1. Transparent cursor theme: 1x1 pixel Xcursor for every shape, set as
system default via XCURSOR_THEME=betterframe-empty. Nuclear fix for
Pi 5 GPU ignoring XCURSOR_SIZE.
2. Full VT lockdown: mask ALL gettys (tty1-6 + templates), logind
NAutoVTs=0 + ReserveVT=0, mask emergency/rescue targets. Ctrl+Alt+Fx
reaches nothing. No login screen ever.
3. Auto-reboot: FailureAction=reboot-force + StartLimitAction=reboot-force
on kiosk unit. If cage/app can't stay running → system reboots rather
than showing a blank screen or login prompt.
4. Purge ALL Pi setup wizards: piwiz, userconf-pi, rpi-first-boot-wizard,
initial-setup, pi-greeter, rpd-plym-splash. Nuke autostart files,
mask systemd units. "Configure your Raspberry" never shows.
Terminal: idle_add_local_once from non-GTK thread silently fails.
Forward ShowTerminalCode/DismissTerminalCode through WorkerMsg channel
which IS polled on the GTK main thread via timeout_add_local.
Journal: try --user-unit first, fall back to unfiltered journal if
permission denied (bfkiosk user may not be in systemd-journal group on
non-reflashed images). Send error line back to admin UI on spawn failure
instead of silent drop.
Three fixes:
1. Terminal code overlay replaces the main display window's child instead
of creating a new gtk::Window (cage compositor only shows one window).
Saves the previous child and restores on dismiss.
2. Code auto-expires after 60s — timeout does NOT increment lockout.
GTK overlay dismissed + pending_code cleared.
3. Journal-start handler already logs but relay might fail silently if
kiosk WS reconnected after admin debug WS connected.
Kiosk side (remote_debug.rs + ws_client.rs refactor):
- Journal streaming: server sends journal-start → kiosk spawns
journalctl -f, pipes lines back as journal-line messages via WS.
journal-stop kills the process. On-demand, not always-on.
- Terminal: server sends terminal-request → kiosk checks lockout +
firmware_channel == "dev" → generates 8-char code displayed on
screen as fullscreen overlay (NOT logged) → server relays admin's
code via terminal-auth → kiosk validates with constant-time compare
→ on success spawns bash, relays I/O as base64 terminal-data.
- Lockout: 3 failed codes per boot → lockout_count++. 3 lockouts
(9 total failures) → permanent (reflash only). Reboot resets
attempt counter, not lockout counter. Successful pairing resets all.
- ws_client.rs rewritten with split reader/writer + tokio::select!
for multiplexing incoming WS messages with outbound journal/terminal
data from sync threads.
Server side (coordinator-ws + routes-admin):
- New admin debug WS endpoint: /ws/admin/debug/:kioskId. Authenticated
via admin API key (query param) or session cookie. Relays messages
bidirectionally between admin browser ↔ kiosk.
- Admin pages: /admin/kiosks/:id/logs (journal viewer with start/
stop/clear) and /admin/kiosks/:id/terminal (code entry + terminal
area). Both open in new tabs from the kiosk detail page.
- Angie proxy config updated with /ws/admin/debug/ location block.
Security:
- Terminal only on dev channel
- Code displayed physically on screen, never logged or stored server-side
- Lockout: 3/boot, 3 lockouts = permanent, pairing resets
- Kiosk responds "locked" without specifying which lockout triggered
Coolify doesn't include .git in Docker build context, causing build
failure. Revert to ARG-based version stamping: compose passes
BF_SERVER_VERSION from Coolify's SOURCE_COMMIT/COOLIFY_GIT_COMMIT
env vars as a build arg, Dockerfile writes it to .bf-version. Removed
git from builder apt install (no longer needed).
importDiscoveredCamera was hardcoded to type="rtsp", losing ONVIF
identity. Camera edit showed RTSP fields, ONVIF event subscription
skipped (checks cam_type=="onvif"), re-discovery impossible.
Now creates type="onvif" with onvif_host/port/username/password stored
on the camera row. Streams still go into camera_streams (unchanged).
Bundle ships onvif fields → kiosk subscribes to PullPoint events.
Also passes host + port as hidden form fields from discover results
page so the add handler has them available. Basic manual camera
creation via UI stays rtsp-only (simpler); discovery flow produces
onvif type.
/opt/betterframe/kiosk/ now owned bfkiosk:bfkiosk so OTA can write
.new/.prev files. Marker path in Rust code aligned with rollback
script expectation (/var/lib/betterframe/kiosk/firmware-applying.json).
