From 0d522cf2c27aa864ad8c764c7148217f52f4e1d7 Mon Sep 17 00:00:00 2001 From: Claude Date: Thu, 17 Sep 2026 15:07:39 +0000 Subject: [PATCH 1/8] fix: ship device firmware, Files, and the desktop hardware units Bare-metal testing on a ThinkPad X230 found gaps of the same shape: a package installed with nothing enabling it, or a package the Hummingbird base never had and no manifest asked for. Reported with journal evidence and a Bluefin side-by-side on the same machine in #97, #98, #99, #100 and #104. Every name below was checked against the two repositories the image actually enables -- the digest-pinned factory metadata layer from Containerfile, and public-hummingbird-x86_64-rpms -- by reading their repodata directly, because this is exactly the change where a plausible-looking package name is the failure mode. packages/utah.toml gains a [hardware] section: linux-firmware, iwlwifi-dvm-firmware, iwlwifi-mvm-firmware, iwlwifi-mld-firmware, iwlegacy-firmware linux-firmware alone would NOT have fixed #97, and that is the trap worth recording. It is a split package here: it Requires only linux-firmware-whence, and it Recommends thirteen device packages -- amd-gpu, atheros, brcmfmac, cirrus, intel-gpu, mt7xxx, nvidia-gpu, nxpwireless, qcom-wwan, realtek, tiwilink and friends -- which weak dependencies bring in for free. Intel wireless is in neither list. Measured against the pinned repository's filelists, iwlwifi-6000g2a-6.ucode -- the exact blob the report names -- iwlwifi 0000:03:00.0: Direct firmware load for iwlwifi-6000g2a-6.ucode failed with error -2 iwlwifi 0000:03:00.0: no suitable firmware found ships in iwlwifi-dvm-firmware. Installing linux-firmware and stopping there would have closed the issue while leaving the radio just as dead. The three iwlwifi generations plus iwlegacy cover Intel wireless from the 3945 era through current parts; everything else arrives through the recommends. microcode_ctl was in the first draft of that section and is now deliberately absent: it exists in neither enabled repository, so naming it would have failed the install transaction outright. CPU errata updates remain a real gap on this base and need a factory recipe first. nautilus joins [gnome], which turns out to need no new factory work at all: it is already in the pinned factory repository at 51~beta-1.hum1.bfin. #100 reported the accessories shipping orphaned -- nautilus-extensions, nautilus-python and nautilus-gsconnect installed with no /usr/bin/nautilus for them to extend -- and the cause was simply that nothing asked for it. Bluefin does not list it either, because Silverblue carries it in the base. gnome-disk-utility and ptyxis are in neither repository and stay open. avahi joins [services]. Hummingbird ships avahi-libs and avahi-glib but not the daemon -- confirmed at 0.9~rc2-8.1.hum1 in their repository -- so geoclue retried forever against nobody (#104): geoclue[2327]: Failed to connect to avahi service: Daemon not running taking mDNS resolution and WiFi-based location with it. [hardware] is a new section rather than more names under [services], because firmware is not a service and this file is careful about that distinction. Sections are enumerated rather than globbed, so install-packages.py has to opt in or the section is silently ignored -- and verify-rpm-contract.py has to as well, or firmware could go missing from a built image with the contract check none the wiser. Both are wired up. configure-services.sh enables bluetooth.service (#98: bluez 5.87 and bluetoothd installed, bluetoothctl shows no adapter at all), input-remapper.service (#99: the GUI cannot reach a root daemon and falls back to a password prompt) and avahi-daemon.service. All three are also added to 85-utah-desktop.preset, which is the part that makes them stick. bootc applies presets on first boot, so a build-time systemctl enable with no preset line behind it is undone -- the trap configure-services.sh's sshd branch already documents. #98 confirmed it from the other side: grepping the deployed image found no preset enabling bluetooth.service anywhere. tests/test_package_resolution.py grows that invariant as a guard, since it is the kind of thing the next unit added will get wrong. Three units predating this test are enabled with no preset entry -- brew-setup, flatpak-nuke-fedora, flatpak-preinstall. They are listed as known exceptions rather than asserted to be correct: they come from ublue packages that may ship their own vendor presets, and confirming that needs the built image, not the source tree. A second test fails if a name in that list stops being enabled, so the exception cannot go stale and quietly widen the hole. Verified by mutation rather than assumed. Dropping the [hardware] line from the resolver, removing bluetooth from the preset, dropping *hardware from the verifier's expected set, and deleting linux-firmware or avahi from the manifest each fail the suite and name what broke. Writing the guard also caught a bug in its own helper: an anchored regex missed the indented sshd enablement inside its conditional, which made the exception list look stale when it was not. Not fixed here: #101 timer The bootc-fetch-apply-updates mask loses to a cross-vendor /etc merge, which carries Bluefin's timers.target.wants symlink across. An image-time mask structurally cannot win against a switcher's /etc; that needs a design decision. #102 BIOS Documentation, and the base ships no BIOS boot payload. #103 getty Needs a bootc karg change as well as a mask, which is a riskier edit than the rest of this and deserves its own review. 33 tests pass, 57 subtests pass. The contract check reports 58 Bluefin, 11 GNOME, 5 firmware and 6 service packages. Co-Authored-By: Claude Opus 5 Claude-Session: https://claude.ai/code/session_01W6BYLUD6KBzF8rBFXKjP7M --- packages/utah.toml | 60 +++++++ scripts/configure-services.sh | 21 +++ scripts/install-packages.py | 3 + scripts/verify-rpm-contract.py | 9 +- .../system-preset/85-utah-desktop.preset | 9 ++ tests/test_package_resolution.py | 149 +++++++++++++++++- 6 files changed, 248 insertions(+), 3 deletions(-) diff --git a/packages/utah.toml b/packages/utah.toml index ac9b02fde..3138b87ab 100644 --- a/packages/utah.toml +++ b/packages/utah.toml @@ -7,6 +7,7 @@ # # Sections: # [gnome] GNOME 51 desktop contract Hummingbird does not ship +# [hardware] firmware and microcode the bootable base leaves out # [build] toolchain needed to build the pinned GNOME extensions # [unavailable] Bluefin contract packages none of Utah's repositories provide # @@ -44,10 +45,61 @@ packages = [ "gtk4", "libadwaita", "mutter", + # nautilus is already in the pinned factory repository at + # 51~beta-1.hum1.bfin; it was simply never asked for, which is why #100 + # found nautilus-extensions, nautilus-python and nautilus-gsconnect + # installed with no /usr/bin/nautilus for them to extend. Bluefin does + # not list it either, because Silverblue carries it in the base. + "nautilus", "xdg-desktop-portal", "xdg-desktop-portal-gnome", ] +[hardware] +# Firmware for real machines. Neither of these is in the Bluefin contract, +# for the same reason the [services] entries below are not: Fedora's base +# image carries them, so Bluefin never has to ask, while Hummingbird's +# bootable base ships effectively no device firmware at all. +# +# On hardware that needs a blob this is not cosmetic. Reported on a ThinkPad +# X230 (#97) as a dead Centrino Advanced-N 6205: +# +# iwlwifi 0000:03:00.0: Direct firmware load for iwlwifi-6000g2a-6.ucode +# failed with error -2 +# iwlwifi 0000:03:00.0: no suitable firmware found +# +# with no iwlwifi blob anywhere under /usr/lib/firmware on the deployed image +# and no linux-firmware in its rpmdb, against ~192 firmware files on Bluefin +# on the same machine. A laptop with no wireless has no way to fetch the fix, +# which is what makes this the sharpest of the parity gaps. +# +# linux-firmware alone is NOT enough, which is the trap in this issue. It is +# a split package here: it Requires only linux-firmware-whence, and it +# Recommends thirteen device packages -- amd-gpu, atheros, brcmfmac, cirrus, +# intel-gpu, mt7xxx, nvidia-gpu, nxpwireless, qcom-wwan, realtek, tiwilink and +# friends -- which weak dependencies bring in for free. Intel wireless is in +# neither list. Measured against the pinned repository's filelists, +# iwlwifi-6000g2a-6.ucode -- the exact blob #97 names -- ships in +# iwlwifi-dvm-firmware, so installing linux-firmware and stopping there would +# have closed the issue while leaving the X230 just as dead. +# +# So the