Releases move from the self-signed CN=unom cert to Azure Artifact Signing (formerly Trusted Signing): account `unomsigning`, profile `unom-io`, signed by the `punktfunk-ci-signing` service principal, which holds only the Artifact Signing Certificate Profile Signer role scoped to that one profile. Both pack scripts gain the backend ahead of the existing .pfx and ephemeral fallbacks, so canary and fork builds are unaffected. Three things that are easy to get wrong, and are handled here rather than discovered in the field: Azure mints a leaf certificate per signing request that expires in about three days. Both scripts previously retried WITHOUT a timestamp when a timestamped sign failed — under Azure that ships an artifact which verifies on the runner and goes untrusted days later, on every user's machine at once. The retry is now gated on the mode: still lenient for a .pfx whose cert outlives the release, a hard failure for Azure. The MSIX manifest Publisher must equal the signer subject byte-for-byte, because package identity is Name + Publisher. The default is now the profile's verified subject, written with `[char]0xFC` escapes rather than literal umlauts so this UTF-8-without-BOM file cannot silently mojibake the DN into one that no longer matches. pack-msix.ps1 now also reads the signature back off the packed .msix and fails on drift — asymmetric on purpose: a subject that disagrees is fatal, a subject that cannot be read is only a warning, since Get-AuthenticodeSignature's .msix support varies by Windows version and signtool has already reported success by then. NOTE this changes package identity, so existing installs need an uninstall, not an upgrade. The updater's leaf-pinning note was wrong and is corrected: update/windows.rs claimed the AUTHENTICODE_SHA256 field made Trusted Signing "a manifest edit", but a per-request leaf is exactly what a leaf pin cannot track — a pin would go stale within days and reject every release after it. Drivers are deliberately untouched: their catalogs keep the DRIVER_CERT_* cert and the installer still plants it as a machine root. The two signatures were always independent (SmartScreen/UAC vs PnP), which is why the installer could move without them. Whether a publicly-trusted catalog would let us drop that root plant is recorded as an unverified follow-up, not assumed. Verified: both scripts parse under the PowerShell 7 AST parser, both workflows are valid YAML, the evaluated Publisher default matches the subject Azure reports for the profile (86 chars, ordinal), rustfmt clean. NOT verified on Windows — the sign path itself needs an on-glass run on .133.
punktfunk Windows client — MSIX packaging
The Windows client ships as signed MSIX packages so Windows boxes get a real package (Start
tile, clean install/uninstall) instead of a loose exe. CI builds + publishes them from
.gitea/workflows/windows-client.yml to Gitea's
generic package registry (https://git.unom.io/unom/-/packages), on every main push that
touches the client (canary) and on vX.Y.Z release tags (stable) — see
Release Channels.
Two architectures, one x64 runner. Both x64 and arm64 packages are produced off the single
x64 Windows runner — x86_64-pc-windows-msvc builds natively, aarch64-pc-windows-msvc is
cross-compiled (the x64 MSVC toolset ships the ARM64 cross compiler; since M10 nothing in the
package links FFmpeg, so neither arch needs a per-arch FFMPEG_DIR tree staged on the runner —
one less thing the ARM64 leg can be missing). Artifacts are arch-suffixed
(..._x64.msix / ..._arm64.msix, each with its matching .cer); pack-msix.ps1 -Arch x64|arm64
stamps the manifest ProcessorArchitecture and names the output. See
windows-client.yml for the cross-build rationale.
What's in the package
pack-msix.ps1 assembles a layout from a cargo build --release and runs makeappx + signtool:
| File | Source |
|---|---|
punktfunk-client.exe |
the release build (the WinUI shell) |
punktfunk-session.exe |
the release build — the Vulkan session client the shell spawns for every stream (sibling resolution, src/spawn.rs). Skia links statically; vulkan-1.dll is a GPU-driver component, never bundled. ARM64 builds it --no-default-features (no Skia console UI) until rust-skia ships aarch64-pc-windows-msvc prebuilts |
Microsoft.WindowsAppRuntime.Bootstrap.dll, resources.pri |
staged by the client's build.rs via windows-reactor-setup::as_framework_dependent() |
SDL3.dll |
auto-staged by the sdl3 crate |
licenses\* |
the project's MIT/Apache texts + the generated THIRD-PARTY-NOTICES.txt (MSIX has no installer EULA page, so attribution ships as files) |
Assets\*.png |
checked-in tile/store logos (rasterized from packaging/flatpak/io.unom.Punktfunk.svg) |
AppxManifest.xml |
the template here, with {VERSION}/{PUBLISHER} substituted |
No FFmpeg DLLs. The client decodes natively since M10 (pf-vkdecode / pf-dxvadec /
OpenH264+rav1d — punktfunk-planning design/client-native-decode.md §6), so nothing here
link-imports libav* and the wildcard avcodec/avformat/avutil/swscale/swresample-*.dll copy is
gone, along with the FFmpeg LGPL notice that accompanied it — shipping that notice now would claim
a dependency the package doesn't have. The host installer is unchanged:
packaging/windows/pack-host-installer.ps1 still ships those DLLs for its AMF/QSV encode path.
