msmtp: resholve queue scripts

[peterhoeg]

Description of changes

Previously, we were doing 3 things to the queueing scripts (msmtpq and
msmtp-queue):

1. replacing binary invocations with @binary@ so we could use
   substituteInPlace and friends to set the proper full paths
2. make queueing directory and log file overridable via environment variables
3. add support to the logging function to log to systemd's journal

The handling of the queue scripts (item 1) was very brittle and didn't catch all
impurities (which for one), so instead use resholve to ensure all paths to
binaries are resolved to their proper paths. That also means less patching.

Item 2 - the connectivity test is now configurable via an environment variable
rather than being forced at build-time.

Items 3 isn't changed.

Additionally, the package has been split into 2 - 1 for binaries and one for
scripts to make it easier to change the scripts without having to rebuild
msmtp fully every time.

Fixes: #185164

Cc: @Jayman2000

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandbox = true set in nix.conf? (See Nix manual)
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 22.11 Release Notes (or backporting 22.05 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
  • (Release notes changes) Ran nixos/doc/manual/md-to-db.sh to update generated release notes
• Fits CONTRIBUTING.md.

[Jayman2000]

I would just not set EMAIL_CONN_TEST at all. According to the connect_test() function, not setting EMAIL_CONN_TEST is supposed to be the same as setting EMAIL_CONN_TEST to p. The line that set it to n just recently got removed from the upstream repo.

[peterhoeg]

With the recent upstream changes, that makes sense - also means patching less which is definitely a plus!

[Jayman2000]

Personally, I would prefer it if we stuck to using the tarballs. I haven’t checked, but I would guess that downloading the tarball is going to use less bandwidth than even a shallow clone. More importantly, it would be more decentralized which I think is important. That being said, GitHub is probably more reliable, so this isn’t that big of a deal.

[peterhoeg]

fetchFromGitHub and friends are pretty efficient as they do fetch tarballs (or zips - don't recall) and with the upcoming changes to nix itself, it will support not having to actually unpack the archive but instead continue to work directly from the downloaded archive.

GH is less likely to be unavailable compared to $SOME_RANDOM_GUYS_OWN_SITE - that's really all.

[peterhoeg]

@GrahamcOfBorg build msmtp

[peterhoeg]

@Jayman2000 are your concerns addressed?

[Jayman2000]

@Jayman2000 are your concerns addressed?

Yes.