nixpkgs: Refresh channels and inputs

Rebase the fork branches and refresh nixpkgs, home-manager, and the
approved ancillary inputs. Update kernel and release metadata, adapt
removed package and Home Assistant options, and keep Determinate Nix
on its tested nixpkgs revision to avoid duplicate Boost patches.

Retire Sharry and its public endpoint because copyparty replaces it.
Document the GitHub mirror gate and require real devshell and system
builds in the upgrade validation workflow.

docs/nixpkgs-upgrade.md

@@ -72,6 +72,12 @@ For **both** branches — `devplayer0` onto `upstream/nixos-unstable`, and `devp
    conflicted.
 6. ⏸ **Push:** only after confirmation. `git push --force-with-lease origin devplayer0
    devplayer0-stable` (force needed — rebase rewrites history).
+7. Wait for the GitHub mirror used by the flake inputs to catch up with the primary fork remote.
+   Compare the local branch tips with
+   `git ls-remote https://github.com/devplayer0/nixpkgs.git refs/heads/devplayer0
+   refs/heads/devplayer0-stable` and do not continue until both match. Updating sooner can leave
+   `nixpkgs-mine` and `nixpkgs-mine-stable` pinned to the pre-rebase commits even though the push
+   succeeded.
 
 ## Phase 3 — Update the pinned inputs
 
@@ -126,7 +132,7 @@ Don't blanket-update. Walk the other inputs deliberately:
 
 1. List inputs and locked revisions from `flake.lock` (or `nix flake metadata`).
 2. For each meaningful input (`libnetRepo`, `devshell`, `determinate-nix`, `ragenix`, `deploy-rs`,
-   `impermanence`, and the packaged apps like `boardie`, `harmonia`, `copyparty`, `sharry`, …),
+   `impermanence`, and the packaged apps like `boardie`, `harmonia`, and `copyparty`),
    compare the locked revision to upstream and summarize notable changes (breaking changes,
    relevant fixes). Many inputs `follows` `nixpkgs-unstable` and already moved in Phase 3.
 3. Propose a per-input update list with reasons; update the approved ones with targeted
@@ -137,7 +143,14 @@ Don't blanket-update. Walk the other inputs deliberately:
 1. `nix flake check --no-build` (broad eval; reproduces CI's cheap checks).
 2. `check-system <host>` on a representative box, and one exercising the stable channel if the
    boxes mix channels. This must exercise the refreshed kernel constants on both channels.
-3. Report eval/build results honestly. On failure, surface the error and stop rather than papering
+3. Build the actual devshell with
+   `nix build --no-link --print-out-paths .#devShells.x86_64-linux.default`. Evaluation does not
+   build its dependencies, so it cannot catch packaging conflicts introduced by inputs such as
+   Determinate Nix.
+4. After the evaluations pass, run `build-system <host>` for one representative NixOS box. Prefer
+   the local box when it is managed by this flake: its full closure is likely to exercise the most
+   relevant packages, home-manager configuration and upgraded kernel. Build only; do not switch.
+5. Report eval/build results honestly. On failure, surface the error and stop rather than papering
    over it.
 
 ## Wrap-up