three iwlwifi generations and iwlegacy are named explicitly, which +# covers Intel wireless from the 3945 era through current parts. Everything +# else arrives through the recommends above. +# +# microcode_ctl is deliberately absent: it exists in neither enabled +# repository -- not in the pinned factory image, not in +# public-hummingbird-x86_64-rpms -- so naming it here would fail the install +# transaction outright. CPU errata updates therefore remain a real gap on +# this base, and closing it needs a factory recipe first. +packages = [ + "linux-firmware", + "iwlwifi-dvm-firmware", + "iwlwifi-mvm-firmware", + "iwlwifi-mld-firmware", + "iwlegacy-firmware", +] + [build] # Build dependencies for scripts/build-gnome-extensions.sh. # @@ -86,7 +138,15 @@ packages = [ # -- and an installed system whose disk is encrypted could not manage its own # keys. The initramfs already carries the crypt and systemd-cryptsetup dracut # modules, so unlocking was never the gap; creating the volume was. +# +# avahi is the same story once more: Hummingbird ships avahi-libs and +# avahi-glib but not the daemon, so anything that speaks to it over D-Bus +# finds nobody home. Reported as geoclue failing on every start (#104) -- +# geoclue[2327]: Failed to connect to avahi service: Daemon not running +# -- which also takes mDNS name resolution and WiFi-based location with it. +# The client libraries being present is what makes this one easy to miss. packages = [ + "avahi", "cryptsetup", "flatpak", "fwupd", diff --git a/scripts/configure-services.sh b/scripts/configure-services.sh index 74d02c5b1..c67e5b923 100644 --- a/scripts/configure-services.sh +++ b/scripts/configure-services.sh @@ -70,6 +70,27 @@ enable_unit ublue-system-setup.service enable_unit systemd-resolved.service enable_unit bootc-unified-storage.service +# Hummingbird's server preset enables none of these, and Bluefin gets them +# from its own 90-default.preset, so on Utah the packages were installed and +# the units never started. Each was reported from the same ThinkPad X230: +# +# bluetooth.service bluez 5.87 and bluetoothd present, yet bluetoothctl +# shows no adapter at all (#98). The BCM20702 in that +# machine runs firmware-less, so enablement is the +# whole fix. +# input-remapper the GUI cannot reach a root daemon and falls back to +# prompting for a password (#99). +# avahi-daemon geoclue retries "Failed to connect to avahi service: +# Daemon not running" forever, taking mDNS resolution +# and WiFi location with it (#104). The daemon itself +# arrives with the avahi entry added to utah.toml. +# +# enable_unit is a no-op when the unit is absent, so a flavor that does not +# carry one of these is unaffected. +enable_unit bluetooth.service +enable_unit input-remapper.service +enable_unit avahi-daemon.service + # Bluefin's Brewfile and Bazaar preinstall hook need the Flathub remote before # first boot. Keep this as a .flatpakrepo descriptor so the remote is available # to both flatpak-preinstall and brew-setup without baking mutable /var state. diff --git a/scripts/install-packages.py b/scripts/install-packages.py index 09d715bab..8e9407277 100755 --- a/scripts/install-packages.py +++ b/scripts/install-packages.py @@ -54,6 +54,9 @@ def contract(base: Path, overlay: Path, major: str | None) -> list[str]: if major: packages += section(base, f"fedora_v{major}") packages += section(overlay, "gnome") + # Firmware is part of the contract too: the bootable base ships almost + # none, so a driver that needs a blob finds nothing to load. + packages += section(overlay, "hardware") # Service packages are part of the desktop contract as well: 40-services.sh # cannot enable what the server base never installed. packages += section(overlay, "services") diff --git a/scripts/verify-rpm-contract.py b/scripts/verify-rpm-contract.py index 08fb22ff3..45156e535 100644 --- a/scripts/verify-rpm-contract.py +++ b/scripts/verify-rpm-contract.py @@ -61,19 +61,24 @@ def main() -> int: if resolved.exists(): contract = [line for line in resolved.read_text().split() if line] gnome_names = set(section(overlay, "gnome")) + hardware_names = set(section(overlay, "hardware")) service_names = set(section(overlay, "services")) - bluefin = [p for p in contract if p not in gnome_names and p not in service_names] + overlay_names = gnome_names | hardware_names | service_names + bluefin = [p for p in contract if p not in overlay_names] gnome = [p for p in contract if p in gnome_names] + hardware = [p for p in contract if p in hardware_names] services = [p for p in contract if p in service_names] else: bluefin = [p for p in section(args.manifest, "fedora") if p not in unavailable] gnome = section(overlay, "gnome") + hardware = section(overlay, "hardware") services = section(overlay, "services") nvidia = list(NVIDIA_PACKAGES) if "nvidia" in flavor else [] - expected = [*bluefin, *gnome, *services, *nvidia] + expected = [*bluefin, *gnome, *hardware, *services, *nvidia] print( f"Verifying {len(bluefin)} Bluefin packages, {len(gnome)} GNOME desktop packages," + f" {len(hardware)} firmware packages," f" {len(services)} desktop service packages, and {len(nvidia)} NVIDIA packages", flush=True, ) diff --git a/system_files/shared/usr/lib/systemd/system-preset/85-utah-desktop.preset b/system_files/shared/usr/lib/systemd/system-preset/85-utah-desktop.preset index ad4db0804..5808177ca 100644 --- a/system_files/shared/usr/lib/systemd/system-preset/85-utah-desktop.preset +++ b/system_files/shared/usr/lib/systemd/system-preset/85-utah-desktop.preset @@ -10,5 +10,14 @@ enable ublue-system-setup.service enable systemd-resolved.service enable bootc-unified-storage.service enable dconf-update.service +# Desktop hardware units. These must be listed here and not only enabled in +# configure-services.sh: bootc applies presets on first boot, so a build-time +# enablement with no preset line behind it gets undone. That is the trap the +# sshd branch of configure-services.sh already documents, and #98 confirmed +# the symptom from the other side -- grepping the deployed image found no +# preset enabling bluetooth.service anywhere. +enable bluetooth.service +enable input-remapper.service +enable avahi-daemon.service disable sshd.service disable rpm-ostree.service diff --git a/tests/test_package_resolution.py b/tests/test_package_resolution.py index 14350d01f..e6ccdab23 100644 --- a/tests/test_package_resolution.py +++ b/tests/test_package_resolution.py @@ -77,6 +77,7 @@ def resolve(self, output, code=1): base.write_text('[fedora]\npackages=["base", "unavailable"]\n' '[fedora_v44]\npackages=["release-specific"]\n') overlay.write_text('[gnome]\npackages=["shell"]\n' + '[hardware]\npackages=["firmware"]\n' '[services]\npackages=["resolver"]\n' '[build]\npackages=["compiler"]\n' '[unavailable]\npackages=["unavailable"]\n') @@ -92,7 +93,8 @@ def test_valid_declined_transaction_includes_every_install_section(self): rc, command = self.resolve("Transaction Summary:\nInstall 12 Packages\nOperation aborted.\n") self.assertEqual(rc, 0) self.assertEqual(command[command.index("install") + 1:], - ["base", "release-specific", "shell", "resolver", "compiler"]) + ["base", "release-specific", "shell", "firmware", "resolver", + "compiler"]) self.assertIn("--assumeno", command) self.assertIn("--disablerepo=*", command) for repo in installer.REPOS: @@ -125,5 +127,150 @@ def test_pins_come_from_containerfile(self): checker.pinned_inputs(path) +class DesktopUnitEnablementTests(unittest.TestCase): + """A build-time enablement with no preset line behind it does not survive. + + bootc applies systemd presets on first boot, so `systemctl enable` at + build time is undone unless 85-utah-desktop.preset agrees. That trap is + documented in configure-services.sh's sshd branch and was confirmed from + the other side by projectbluefin/utah#98, where the deployed image had no + preset entry enabling bluetooth.service. Hold the two files in agreement + so the next unit added to one cannot silently omit the other. + """ + + PRESET = ROOT / ("system_files/shared/usr/lib/systemd/system-preset/" + "85-utah-desktop.preset") + SERVICES = ROOT / "scripts/configure-services.sh" + + def preset_directives(self, verb): + return { + line.split()[1] + for line in self.PRESET.read_text().splitlines() + if line.startswith(f"{verb} ") + } + + def script_units(self, function): + # Leading whitespace matters: sshd is enabled inside a conditional + # block, so an anchored pattern misses it and the allowlist below + # would look stale when it is not. + return set( + re.findall(rf"^[ \t]*{function} (\S+)$", self.SERVICES.read_text(), re.M) + ) + + # sshd is deliberately excluded: configure-services.sh rewrites the preset + # in place for the opt-in debug build rather than shipping it enabled, + # which is the one case where the two files may legitimately disagree. + # + # The other three predate this test and are NOT asserted to be correct. + # They come from ublue packages that may ship their own vendor presets, in + # which case Utah's preset has nothing to add -- but that was not verified + # here, because it needs the built image rather than the source tree. They + # are listed so the guard below can be exact about what it does not yet + # cover, instead of being weakened into passing for everything. If one of + # them turns out to have no preset behind it either, it is the same bug as + # #98 and belongs in the preset. + WITHOUT_PRESET = { + "sshd.service", + "brew-setup.service", + "flatpak-nuke-fedora.service", + "flatpak-preinstall.service", + } + + def test_no_new_unit_is_enabled_without_a_preset_entry(self): + enabled = self.script_units("enable_unit") - self.WITHOUT_PRESET + missing = sorted(enabled - self.preset_directives("enable")) + self.assertEqual( + missing, [], + "enabled at build time with no preset entry, so bootc's first-boot " + f"preset application will undo it: {missing}", + ) + + def test_the_exception_list_does_not_cover_absent_units(self): + # A stale allowlist silently widens the hole above. Every name in it + # must still be a unit configure-services.sh actually enables. + enabled = self.script_units("enable_unit") + stale = sorted(self.WITHOUT_PRESET - enabled) + self.assertEqual(stale, [], f"allowlisted but no longer enabled: {stale}") + + def test_the_reported_desktop_units_are_enabled(self): + # projectbluefin/utah#98, #99 and #104: each package was installed and + # its unit never started. Name them so a refactor cannot drop one. + for unit in ("bluetooth.service", "input-remapper.service", "avahi-daemon.service"): + with self.subTest(unit=unit): + self.assertIn(unit, self.script_units("enable_unit")) + self.assertIn(unit, self.preset_directives("enable")) + + +class HardwareContractTests(unittest.TestCase): + """Firmware is in the install set and in what the image verifies.""" + + OVERLAY = ROOT / "packages/utah.toml" + verifier = None + + @classmethod + def setUpClass(cls): + cls.verifier = load("verify-rpm-contract") + + def test_firmware_packages_are_declared(self): + # projectbluefin/utah#97: no linux-firmware in the deployed rpmdb, and + # iwlwifi found no blob to load. + declared = installer.section(self.OVERLAY, "hardware") + self.assertIn("linux-firmware", declared) + + def test_intel_wireless_firmware_is_named_explicitly(self): + # The heart of #97. linux-firmware is a split package that requires + # only linux-firmware-whence and recommends thirteen device packages, + # none of them Intel wireless -- while iwlwifi-6000g2a-6.ucode, the + # blob the report names, ships in iwlwifi-dvm-firmware. Depending on + # linux-firmware alone closes the issue without fixing the radio, so + # pin the explicit set. + declared = set(installer.section(self.OVERLAY, "hardware")) + for pkg in ("iwlwifi-dvm-firmware", "iwlwifi-mvm-firmware", + "iwlwifi-mld-firmware", "iwlegacy-firmware"): + with self.subTest(package=pkg): + self.assertIn(pkg, declared) + + def test_no_package_is_requested_from_outside_the_enabled_repositories(self): + # microcode_ctl was in the first draft of [hardware] and is in neither + # enabled repository, which would have failed the install transaction. + # Nothing here can assert repository contents offline, so hold the one + # name that was measured and found missing. + declared = set(installer.section(self.OVERLAY, "hardware")) + self.assertNotIn( + "microcode_ctl", declared, + "microcode_ctl is in neither the pinned factory image nor " + "public-hummingbird-x86_64-rpms; it needs a factory recipe first", + ) + + def test_nautilus_is_in_the_install_set(self): + # projectbluefin/utah#100: the extensions shipped without the app. + # nautilus is in the pinned factory repository at 51~beta-1.hum1.bfin, + # so this needs no new factory work. + contract = installer.contract(ROOT / "packages/bluefin.toml", self.OVERLAY, "44") + self.assertIn("nautilus", contract) + + def test_hardware_section_reaches_the_install_set(self): + contract = installer.contract(ROOT / "packages/bluefin.toml", self.OVERLAY, "44") + for pkg in installer.section(self.OVERLAY, "hardware"): + with self.subTest(package=pkg): + self.assertIn(pkg, contract) + + def test_avahi_daemon_is_in_the_install_set(self): + # projectbluefin/utah#104: avahi-libs and avahi-glib shipped, the + # daemon did not, so geoclue had nothing to talk to. + contract = installer.contract(ROOT / "packages/bluefin.toml", self.OVERLAY, "44") + self.assertIn("avahi", contract) + + def test_verifier_asserts_the_hardware_section(self): + # The off-image --check path builds `expected` from the manifest. If + # [hardware] were missing from it, firmware could go absent from a + # built image without the contract check noticing. + section = self.verifier.section(self.OVERLAY, "hardware") + self.assertTrue(section) + source = (ROOT / "scripts/verify-rpm-contract.py").read_text() + self.assertIn('hardware = section(overlay, "hardware")', source) + self.assertIn("*hardware,", source) + + if __name__ == "__main__": unittest.main() From 31d765bddb78eac0b1cb1e2771e436d05c7da05f Mon Sep 17 00:00:00 2001 From: Claude Date: Thu, 17 Sep 2026 16:55:25 +0000 Subject: [PATCH 2/8] fix: restore the bluefin.toml parity contract packages/bluefin.toml is meant to be a byte-for-byte copy of upstream Bluefin's build_files/packages/base.toml. utah.toml says so in as many words -- "it is the parity contract, and any drift there is a parity bug" -- and just check-parity exists to make that loud. It has drifted, so that check is red on main today and every open pull request inherits the failure, #94 and #105 among them. The drift is four hunks: firefox moved into [fedora], and libgda, libgda-sqlite and ppp missing. Syncing the file alone would have been worse than the failure it fixes. All three missing names were measured against both repositories the image actually enables -- the digest-pinned factory metadata layer from Containerfile, and public-hummingbird-x86_64-rpms: libgda absent from both libgda-sqlite absent from both ppp absent from both, and nothing Provides it So a bare sync hands three unresolvable names to the install transaction and turns a failing lint into a failing image build. The sync and the [unavailable] entries have to land together, which is why they are one commit. Each entry carries a tracking issue, as that list requires: libgda, libgda-sqlite (#106) Upstream added both for the maintained Copyous extension (bluefin#1210). Hummingbird carries no package whose name contains gda at all. One factory recipe covers both, because libgda-sqlite is a subpackage of the libgda SRPM rather than its own source -- Fedora's libgda.spec declares %package sqlite alongside ui, tools, sqlcipher, mysql and postgres. ppp (#107) Worth more than dial-up trivia. Hummingbird ships NetworkManager-ppp, both 1.58.0-1.hum1 and 1.58.1-1.hum1, and its metadata carries Requires: ppp. With ppp nowhere in the enabled set that plugin is uninstallable from Hummingbird's own repository, so NetworkManager's PPPoE and PPTP paths cannot be installed at all. It is the same shape as avahi (#104) and nautilus (#100): a plugin shipped without the daemon behind it. firefox needed no handling. It is already in [unavailable], and contract() filters that set out of every section, so moving it between sections cannot put it back in the image -- which matters, because Utah ships org.mozilla.firefox as a Flatpak and re-adding the RPM would regress that decision. This change installs nothing new and removes nothing. Proven rather than argued: contract() over main's two files and over this branch's both return the same 74 packages, with no additions and no removals, and libgda, libgda-sqlite, ppp and firefox are each absent from the resulting set. 23 tests pass, 45 subtests pass. The parity diff against upstream is now empty, the RPM contract check reports 58 Bluefin, 10 GNOME and 5 service packages, and install-packages.py's own guard confirms no [unavailable] package appears in the install set. Noted but deliberately not changed: evolution-ews-core, fish and zsh predate this and carry no issue reference, so the "every entry MUST carry a tracking issue" rule is not currently enforced by anything. Worth a guard, but