[ryanrhughes]

Automated update of a package that tracks an upstream vendor release
feed or branch tip rather than the AUR.

Release watches and providers are declared in .omarchy/package.json;
exceptional feeds use .omarchy/upstream.sh. This PR auto-merges once
the build checks pass. A failing build leaves it open for a maintainer
without holding back any other package.

[maralcbr]

Thanks @ryanrhughes. The bump itself is right: 106f944 is upstream's v3.1.1 tag, pkgrel is back to 1, and v3.1.0..v3.1.1 upstream only touches CI, release metadata and the meson wraps, which we don't use because build() passes --wrap-mode=nofallback. The aarch64 build ran for real: both checksums passed and it produced iptsd-3.1.1-1.

self-tests is red, though, and it's required, so this can't auto-merge. Two things stand in the way:

1. pkgbuilds/iptsd/test.sh:14-15 pins the old commit. It checks _commit == a83bc12… ("pinned v3.1.0 commit"), so it fails at this head and will fail on every future sync. Running it locally gives not ok - iptsd builds from the pinned v3.1.0 commit; with master's PKGBUILD it passes. Please check the shape instead, e.g. [[ $_commit =~ ^[0-9a-f]{40}$ ]] || fail "iptsd builds from a pinned upstream commit", and keep the check at lines 16-17 that the source is fetched at _commit.
2. The channels key, which isn't this PR's fault. CI actually stops earlier, in tests/package-metadata.sh: invalid channels for iptsd ... invalid channels for surface-snapdragon-addresses: channels is gone (helpers/package-metadata.sh:488-491). Both packages landed (#826, #972) with "channels": ["edge"] after 61ff008 retired the key, and the test validates every package, so master's self-tests fails the same way and every package PR is blocked by it. That's also why the test.sh failure above doesn't show in CI yet. Could someone drop channels from both package.json files in a separate PR on master? It's already ignored for routing, so nothing changes in where either package publishes.

Smaller thing:

• pkgbuilds/iptsd/README.package.md:3 still says v3.1.0. Suggest version-neutral wording, "built from a pinned upstream release of linux-surface/iptsd", so syncs don't leave it stale. It's installed by the package, so the second entry in sha256sums needs updating with it.

With the test.sh fix in and master's channels fix landed, this is good to merge from my side.

[greptile-apps]

Confidence Score: 4/5

[Low impact] The update reaches users of one aarch64 Surface digitizer package.

The PR is not ready to merge because test.sh still rejects the new upstream commit.

Findings

1.  New pin breaks CI ▶
2.  Documentation names the old release ▶

Summary

Updates iptsd from 3.1.0-2 to 3.1.1-1.

• The Surface digitizer package now builds iptsd 3.1.1.

_{Reviews (41) · Last reviewed commit: "Sync upstream (six-hourly): iptsd 3.1.1-..." · Reviewed by Greptile}

[greptile-apps]

Documentation names the old release

The package now uses v3.1.1, but the installed README.package.md still says it was built from v3.1.0. That mismatch can confuse bug reports. Update the document's version reference and refresh its checksum in PKGBUILD.

[greptile-apps]

New pin breaks CI

test.sh sources this PKGBUILD but still requires _commit to equal a83bc1232f7096f8b33b50fdbda249cd640de670. The new value fails that check and makes the test exit with an error. The PR workflow runs this test, so its self-tests job will fail.

Update the test to match the new release or check the pin without hardcoding the old release.

[greptile-apps]

Greptile has paused reviews on this repository — it used its 750 free open-source review credits for this billing period. Reviews resume automatically on November 21. To continue before then, an organization admin can keep reviews running past the free credits — those bill as normal usage.