Coolify pulls from GitHub and runs docker compose build — no guaranteed
env vars like SOURCE_COMMIT. Previous approach relied on ARG/ENV
passthrough that silently defaulted to "dev".
Fix: install git in the builder stage, COPY .git into context, run
git describe --tags --always to derive the version, write it to
/app/server/.bf-version. version.ts reads this file as a fallback
between env vars and the "dev" literal.
Chain: BF_SERVER_VERSION env → BF_BUILD_VERSION env → .bf-version file
→ COOLIFY_GIT_COMMIT env → SOURCE_COMMIT env → "dev".
Also: fix .gitignore for rauc-signing/ (was under wrong path).
Use hx-retarget/hx-reswap response headers to replace full grid
when cell dimensions change (overlap resolution may move other cells).
Single-cell swap when only content fields change.
Kernel dropped to initramfs because root=LABEL=BF_ROOT_A in cmdline.txt
but the ext4 filesystem had no label set (pi-gen's default is unlabeled).
dd copies raw bytes — any label must be set on the standalone file BEFORE
writing into the output image.
Add e2label BF_ROOT_A on rootfs.ext4 + fatlabel BF_BOOT_A / BF_BOOT_B
on each bootfs copy after patching cmdline.txt but before dd.
losetup -fP partition scanning failed on CI runner ("failed to open
partition 1"). Rewrite to parse partition start/size from sfdisk -J
(JSON output) via jq, then dd with skip/seek at exact sector offsets.
Only uses losetup for individual file images (selector.vfat, rootfs,
bootfs) where partition scanning isn't needed.
Also: add jq to CI apt install, drop xz compression from -9 to -6
(faster, still ~85% ratio on rootfs), free source image earlier to
avoid disk exhaustion on runners with tight scratch.
Camera edit page gains a "Live Events" panel that auto-refreshes every
5s via htmx. Shows last 20 events for this camera from event_log:
topic, source type, timestamp, and raw payload JSON. Surfaces ALL
ONVIF topics including unknown ones — if a camera produces an event
type we haven't seen before, it shows up here immediately.
queryEvents gains camera_id + source_type filters. Route
GET /admin/cameras/:id/events returns an HTML fragment with the event
table rows.
Replace naive per-cell shift with iterative overlap resolver. When any
cell grows (direction expand, dim/delta resize, or cell edit with new
spans), all overlapping cells get pushed along the expansion axis, with
cascading to prevent pushed cells from overlapping each other.
Fixes: expanding left block goes under right block instead of pushing it.
- Bake @flowfuse/node-red-dashboard into Node-RED Docker image
- Fire-and-forget syncDashboardsFromNodered() on GET /admin/entities
so dashboard tabs appear without manual sync button click
Adds Clone Layout button to layout edit page. Duplicates the layout
with all cells, label bindings, and display attachments. Name gets
"(copy)" suffix with dedup.
New kiosk/src/onvif_events.rs: for each ONVIF camera in the bundle,
creates a PullPoint subscription, polls every 3s, parses
NotificationMessage XML into structured JSON (topic + source key/values
+ data key/values + timestamp), and POSTs to /api/kiosk/event with
source_type=onvif + camera_id.
Forwards ALL event topics: motion, ANPR (LicensePlateRecognition),
line crossing, intrusion, digital input, analytics, tamper — everything
the camera exposes. Node-RED sorts what matters.
Subscription lifecycle:
- CreatePullPointSubscription with 60s InitialTerminationTime
- Renew every 55s before timeout
- Unsubscribe on bundle change / shutdown
- Auto-resubscribe on pull/renew failure (30s backoff)
- Generation tracking via Weak<()> so old workers self-terminate
when start() is called with a new bundle
WSSE PasswordDigest auth for SOAP calls — same scheme the server's
onvif.ts uses. sha1 crate added.
BundleCamera extended with onvif_host/port/username/password_encrypted
fields (server already ships them; kiosk just wasn't deserializing).
Gated by BF_ENABLE_ONVIF_EVENTS=1. Enabled by default in the pi-gen
image env file.
TODO: cluster-key-based decryption of onvif_password_encrypted. For
now relies on the RTSP URI having plaintext credentials embedded (which
the ONVIF import path already ensures via rtspWithCredentials).
Two fixes:
1. When admin renames a camera, the linked entity's name now syncs
automatically so the entity list doesn't drift from the camera list.
2. ONVIF discovery now calls GetDeviceInformation before GetProfiles
(best-effort, catches auth-gated devices). Pulls Manufacturer + Model
and uses the combined string as the camera's proposed name instead of
the raw IP. E.g. "Hikvision DS-2CD2146G2" instead of "192.168.74.8".