not in a commit whose job is restoring the contract. Co-Authored-By: Claude Opus 5 Claude-Session: https://claude.ai/code/session_01W6BYLUD6KBzF8rBFXKjP7M --- packages/bluefin.toml | 6 +++++- packages/utah.toml | 20 ++++++++++++++++++++ 2 files changed, 25 insertions(+), 1 deletion(-) diff --git a/packages/bluefin.toml b/packages/bluefin.toml index 8d5ef38f5..4003ddd1b 100644 --- a/packages/bluefin.toml +++ b/packages/bluefin.toml @@ -46,7 +46,6 @@ packages = [ "firewall-config", "fish", "flatpak-spawn", - "firefox", "fzf", "gcc", "gcc-c++", @@ -71,6 +70,9 @@ packages = [ "libblockdev-mpath", "libcamera-gstreamer", "libcamera-tools", + # Runtime dependencies for the maintained Copyous extension (#1210). + "libgda", + "libgda-sqlite", "libratbag-ratbagd", "libva-utils", "libxcrypt-compat", @@ -82,6 +84,7 @@ packages = [ "openrgb-udev-rules", "openssh-askpass", "pipewire-libs-extra", + "ppp", "python3-gnome-ponytail-daemon", "slitherer", "squashfs-tools", @@ -120,6 +123,7 @@ packages = [ "fedora-chromium-config", "fedora-chromium-config-gnome", "fedora-third-party", + "firefox", "firefox-langpacks", "gnome-extensions-app", "gnome-shell-extension-background-logo", diff --git a/packages/utah.toml b/packages/utah.toml index ac9b02fde..75c312533 100644 --- a/packages/utah.toml +++ b/packages/utah.toml @@ -133,6 +133,23 @@ packages = [ # its Qt WebEngine dependency closure. Match the factory's installer policy # until that path is supported; tracked by #83. # +# libgda, libgda-sqlite +# Upstream added both as runtime dependencies for the maintained Copyous +# extension (bluefin#1210). Neither is in the pinned factory image or in +# public-hummingbird-x86_64-rpms -- Hummingbird carries no package whose +# name contains gda at all. One factory recipe would cover both, because +# libgda-sqlite is a subpackage of the libgda SRPM rather than its own +# source. Tracked by #106. +# +# ppp +# In neither repository, and nothing in either Provides it. This one costs +# more than dial-up: Hummingbird ships NetworkManager-ppp (1.58.0-1.hum1 +# and 1.58.1-1.hum1) whose metadata carries Requires: ppp, so that plugin +# is uninstallable from their own repository and NetworkManager's PPPoE and +# PPTP paths cannot be installed at all. The same shape as avahi (#104) and +# nautilus (#100): a plugin shipped without the daemon behind it. +# Tracked by #107. +# # grub2-efi-x64-cdboot and pipewire-libs-extra are supplied by the pinned # September 10 factory candidate and return to the install contract (#35). packages = [ @@ -140,6 +157,9 @@ packages = [ "evolution-ews-core", "firefox", "fish", + "libgda", + "libgda-sqlite", + "ppp", "slitherer", "zsh", ] From 3ee771dd2ae00f70e956055c25c15af89b11c00f Mon Sep 17 00:00:00 2001 From: Claude Date: Thu, 17 Sep 2026 17:02:06 +0000 Subject: [PATCH 3/8] fix: drop nautilus and avahi, whose closures do not resolve CI resolved the real install transaction and found that two of the four package additions cannot be installed. Name availability is not installability, and checking the former is the mistake this corrects: I read both names out of repodata, confirmed they exist, and stopped there. Problem 1: conflicting requests - nothing provides libgexiv2-0.16.so.4()(64bit) needed by nautilus-51~beta-1.hum1.bfin.x86_64 from utah-packages - nothing provides libportal-gtk4.so.1()(64bit) needed by nautilus... - nothing provides libportal.so.1()(64bit) needed by nautilus... Problem 2: cannot install the best candidate for the job - nothing provides libdaemon.so.0()(64bit) needed by avahi-0.9~rc4-0.1.1.hum1.x86_64 from public-hummingbird-x86_64-rpms - nothing provides libdaemon >= 0.11 needed by avahi... gexiv2, libportal, exiv2 and libdaemon are in no repository this image enables, and in no factory recipe either -- not in the pinned factory image, not in public-hummingbird-x86_64-rpms, and not among utah-packages#70's 340 recipes. So neither package becomes installable when #70 lands. nautilus needs three new recipes (libportal, gexiv2, and exiv2 beneath it); avahi needs one (libdaemon). Both move to [unavailable] with that recorded, tracked by the issues that already exist for them: #100 for nautilus, #104 for avahi. The parity debt list is the honest place for a package whose blocker is named and unfixed here. This also corrects what I put on #100. I told the reporter nautilus needed no factory work because the recipe was already in the pinned repository. The recipe is there; the closure is not, and that is the same wall from one step further in. The avahi-daemon.service enablement goes with it, from both configure-services.sh and 85-utah-desktop.preset. A preset line for a unit whose package cannot be installed is dead configuration that reads like a fix. tests/test_package_resolution.py grows test_no_unit_is_enabled_for_an_uninstallable_package so the next attempt cannot reintroduce that shape, and test_packages_whose_closure_does_not_resolve_stay_out pins both packages with the failing sonames in its docstring. What survives is the part that actually resolved, and it is the part that mattered most: [hardware] with linux-firmware and the four Intel wireless firmware packages. None of those appeared in the transaction problems. #97 -- a ThinkPad X230 with no working WiFi -- is still fixed by this PR, as are #98 and #99, whose packages were installed all along and only needed enabling. 79 packages in the install set, no [unavailable] overlap, parity clean against upstream, 33 tests and 58 subtests pass, contract check reports 58 Bluefin, 10 GNOME, 5 firmware and 5 service packages. Co-Authored-By: Claude Opus 5 Claude-Session: https://claude.ai/code/session_01W6BYLUD6KBzF8rBFXKjP7M --- packages/utah.toml | 35 +++++++++------ scripts/configure-services.sh | 10 ++--- .../system-preset/85-utah-desktop.preset | 1 - tests/test_package_resolution.py | 44 +++++++++++++------ 4 files changed, 56 insertions(+), 34 deletions(-) diff --git a/packages/utah.toml b/packages/utah.toml index 93b1c307d..ecd3f5714 100644 --- a/packages/utah.toml +++ b/packages/utah.toml @@ -45,12 +45,6 @@ packages = [ "gtk4", "libadwaita", "mutter", - # nautilus is already in the pinned factory repository at - # 51~beta-1.hum1.bfin; it was simply never asked for, which is why #100 - # found nautilus-extensions, nautilus-python and nautilus-gsconnect - # installed with no /usr/bin/nautilus for them to extend. Bluefin does - # not list it either, because Silverblue carries it in the base. - "nautilus", "xdg-desktop-portal", "xdg-desktop-portal-gnome", ] @@ -138,15 +132,7 @@ packages = [ # -- and an installed system whose disk is encrypted could not manage its own # keys. The initramfs already carries the crypt and systemd-cryptsetup dracut # modules, so unlocking was never the gap; creating the volume was. -# -# avahi is the same story once more: Hummingbird ships avahi-libs and -# avahi-glib but not the daemon, so anything that speaks to it over D-Bus -# finds nobody home. Reported as geoclue failing on every start (#104) -- -# geoclue[2327]: Failed to connect to avahi service: Daemon not running -# -- which also takes mDNS name resolution and WiFi-based location with it. -# The client libraries being present is what makes this one easy to miss. packages = [ - "avahi", "cryptsetup", "flatpak", "fwupd", @@ -193,6 +179,25 @@ packages = [ # its Qt WebEngine dependency closure. Match the factory's installer policy # until that path is supported; tracked by #83. # +# avahi +# Hummingbird ships avahi-libs and avahi-glib but not the daemon, so +# geoclue retries forever against nobody (#104). The daemon package does +# exist in their repository, but it is not installable here: avahi +# 0.9~rc4-0.1.1.hum1 needs libdaemon.so.0 and libdaemon >= 0.11, and +# libdaemon is in no repository this image enables -- not Hummingbird, not +# the pinned factory image, and not the factory inventory either. A +# libdaemon recipe has to come first. Tracked by #104. +# +# nautilus +# The accessories ship orphaned -- nautilus-extensions, nautilus-python and +# nautilus-gsconnect installed with no /usr/bin/nautilus (#100). The recipe +# is not the gap: nautilus 51~beta-1.hum1.bfin is in the pinned factory +# repository already. Its dependency closure is. It needs +# libgexiv2-0.16.so.4, libportal.so.1 and libportal-gtk4.so.1, and gexiv2 +# and libportal are in no repository this image enables and in no factory +# recipe -- including utah-packages#70's 340. gexiv2 additionally needs +# exiv2, which is equally absent. Three recipes, not one. Tracked by #100. +# # libgda, libgda-sqlite # Upstream