Why an "unpackaged" WinUI app packages cleanly
main calls windows_reactor::bootstrap(), which runs MddBootstrapInitialize2 with
OnPackageIdentity_NOOP (crates/libs/reactor/src/bootstrap.rs), so under MSIX package
identity the App SDK bootstrapper is a no-op and the runtime is resolved from the manifest's
<PackageDependency> on Microsoft.WindowsAppRuntime.2 instead (reactor pins
WINDOWSAPPSDK_RELEASE_MAJORMINOR = 0x20000 = 2.0). It's a full-trust Win32 app
(EntryPoint="Windows.FullTrustApplication" + runFullTrust) because it owns raw D3D11, Win32
low-level input hooks, WASAPI and SDL3.
Versioning
MSIX requires a strictly 4-part numeric version. The workflow computes:
vX.Y.Ztag →X.Y.Z.0(THE release; any-rc/+metasuffix is dropped for MSIX). Published to the stablelatest/alias and attached to the unified Gitea Release.mainpush /workflow_dispatch→0.3.<run_number>.0(canary, climbs by run number;canary/alias).
Signing & install
CI signs every build with Azure Artifact Signing (formerly Trusted Signing) — account
unomsigning, certificate profile unom-io, endpoint https://neu.codesigning.azure.net/. That
chain is publicly trusted, so there is nothing to import:
# install the package for your CPU (and re-run for each upgrade)
Add-AppxPackage -Path .\punktfunk-client-windows_<ver>_x64.msix # Intel/AMD
Add-AppxPackage -Path .\punktfunk-client-windows_<ver>_arm64.msix # ARM64 (Snapdragon, etc.)
The MSIX declares a dependency on the Windows App SDK 2.x runtime; install
the App SDK runtime if Add-AppxPackage reports a missing
Microsoft.WindowsAppRuntime.2 framework.
How signing resolves
pack-msix.ps1 picks a backend in this order:
- Azure Artifact Signing when
AZURE_CODESIGNING_ENDPOINT/_ACCOUNT/_PROFILEare all set (the workflow sets them; they aren't secret). Credentials come fromAZURE_TENANT_ID/AZURE_CLIENT_ID/AZURE_CLIENT_SECRET— thepunktfunk-ci-signingservice principal, which holds only the Artifact Signing Certificate Profile Signer role scoped to theunom-ioprofile. Keys are HSM-backed and never leave Azure, so there is no.pfxand no.ceris emitted. MSIX_CERT_PFX_B64/MSIX_CERT_PASSWORD— the older stable self-signed cert (CN=unom, public half checked in aspunktfunk-codesign.cer), kept as a fallback.- An ephemeral self-signed cert (forks / local builds with no secrets at all).
Modes 2 and 3 still export a .cer to import into Cert:\LocalMachine\TrustedPeople first. On a
v* tag, a build with no real signing backend fails closed rather than shipping a throwaway.
Two things about Azure mode that are easy to get wrong:
- Timestamping is mandatory, not best-effort. Azure mints a leaf cert per request that expires in about three days. An untimestamped signature therefore stops verifying within days of release, so the script refuses to retry without one (modes 2 and 3 keep the old best-effort retry).
- The manifest
Publishermust equal the signer's subject exactly, because MSIX package identity is Name + Publisher. The default-Publisheris theunom-ioprofile's verified subject; after signing, the script reads the signature back off the.msixand fails the build on any drift. Changing it makes a different package — existing installs must be uninstalled, not upgraded.
Building locally
On the Windows runner / dev VM (MSVC + Windows SDK present), after a release build:
# x64
cargo build --release -p punktfunk-client-windows --target x86_64-pc-windows-msvc
pwsh -File clients/windows/packaging/pack-msix.ps1 `
-Version 0.2.0.0 -TargetDir C:\t\x86_64-pc-windows-msvc\release -OutDir C:\t\msix
# arm64 (cross-compiled; no extra environment — the client links no FFmpeg)
cargo build --release -p punktfunk-client-windows --target aarch64-pc-windows-msvc
pwsh -File clients/windows/packaging/pack-msix.ps1 `
-Version 0.2.0.0 -Arch arm64 -TargetDir C:\t\aarch64-pc-windows-msvc\release -OutDir C:\t\msix
Validated end-to-end on the build VM (pack → sign → Add-AppxPackage → framework-dependency
resolution). The only step that needs a real display is launching the WinUI window (same
on-glass constraint as the rest of the client).