I will update the PR following this comment.

This needs another rebase.

This needs another rebase.

Done

This is the second time I rebased my PR on master after more recent PRs than mine were merged. What guides the order in which PRs are merged?

Thanks!

This is the second time I rebased my PR on master after more recent PRs than mine were merged. What guides the order in which PRs are merged?

There's no specific order, but this PR is quite big, which is why it's harder to review.

Networking-Tests are looking good.

/nix/store/4y8k9isgglvhycvq59kg50hx62rw08kn-vm-test-run-Bond-Networking-Networkd
/nix/store/0kmcb3mwlm63hjkbd50h075d61ncqf0i-vm-test-run-Bridge-Networking-Networkd
/nix/store/dcz18v4qz82sx9ynr3wyqimzrmaigpgg-vm-test-run-OneInterfaceDHCP-Networking-Networkd
/nix/store/1g9p0kgpd8nzbv6kqkkcq08nh8hmvmxi-vm-test-run-SimpleDHCP-Networking-Networkd
/nix/store/3014l21l3qrg01s3x33ys08fgjqrs9lw-vm-test-run-Link-Networking-Networkd
/nix/store/jllmqcvalzn1imsb6x0r57dfcdw2v7cz-vm-test-run-Loopback-Networking-Networkd
/nix/store/431izfl3mb82crr6kaw143ky0r04c9yl-vm-test-run-MACVLAN-Networking-Networkd
/nix/store/p7vj2zfcl228a9p5ri27ycrdm1yix6v9-vm-test-run-Privacy-Networking-Networkd
/nix/store/fs1rk6x92h83gfm1f8pm4ps8158g17ph-vm-test-run-routes-Networking-Networkd
/nix/store/wvh4cha8ssx3ypplwqp75wf1phiy46yx-vm-test-run-Sit-Networking-Networkd
/nix/store/blbh606jdzsbgdadbavj8nl0ic00c4d0-vm-test-run-Static-Networking-Networkd
/nix/store/mgv2b09h658glqn8lgz2s3akxdwzf7d7-vm-test-run-Virtual-Networking-Networkd
/nix/store/8kfj6r2kh6ab2w5g6ix304fr21zlw2vz-vm-test-run-vlan-Networking-Networkd

I tried reviewing this, but it's really hard to do.

e9d13d3 added some commented-out checks due to Nix' <= 2.2 integer size, and 70407f0 commented them back in. Could this be squashed together, so we only add the checks in a single commit?

As much as I could try, each of these 2 commits do one logical thing and one thing only, as described by the commit messages

• e9d13d3: update options for systemd 245. That commit synchronizes the code with the man page of networkd and, for homogeneity with the rest of the code, keeps avoiding the use of assertRange with 64bits integers.
• 70407f0: use assertRange with 64bits integers. This commit takes into account this comment. Perhaps it would better fit as a separate PR after this one would have been merged. Since the PR is about cleaning up the networkd module, I included this commit here.

I apologize if my commit messages were not clear enough.

Also, e9d13d3 does a bit too much at once. For example, it removes OriginalName. Is this intended?

Again, that commit synchronizes the code with the man page of networkd. The best way to review that I would suggest is to read the man page. For the specifics of OriginalName, refer to man systemd.link. I also try to double check by contacting the author here, without success. However, you kind of confirmed it yourself here. To facilitate our work, let's discuss specific code issues/questions with the review feature of Github.

Could you revise the commits, and make it these changes easier to review? Maybe something like this also deserves a separate commit.

I do not think it is beneficial because the commit does one clear logical thing. If you disagree, please describe what you mean by "these changes" because, as you may have noted, this is not the only thing I removed in this commit e9d13d3.

Thanks again for your patience! I'll try to react quicker on further updates.

Thank you as well for reviewing this.

Could we move on with this PR?

I gave this another close look, and while it is pretty hard to review due to the amount of changes, this seems to be fine, and should make things more readable and understandable in the longer run.

Thanks for the work you put in this, and your patience with me :-)