added both as runtime dependencies for the maintained Copyous # extension (bluefin#1210). Neither is in the pinned factory image or in @@ -214,11 +219,13 @@ packages = [ # September 10 factory candidate and return to the install contract (#35). packages = [ "anaconda-live", + "avahi", "evolution-ews-core", "firefox", "fish", "libgda", "libgda-sqlite", + "nautilus", "ppp", "slitherer", "zsh", diff --git a/scripts/configure-services.sh b/scripts/configure-services.sh index c67e5b923..d8df7f08e 100644 --- a/scripts/configure-services.sh +++ b/scripts/configure-services.sh @@ -80,16 +80,16 @@ enable_unit bootc-unified-storage.service # whole fix. # input-remapper the GUI cannot reach a root daemon and falls back to # prompting for a password (#99). -# avahi-daemon geoclue retries "Failed to connect to avahi service: -# Daemon not running" forever, taking mDNS resolution -# and WiFi location with it (#104). The daemon itself -# arrives with the avahi entry added to utah.toml. +# +# avahi-daemon is deliberately NOT here. #104 is real, but avahi cannot be +# installed on this base at all -- it needs libdaemon, which no enabled +# repository carries -- so enabling its unit would be dead configuration +# claiming a fix that is not there. It goes back in with the package. # # enable_unit is a no-op when the unit is absent, so a flavor that does not # carry one of these is unaffected. enable_unit bluetooth.service enable_unit input-remapper.service -enable_unit avahi-daemon.service # Bluefin's Brewfile and Bazaar preinstall hook need the Flathub remote before # first boot. Keep this as a .flatpakrepo descriptor so the remote is available diff --git a/system_files/shared/usr/lib/systemd/system-preset/85-utah-desktop.preset b/system_files/shared/usr/lib/systemd/system-preset/85-utah-desktop.preset index 5808177ca..9a7fe9f56 100644 --- a/system_files/shared/usr/lib/systemd/system-preset/85-utah-desktop.preset +++ b/system_files/shared/usr/lib/systemd/system-preset/85-utah-desktop.preset @@ -18,6 +18,5 @@ enable dconf-update.service # preset enabling bluetooth.service anywhere. enable bluetooth.service enable input-remapper.service -enable avahi-daemon.service disable sshd.service disable rpm-ostree.service diff --git a/tests/test_package_resolution.py b/tests/test_package_resolution.py index e6ccdab23..37b7bc215 100644 --- a/tests/test_package_resolution.py +++ b/tests/test_package_resolution.py @@ -193,13 +193,22 @@ def test_the_exception_list_does_not_cover_absent_units(self): self.assertEqual(stale, [], f"allowlisted but no longer enabled: {stale}") def test_the_reported_desktop_units_are_enabled(self): - # projectbluefin/utah#98, #99 and #104: each package was installed and - # its unit never started. Name them so a refactor cannot drop one. - for unit in ("bluetooth.service", "input-remapper.service", "avahi-daemon.service"): + # projectbluefin/utah#98 and #99: each package was installed and its + # unit never started. Name them so a refactor cannot drop one. + for unit in ("bluetooth.service", "input-remapper.service"): with self.subTest(unit=unit): self.assertIn(unit, self.script_units("enable_unit")) self.assertIn(unit, self.preset_directives("enable")) + def test_no_unit_is_enabled_for_an_uninstallable_package(self): + # avahi-daemon was enabled here until CI resolved the real + # transaction and showed avahi cannot be installed at all (it needs + # libdaemon, which no enabled repository carries). Enabling a unit + # whose package is in [unavailable] is dead configuration that reads + # like a fix. + enabled = self.script_units("enable_unit") | self.preset_directives("enable") + self.assertNotIn("avahi-daemon.service", enabled) + class HardwareContractTests(unittest.TestCase): """Firmware is in the install set and in what the image verifies.""" @@ -242,24 +251,31 @@ def test_no_package_is_requested_from_outside_the_enabled_repositories(self): "public-hummingbird-x86_64-rpms; it needs a factory recipe first", ) - def test_nautilus_is_in_the_install_set(self): - # projectbluefin/utah#100: the extensions shipped without the app. - # nautilus is in the pinned factory repository at 51~beta-1.hum1.bfin, - # so this needs no new factory work. - contract = installer.contract(ROOT / "packages/bluefin.toml", self.OVERLAY, "44") - self.assertIn("nautilus", contract) - def test_hardware_section_reaches_the_install_set(self): contract = installer.contract(ROOT / "packages/bluefin.toml", self.OVERLAY, "44") for pkg in installer.section(self.OVERLAY, "hardware"): with self.subTest(package=pkg): self.assertIn(pkg, contract) - def test_avahi_daemon_is_in_the_install_set(self): - # projectbluefin/utah#104: avahi-libs and avahi-glib shipped, the - # daemon did not, so geoclue had nothing to talk to. + def test_packages_whose_closure_does_not_resolve_stay_out(self): + """Name availability is not installability, which CI proved the hard way. + + Both of these exist by name in a repository the image enables, so a + name lookup says yes. Resolving the real transaction says no: + + nothing provides libgexiv2-0.16.so.4 needed by nautilus... + nothing provides libportal.so.1 needed by nautilus... + nothing provides libdaemon.so.0 needed by avahi... + + gexiv2, libportal, exiv2 and libdaemon are in no enabled repository + and in no factory recipe, so neither package can be installed until + those exist. #100 and #104 track them. + """ contract = installer.contract(ROOT / "packages/bluefin.toml", self.OVERLAY, "44") - self.assertIn("avahi", contract) + for pkg in ("nautilus", "avahi"): + with self.subTest(package=pkg): + self.assertNotIn(pkg, contract) + self.assertIn(pkg, installer.section(self.OVERLAY, "unavailable")) def test_verifier_asserts_the_hardware_section(self): # The off-image --check path builds `expected` from the manifest. If From 3446d62974b233df2245353d2d6bcf6f800e2052 Mon Sep 17 00:00:00 2001 From: Claude Date: Fri, 18 Sep 2026 15:40:19 +0000 Subject: [PATCH 4/8] Install what Bluefin inherits from Fedora's base and Hummingbird already has The image-level parity report (#117) found 277 packages in Bluefin's image that Utah's lacks and that public-hummingbird-x86_64-rpms carries today: Bluefin never names them because Silverblue's base supplies them, and Hummingbird's server-shaped bootable base leaves them out. A new [parity] section in packages/utah.toml adds the desktop-relevant ones -- the NetworkManager Bluetooth and WWAN plugins and connectivity check, ALSA use-case profiles, the English language pack and core font set, and the shell tooling Bluefin users reach for -- 30 names, each with its Requires resolved against Hummingbird plus the pinned factory repository before it went in. Dakota ships the same networking, audio and font pieces. What did not resolve is recorded beside the list rather than dropped silently: cups needs cups-filters, vim-enhanced needs gpm, the emoji and math default fonts need google-noto-emoji-color-fonts and stix-fonts, and fprintd and alsa-utils are in no repository at all. NetworkManager-wifi waits for the factory image that carries wpa_supplicant and wireless-regdb (utah-packages#133). install-packages.py adds the section to the install set, verify-rpm-contract asserts it on the built image like [hardware], and the tests cover both paths plus a check that [parity] never duplicates a name another section already claims (unzip was one, via [build]). Co-Authored-By: Claude Fable 5.1 Claude-Session: https://claude.ai/code/session_01W6BYLUD6KBzF8rBFXKjP7M --- docs/skills/package-contract.md | 3 ++ packages/utah.toml | 65 ++++++++++++++++++++++++++++++++ scripts/install-packages.py | 3 ++ scripts/verify-rpm-contract.py | 9 +++-- tests/test_package_resolution.py | 28 +++++++++++++- 5 files changed, 103 insertions(+), 5 deletions(-) diff --git a/docs/skills/package-contract.md b/docs/skills/package-contract.md index f9aefcdae..155d14a79 100644 --- a/docs/skills/package-contract.md +++ b/docs/skills/package-contract.md @@ -40,6 +40,9 @@ policy for changing them. - `[gnome]` — GNOME 51 desktop contract Hummingbird does not ship. - `[build]` — toolchain needed to build the pinned GNOME extensions (`scripts/build-gnome-extensions.sh`). + - `[parity]` — what Bluefin inherits from Fedora's base image and Hummingbird + has in its repository but not in its bootable base; each entry's Requires + was resolved against Hummingbird plus the pinned factory before adding it. - `[services]` — desktop services Bluefin adds on top of the server base. - `[unavailable]` — Bluefin contract packages none of Utah's repositories provide. diff --git a/packages/utah.toml b/packages/utah.toml index ecd3f5714..b860ff335 100644 --- a/packages/utah.toml +++ b/packages/utah.toml @@ -8,6 +8,7 @@ # Sections: # [gnome] GNOME 51 desktop contract Hummingbird does not ship # [hardware] firmware and microcode the bootable base leaves out +# [parity] what Bluefin gets from Fedora's base image and Hummingbird has # [build] toolchain needed to build the pinned GNOME extensions # [unavailable] Bluefin contract packages none of Utah's repositories provide # @@ -94,6 +95,70 @@ packages = [ "iwlegacy-firmware", ] +[parity] +# Packages Bluefin's image carries without naming them in bluefin.toml, because +# Fedora's Silverblue base supplies them, and which Hummingbird's repository +# has but its server-shaped bootable base leaves out. Measured, not guessed: +# the image-level parity report (#117, scripts/check-image-parity.py) listed +# 277 such names against public-hummingbird-x86_64-rpms, and each entry here +# had its Requires resolved against Hummingbird plus the pinned factory +# repository before it was added. Aligned with Dakota where it ships the same +# piece: NetworkManager's Bluetooth and WWAN plugins, the connectivity check, +# ALSA use-case profiles, the language pack and the core font set. +# +# Left out, with the dependency Hummingbird does not carry: +# cups, cups-client need cups-filters -- no driverless-only cups +# exists as a package, so printing stays a gap +# vim-enhanced needs libgpm.so.2 (vim-minimal is in the base) +# default-fonts-core-emoji needs google-noto-emoji-color-fonts +# default-fonts-core-math needs stix-fonts +# fprintd, fprintd-pam in no repository; authselect's with-fingerprint +# is enabled but has no daemon behind it yet +# alsa-utils in no repository; Dakota builds it for the +# mixer-state restore (dakota#1027) +# unzip already in [build], which the image keeps +# +# NetworkManager-wifi is deliberately not here yet: its supplicant and the +# regulatory database are being built by utah-packages#133, and it joins this +# list when the factory image that carries them is pinned (#97, +# utah-packages#136). avahi and nautilus are the same shape, see [unavailable]. +packages = [ + # Networking + "NetworkManager-bluetooth", + "NetworkManager-wwan", + "NetworkManager-config-connectivity-fedora", + "bind-utils", + "iptables-nft", + # Audio + "alsa-ucm", + # Locale and fonts + "langpacks-en", + "langpacks-fonts-en", + "abattis-cantarell-fonts", + "default-fonts-core-sans", + "default-fonts-core-serif", + "default-fonts-core-mono", + "google-noto-sans-mono-vf-fonts", + "google-noto-serif-vf-fonts", + "google-noto-sans-symbols-2-fonts", + "google-noto-sans-math-fonts", + # Shell and tooling Bluefin users reach for + "bash-completion", + "man-db", + "less", + "which", + "file", + "hostname", + "xxd", + "zip", + "dbus-tools", + "keyutils", + "crypto-policies-scripts", + "poppler-utils", + "python3-pip", + "systemd-container", +] + [build] # Build dependencies for scripts/build-gnome-extensions.sh. # diff --git a/scripts/install-packages.py b/scripts/install-packages.py index 8e9407277..149bd7d4a 100755 --- a/scripts/install-packages.py +++ b/scripts/install-packages.py @@ -57,6 +57,9 @@ def contract(base: Path, overlay: Path, major: str | None) -> list[str]: # Firmware is part of the contract too: the bootable base ships almost # none, so a driver that needs a blob finds nothing to load. packages += section(overlay, "hardware") + # Parity with what Bluefin inherits from Fedora's base image and Hummingbird + # has in its repository but not in its bootable base. + packages += section(overlay, "parity") # Service packages are part of the desktop contract as well: 40-services.sh # cannot enable what the server base never installed. packages += section(overlay, "services") diff --git a/scripts/verify-rpm-contract.py b/scripts/verify-rpm-contract.py index 45156e535..9789f73c3 100644 --- a/scripts/verify-rpm-contract.py +++ b/scripts/verify-rpm-contract.py @@ -62,23 +62,26 @@ def main() -> int: contract = [line for line in resolved.read_text().split() if line] gnome_names = set(section(overlay, "gnome")) hardware_names = set(section(overlay, "hardware")) + parity_names = set(section(overlay, "parity")) service_names = set(section(overlay, "services")) - overlay_names = gnome_names | hardware_names | service_names + overlay_names = gnome_names | hardware_names | parity_names | service_names bluefin = [p for p in contract if p not in overlay_names] gnome = [p for p in contract if p in gnome_names] hardware = [p for p in contract if p in hardware_names] + parity = [p for p in contract if p in parity_names] services = [p for p in contract if p in service_names] else: bluefin = [p for p in section(args.manifest, "fedora") if p not in unavailable] gnome = section(overlay, "gnome") hardware = section(overlay, "hardware") + parity = section(overlay, "parity") services = section(overlay, "services") nvidia = list(NVIDIA_PACKAGES) if "nvidia" in flavor else [] - expected = [*bluefin, *gnome, *hardware, *services, *nvidia] + expected = [*bluefin, *gnome, *hardware, *parity, *services, *nvidia] print( f"Verifying {len(bluefin)} Bluefin packages, {len(gnome)} GNOME desktop packages," - f" {len(hardware)} firmware packages," + f" {len(hardware)} firmware packages, {len(parity)} parity packages," f" {len(services)} desktop service packages, and {len(nvidia)} NVIDIA packages", flush=True, ) diff --git a/tests/test_package_resolution.py b/tests/test_package_resolution.py index 37b7bc215..9e299911f 100644 --- a/tests/test_package_resolution.py +++ b/tests/test_package_resolution.py @@ -78,6 +78,7 @@ def resolve(self, output, code=1): '[fedora_v44]\npackages=["release-specific"]\n') overlay.write_text('[gnome]\npackages=["shell"]\n' '[hardware]\npackages=["firmware"]\n' + '[parity]\npackages=["manpages"]\n' '[services]\npackages=["resolver"]\n' '[build]\npackages=["compiler"]\n' '[unavailable]\npackages=["unavailable"]\n') @@ -93,8 +94,8 @@ def test_valid_declined_transaction_includes_every_install_section(self): rc, command = self.resolve("Transaction Summary:\nInstall 12 Packages\nOperation aborted.\n") self.assertEqual(rc, 0) self.assertEqual(command[command.index("install") + 1:], - ["base", "release-specific", "shell", "firmware", "resolver", - "compiler"]) + ["base", "release-specific", "shell", "firmware", "manpages", + "resolver", "compiler"]) self.assertIn("--assumeno", command) self.assertIn("--disablerepo=*", command) for repo in installer.REPOS: @@ -277,6 +278,29 @@ def test_packages_whose_closure_does_not_resolve_stay_out(self): self.assertNotIn(pkg, contract) self.assertIn(pkg, installer.section(self.OVERLAY, "unavailable")) + def test_parity_section_reaches_the_install_set(self): + contract = installer.contract(ROOT / "packages/bluefin.toml", self.OVERLAY, "44") + for pkg in installer.section(self.OVERLAY, "parity"): + with self.subTest(package=pkg): + self.assertIn(pkg, contract) + + def test_parity_section_names_nothing_the_contract_already_asks_for(self): + # A name in [parity] that bluefin.toml or another overlay section already + # lists would be a duplicate claim, and the verifier rejects duplicates. + parity = installer.section(self.OVERLAY, "parity") + others = set(installer.section(ROOT / "packages/bluefin.toml", "fedora")) + for name in ("gnome", "hardware", "services", "build", "unavailable"): + others |= set(installer.section(self.OVERLAY, name)) + self.assertEqual(sorted(set(parity) & others), []) + self.assertEqual(len(set(parity)), len(parity)) + + def test_verifier_asserts_the_parity_section(self): + section = self.verifier.section(self.OVERLAY, "parity") + self.assertTrue(section) + source = (ROOT / "scripts/verify-rpm-contract.py").read_text() + self.assertIn('parity = section(overlay, "parity")', source) + self.assertIn("*parity,", source) + def test_verifier_asserts_the_hardware_section(self): # The off-image --check path builds `expected` from the manifest. If # [hardware] were missing from it, firmware could go absent from a From 7cf27018b74fb7c8948cc5814ec68e48d120546b Mon Sep 17 00:00:00 2001 From: Claude Date: Fri, 18 Sep 2026 16:29:51 +0000 Subject: [PATCH 5/8] Revert "Install what Bluefin inherits from Fedora's base and Hummingbird already has" This reverts commit 3446d62974b233df2245353d2d6bcf6f800e2052. --- docs/skills/package-contract.md | 3 -- packages/utah.toml | 65 -------------------------------- scripts/install-packages.py | 3 -- scripts/verify-rpm-contract.py | 9 ++--- tests/test_package_resolution.py | 28 +------------- 5 files changed, 5 