people using it, so that does give me a little pause, but it does seem to work quite nicely.

I am curious though about what specifically Hatch offers that others don't. I hadn't looked into build-backends until this point, but there are packages like https://github.com/pypa/flit maintained by pypa. What makes Hatch better than others that have a larger community maintaining them?

[ofek]

I can briefly tell you why we moved at Datadog!

setuptools

Disadvantages:

• Support for pyproject.toml-based metadata declaration is experimental atm
• No support for editable installation standards (PEP 660) so pip install -e ... still requires a setup.py file
• Based on a deprecated standard library module that it now must vendor
• Slow to get things merged since the code quality issue is exacerbated by the fact that it is mostly maintained by just one person

Flit

flit-core is the build backend behind Flit.

Disadvantages:

• No mechanism for extending behavior by design, so we could never do things like versioning with Git tags
• Very opinionated about project layouts since it targets beginners e.g. only supports implicit namespace packages, must define certain data like __version__ inside the root __init__.py file, etc.

Poetry

poetry-core is the build backend behind Poetry.

Disadvantages:

• Persistent unwillingness to adopt new standards if they are deemed suboptimal (see comments on PEP 621 and PEP 665)

Hatchling

• In comparison to setuptools, Hatchling has reproducible builds, supports the modern PEP 660 editable installation standard, has a more intuitive plugin system, and the codebase is an order of magnitude smaller thus making it more auditable and less prone to bugs. For example, the logic to build a wheel resides in this one file.
• In comparison to flit-core, Hatchling is extensible by design. For example, there is already a plugin for setuptools_scm (which despite the legacy naming is actually now decoupled from setuptools) and a plugin for compiling code with Mypyc.
• In comparison to poetry-core, Hatchling never strays from community standards and is in fact eager to adopt new ones (it's the only backend that currently supports PEP 639).

[ofek]

I don't have a comprehensive list of users yet, but more projects switched recently too like tox

[facelessuser]

Hey thanks for the info. Sounds like hatch is the way to go. Thanks for your help and patience.

[facelessuser]

@gir-bot lgtm