insertions(+), 103 deletions(-) diff --git a/docs/skills/package-contract.md b/docs/skills/package-contract.md index 155d14a79..f9aefcdae 100644 --- a/docs/skills/package-contract.md +++ b/docs/skills/package-contract.md @@ -40,9 +40,6 @@ policy for changing them. - `[gnome]` — GNOME 51 desktop contract Hummingbird does not ship. - `[build]` — toolchain needed to build the pinned GNOME extensions (`scripts/build-gnome-extensions.sh`). - - `[parity]` — what Bluefin inherits from Fedora's base image and Hummingbird - has in its repository but not in its bootable base; each entry's Requires - was resolved against Hummingbird plus the pinned factory before adding it. - `[services]` — desktop services Bluefin adds on top of the server base. - `[unavailable]` — Bluefin contract packages none of Utah's repositories provide. diff --git a/packages/utah.toml b/packages/utah.toml index b860ff335..ecd3f5714 100644 --- a/packages/utah.toml +++ b/packages/utah.toml @@ -8,7 +8,6 @@ # Sections: # [gnome] GNOME 51 desktop contract Hummingbird does not ship # [hardware] firmware and microcode the bootable base leaves out -# [parity] what Bluefin gets from Fedora's base image and Hummingbird has # [build] toolchain needed to build the pinned GNOME extensions # [unavailable] Bluefin contract packages none of Utah's repositories provide # @@ -95,70 +94,6 @@ packages = [ "iwlegacy-firmware", ] -[parity] -# Packages Bluefin's image carries without naming them in bluefin.toml, because -# Fedora's Silverblue base supplies them, and which Hummingbird's repository -# has but its server-shaped bootable base leaves out. Measured, not guessed: -# the image-level parity report (#117, scripts/check-image-parity.py) listed -# 277 such names against public-hummingbird-x86_64-rpms, and each entry here -# had its Requires resolved against Hummingbird plus the pinned factory -# repository before it was added. Aligned with Dakota where it ships the same -# piece: NetworkManager's Bluetooth and WWAN plugins, the connectivity check, -# ALSA use-case profiles, the language pack and the core font set. -# -# Left out, with the dependency Hummingbird does not carry: -# cups, cups-client need cups-filters -- no driverless-only cups -# exists as a package, so printing stays a gap -# vim-enhanced needs libgpm.so.2 (vim-minimal is in the base) -# default-fonts-core-emoji needs google-noto-emoji-color-fonts -# default-fonts-core-math needs stix-fonts -# fprintd, fprintd-pam in no repository; authselect's with-fingerprint -# is enabled but has no daemon behind it yet -# alsa-utils in no repository; Dakota builds it for the -# mixer-state restore (dakota#1027) -# unzip already in [build], which the image keeps -# -# NetworkManager-wifi is deliberately not here yet: its supplicant and the -# regulatory database are being built by utah-packages#133, and it joins this -# list when the factory image that carries them is pinned (#97, -# utah-packages#136). avahi and nautilus are the same shape, see [unavailable]. -packages = [ - # Networking - "NetworkManager-bluetooth", - "NetworkManager-wwan", - "NetworkManager-config-connectivity-fedora", - "bind-utils", - "iptables-nft", - # Audio - "alsa-ucm", - # Locale and fonts - "langpacks-en", - "langpacks-fonts-en", - "abattis-cantarell-fonts", - "default-fonts-core-sans", - "default-fonts-core-serif", - "default-fonts-core-mono", - "google-noto-sans-mono-vf-fonts", - "google-noto-serif-vf-fonts", - "google-noto-sans-symbols-2-fonts", - "google-noto-sans-math-fonts", - # Shell and tooling Bluefin users reach for - "bash-completion", - "man-db", - "less", - "which", - "file", - "hostname", - "xxd", - "zip", - "dbus-tools", - "keyutils", - "crypto-policies-scripts", - "poppler-utils", - "python3-pip", - "systemd-container", -] - [build] # Build dependencies for scripts/build-gnome-extensions.sh. # diff --git a/scripts/install-packages.py b/scripts/install-packages.py index 149bd7d4a..8e9407277 100755 --- a/scripts/install-packages.py +++ b/scripts/install-packages.py @@ -57,9 +57,6 @@ def contract(base: Path, overlay: Path, major: str | None) -> list[str]: # Firmware is part of the contract too: the bootable base ships almost # none, so a driver that needs a blob finds nothing to load. packages += section(overlay, "hardware") - # Parity with what Bluefin inherits from Fedora's base image and Hummingbird - # has in its repository but not in its bootable base. - packages += section(overlay, "parity") # Service packages are part of the desktop contract as well: 40-services.sh # cannot enable what the server base never installed. packages += section(overlay, "services") diff --git a/scripts/verify-rpm-contract.py b/scripts/verify-rpm-contract.py index 9789f73c3..45156e535 100644 --- a/scripts/verify-rpm-contract.py +++ b/scripts/verify-rpm-contract.py @@ -62,26 +62,23 @@ def main() -> int: contract = [line for line in resolved.read_text().split() if line] gnome_names = set(section(overlay, "gnome")) hardware_names = set(section(overlay, "hardware")) - parity_names = set(section(overlay, "parity")) service_names = set(section(overlay, "services")) - overlay_names = gnome_names | hardware_names | parity_names | service_names + overlay_names = gnome_names | hardware_names | service_names bluefin = [p for p in contract if p not in overlay_names] gnome = [p for p in contract if p in gnome_names] hardware = [p for p in contract if p in hardware_names] - parity = [p for p in contract if p in parity_names] services = [p for p in contract if p in service_names] else: bluefin = [p for p in section(args.manifest, "fedora") if p not in unavailable] gnome = section(overlay, "gnome") hardware = section(overlay, "hardware") - parity = section(overlay, "parity") services = section(overlay, "services") nvidia = list(NVIDIA_PACKAGES) if "nvidia" in flavor else [] - expected = [*bluefin, *gnome, *hardware, *parity, *services, *nvidia] + expected = [*bluefin, *gnome, *hardware, *services, *nvidia] print( f"Verifying {len(bluefin)} Bluefin packages, {len(gnome)} GNOME desktop packages," - f" {len(hardware)} firmware packages, {len(parity)} parity packages," + f" {len(hardware)} firmware packages," f" {len(services)} desktop service packages, and {len(nvidia)} NVIDIA packages", flush=True, ) diff --git a/tests/test_package_resolution.py b/tests/test_package_resolution.py index 9e299911f..37b7bc215 100644 --- a/tests/test_package_resolution.py +++ b/tests/test_package_resolution.py @@ -78,7 +78,6 @@ def resolve(self, output, code=1): '[fedora_v44]\npackages=["release-specific"]\n') overlay.write_text('[gnome]\npackages=["shell"]\n' '[hardware]\npackages=["firmware"]\n' - '[parity]\npackages=["manpages"]\n' '[services]\npackages=["resolver"]\n' '[build]\npackages=["compiler"]\n' '[unavailable]\npackages=["unavailable"]\n') @@ -94,8 +93,8 @@ def test_valid_declined_transaction_includes_every_install_section(self): rc, command = self.resolve("Transaction Summary:\nInstall 12 Packages\nOperation aborted.\n") self.assertEqual(rc, 0) self.assertEqual(command[command.index("install") + 1:], - ["base", "release-specific", "shell", "firmware", "manpages", - "resolver", "compiler"]) + ["base", "release-specific", "shell", "firmware", "resolver", + "compiler"]) self.assertIn("--assumeno", command) self.assertIn("--disablerepo=*", command) for repo in installer.REPOS: @@ -278,29 +277,6 @@ def test_packages_whose_closure_does_not_resolve_stay_out(self): self.assertNotIn(pkg, contract) self.assertIn(pkg, installer.section(self.OVERLAY, "unavailable")) - def test_parity_section_reaches_the_install_set(self): - contract = installer.contract(ROOT / "packages/bluefin.toml", self.OVERLAY, "44") - for pkg in installer.section(self.OVERLAY, "parity"): - with self.subTest(package=pkg): - self.assertIn(pkg, contract) - - def test_parity_section_names_nothing_the_contract_already_asks_for(self): - # A name in [parity] that bluefin.toml or another overlay section already - # lists would be a duplicate claim, and the verifier rejects duplicates. - parity = installer.section(self.OVERLAY, "parity") - others = set(installer.section(ROOT / "packages/bluefin.toml", "fedora")) - for name in ("gnome", "hardware", "services", "build", "unavailable"): - others |= set(installer.section(self.OVERLAY, name)) - self.assertEqual(sorted(set(parity) & others), []) - self.assertEqual(len(set(parity)), len(parity)) - - def test_verifier_asserts_the_parity_section(self): - section = self.verifier.section(self.OVERLAY, "parity") - self.assertTrue(section) - source = (ROOT / "scripts/verify-rpm-contract.py").read_text() - self.assertIn('parity = section(overlay, "parity")', source) - self.assertIn("*parity,", source) - def test_verifier_asserts_the_hardware_section(self): # The off-image --check path builds `expected` from the manifest. If # [hardware] were missing from it, firmware could go absent from a From ddb85f66d2d25e5d1f4ed42dc6e60c1a60ed5042 Mon Sep 17 00:00:00 2001 From: James Reilly Date: Mon, 21 Sep 2026 12:08:07 +0530 Subject: [PATCH 6/8] fix: wire input-remapper through desktop contract and docs MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The X230 report enabled input-remapper.service in configure-services.sh and the preset, but nothing asserted it: contracts/bluefin-desktop.toml [services].enabled did not list it, so verify-desktop-contract never checked the fix and a later regression would be silent — the same gap #98's bluetooth work was deferred to #125 for. Add input-remapper.service to the contract and to the desktop-contract skill doc that enumerates the asserted units. Also bring the package skill and README counts current: the overlay now carries [hardware] and [parity] (49 Utah additions, 11 unavailable), which the manifest header already describes but the docs still quoted as 12 and 4, and the install-set sentence omitted the two new sections. Assisted-by: Muse Spark via Muse Code Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --- README.md | 4 ++-- contracts/bluefin-desktop.toml | 1 + docs/skills/desktop-contract.md | 2 +- docs/skills/package-contract.md | 12 +++++++----- 4 files changed, 11 insertions(+), 8 deletions(-) diff --git a/README.md b/README.md index 29ecdf750..caecb7ff1 100644 --- a/README.md +++ b/README.md @@ -61,8 +61,8 @@ being noticed later. | | count | | --- | --- | | Bluefin contract installed | **61** | -| Utah additions (GNOME 51, desktop services) | 12 | -| Genuinely unavailable | **4** | +| Utah additions (GNOME 51, hardware, parity, desktop services) | **49** | +| Genuinely unavailable | **11** | The install writes its resolved list to `/usr/share/utah/contract.txt` and the verify step asserts *that file*, so the two cannot disagree. diff --git a/contracts/bluefin-desktop.toml b/contracts/bluefin-desktop.toml index 2c3268211..fa4c8136e 100644 --- a/contracts/bluefin-desktop.toml +++ b/contracts/bluefin-desktop.toml @@ -136,6 +136,7 @@ enabled = [ "flatpak-nuke-fedora.service", "flatpak-preinstall.service", "gdm.service", + "input-remapper.service", "ublue-system-setup.service", "uupd.timer", ] diff --git a/docs/skills/desktop-contract.md b/docs/skills/desktop-contract.md index a27f71665..a09ee247a 100644 --- a/docs/skills/desktop-contract.md +++ b/docs/skills/desktop-contract.md @@ -57,7 +57,7 @@ The TOML's sections are the contract's table of contents: - **`[services]`** — systemd units the preset must enable: `gdm.service`, `ublue-system-setup.service`, `flatpak-preinstall.service`, `flatpak-nuke-fedora.service`, `brew-setup.service`, `dconf-update.service`, - `bootc-unified-storage.service`, `uupd.timer`. + `bootc-unified-storage.service`, `input-remapper.service`, `uupd.timer`. ## GNOME extensions are pinned submodules diff --git a/docs/skills/package-contract.md b/docs/skills/package-contract.md index 412da7e4c..32ee66e29 100644 --- a/docs/skills/package-contract.md +++ b/docs/skills/package-contract.md @@ -38,12 +38,13 @@ policy for changing them. the header comment of that file (cite it; do not move or copy it): - `[gnome]` — GNOME 51 desktop contract Hummingbird does not ship. - - `[build]` — toolchain needed to build the pinned GNOME extensions - (`scripts/build-gnome-extensions.sh`). + - `[hardware]` — firmware and microcode the bootable base leaves out. - `[parity]` — what Bluefin inherits from Fedora's base image and Hummingbird has in its repository but not in its bootable base; CI's package availability step resolves the real transaction and is the gate on every name there. + - `[build]` — toolchain needed to build the pinned GNOME extensions + (`scripts/build-gnome-extensions.sh`). - `[services]` — desktop services Bluefin adds on top of the server base. - `[unavailable]` — Bluefin contract packages none of Utah's repositories provide. @@ -111,7 +112,8 @@ about installation. is a **build failure**. `just check-repos` reads the base and package-image digests from `Containerfile`, copies the same repository configuration, and runs `install-packages.py --resolve` inside that base. This includes the - release-specific Bluefin section, GNOME, services, and extension build tools. + release-specific Bluefin section, GNOME, hardware, parity, services, and + extension build tools. It needs Podman and network access. A name lookup on GitHub Pages is not evidence that the pinned OCI repository is complete or ABI-compatible. The factory's leading metadata layer is checked against the pinned manifest @@ -125,8 +127,8 @@ about installation. (`just check-parity`). Current counts, per the README "Package parity" section: 61 Bluefin contract -packages installed, 12 Utah additions (GNOME 51, desktop services), 4 -genuinely unavailable. +packages installed, 49 Utah additions (GNOME 51, hardware, parity, desktop +services), 11 genuinely unavailable. ## Verification From 4c6173ebfa4cb05d7f3464311118441f77f05fed Mon Sep 17 00:00:00 2001 From: "sec-check[bot]" Date: Mon, 21 Sep 2026 03:53:41 -0400 Subject: [PATCH 7/8] docs: correct Bluefin contract installed count to 58 The parity table and the package-contract skill doc still claimed 61 Bluefin contract packages installed. The contract check resolves [fedora] (65 names) minus the 7 [unavailable] names it contains, which is 58: scripts/verify-rpm-contract.py --check packages/bluefin.toml packages/utah.toml Verifying 58 Bluefin packages, 11 GNOME desktop packages, 5 firmware packages, 28 parity packages, 5 desktop service packages, and 0 NVIDIA packages The additions (49) and unavailable (11) counts in the same rows were already updated; this brings the third figure in line with them. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --- README.md | 2 +- docs/skills/package-contract.md | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index caecb7ff1..f64c6855a 100644 --- a/README.md +++ b/README.md @@ -60,7 +60,7 @@ being noticed later. | | count | | --- | --- | -| Bluefin contract installed | **61** | +| Bluefin contract installed | **58** | | Utah additions (GNOME 51, hardware, parity, desktop services) | **49** | | Genuinely unavailable | **11** | diff --git a/docs/skills/package-contract.md b/docs/skills/package-contract.md index 32ee66e29..7d4d254ef 100644 --- a/docs/skills/package-contract.md +++ b/docs/skills/package-contract.md @@ -126,7 +126,7 @@ about installation. - Drift in `packages/bluefin.toml` from upstream is a CI failure (`just check-parity`). -Current counts, per the README "Package parity" section: 61 Bluefin contract +Current counts, per the README "Package parity" section: 58 Bluefin contract packages installed, 49 Utah additions (GNOME 51, hardware, parity, desktop services), 11 genuinely unavailable. From 40a5a4206595d1a65c287ff0318944f564b12d33 Mon Sep 17 00:00:00 2001 From: castrojo Date: Tue, 22 Sep 2026 15:51:10 -0400 Subject: [PATCH 8/8] chore: regenerate site data for firmware and hardware units --- site/data/packages.json | 31 ++++++++++++++++++++++++++++--- 1 file changed, 28 insertions(+), 3 deletions(-) diff --git a/site/data/packages.json b/site/data/packages.json index 5ca8f5a1a..1b9347f0f 100644 --- a/site/data/packages.json +++ b/site/data/packages.json @@ -1,5 +1,5 @@ { - "generated_at": "2026-09-19", + "generated_at": "2026-09-22", "generator": "scripts/generate-site-data.py", "groups": [ { @@ -123,6 +123,19 @@ "title": "Base-image parity", "transient": false }, + { + "blurb": "Device firmware the bootable base leaves out, so a driver that needs a blob can find one.", + "id": "hardware", + "packages": [ + "iwlegacy-firmware", + "iwlwifi-dvm-firmware", + "iwlwifi-mld-firmware", + "iwlwifi-mvm-firmware", + "linux-firmware" + ], + "title": "Firmware", + "transient": false + }, { "blurb": "Units Bluefin adds on top of the server base; the preset cannot enable what was never installed.", "id": "services", @@ -151,8 +164,8 @@ } ], "totals": { - "installed": 102, - "unavailable": 9 + "installed": 107, + "unavailable": 11 }, "unavailable": [ { @@ -161,6 +174,12 @@ ], "name": "anaconda-live" }, + { + "issues": [ + 104 + ], + "name": "avahi" + }, { "issues": [], "name": "evolution-ews-core" @@ -187,6 +206,12 @@ ], "name": "libgda-sqlite" }, + { + "issues": [ + 100 + ], + "name": "nautilus" + }, { "issues": [ 107