Falls back to host IP when the device omits the info.
Camera edit page now shows a "Kiosk Subscriptions" table: every kiosk
whose layouts reference this camera, which specific layout names, and
whether the camera is in the kiosk's active layout (green "active"
badge) or just bundled (gray "bundled" badge).
Snapshot route switched from listKiosksRenderingCamera (active-only)
to listKiosksWithCameraInBundle (any layout). The kiosk's LAN endpoint
opens a one-shot RTSP connection from its own network position even when
the camera isn't on screen — no warm pipeline needed. Server falls back
to direct pull only when NO kiosk has the camera in any layout at all.
Every column + table added inside an already-passed PRAGMA user_version
entry is re-created here with IF NOT EXISTS / addColumnIfNotExists so
existing deploys finally pick them up:
- kiosks: reported_hostname, network_interfaces_json
- kiosks: os_update_channel, os_update_target_version, os_update_last_*
- kiosks: managed_image, managed_config_*
- displays: active_layout_id
- os_update_releases table + indices
- os_update_rollouts table + indices
Rule going forward: NEVER add columns/tables inside existing migration
entries. Always append a NEW entry at the end of the MIGRATIONS array.
`secrets` context isn't available in step-level `if:` expressions inside
a reusable workflow_call. Move the secret-presence check to job-level
env (HAS_RAUC_SECRETS, HAS_AUTOIMPORT) and reference those in step if:.
New module kiosk/src/at_rest.rs. Derives an AES-256-GCM key via HKDF
from a Pi-bound value:
1. /proc/device-tree/serial-number (Pi 5 firmware exposes it)
2. /proc/cpuinfo Serial line (older kernels)
3. /etc/machine-id (non-Pi dev fallback)
File format: "BFE1" magic || 12-byte random nonce || ciphertext+tag.
Atomic write via tempfile + rename so a crash mid-write can't leave a
half-encrypted file.
Wired into kiosk/src/server.rs at every file I/O touching sensitive
state:
- kiosk.key (bearer token to BF server)
- local.key (LAN-side API auth key)
- bundle.json (cached bundle with RTSP credentials in URL form)
Migration: read paths tolerate legacy plaintext (kiosks upgraded from a
pre-at_rest build) AND re-store as ciphertext on the first read. One-
shot upgrade — subsequent boots skip the migration write.
Threat model defended: SD card extraction. Attacker who pulls the card
can't decrypt without also having the same physical Pi (CPU serial is
hardware-bound). Doesn't defeat an attacker who has both — at that
point they ARE the kiosk. Bar is raised from "trivially extract every
camera password" to "must steal the device intact."
Not defended: TPM-style attestation, remote attestation, sealed boot.
Pi 5 has no TPM and we don't ship a secure-boot config.
Tests in-module: round-trip short bytes, round-trip JSON, legacy
plaintext passthrough.
Mirrors the kiosk-firmware admin shape one-for-one against OS RAUC
bundles:
/admin/os-updates release list, yank
/admin/os-updates/rollouts rollout list + create
/admin/os-updates/rollouts/:id/state pause/resume/complete
/admin/kiosks/:id/os-update per-kiosk channel + pin
Templates: OsUpdatePage, OsUpdateRolloutsPage, KioskOsUpdatePanel.
KioskOsUpdatePanel is rendered next to the existing KioskFirmwarePanel
on the kiosk detail page so OS + app state sit side-by-side. The
"how bundles get here" sidebar on the list page documents the four
GitHub secrets needed (signing cert/key + autoimport URL/key) so a
new operator doesn't have to dig through scripts/ to find them.
Nav gains an OS Updates entry between Firmware and Labels. Activates
on activeNav="os-updates".
Repo + import endpoint already existed (audit confirmed earlier). All
admin routes use them as-is.
Two image-side hardening pieces both small enough to ship together.
deploy/nftables/nftables.conf — single ruleset installed at /etc/nftables.conf.
Default-drop input. Allowed: loopback, established/related, ratelimited
ICMP, kiosk local API :18090 from RFC1918 / RFC4193 / link-local sources
only. SSH stays gated by sshd-disabled (image build sets enable-ssh: 0
and 01-run-chroot masks it); the firewall rule for :22 is left commented
in for triage scenarios. Forward dropped. Output left wide open — kiosk
needs to dial out to arbitrary RTSP cameras + the BF server (which may
live on the public internet) without explicit allowlisting.
deploy/systemd/betterframe-firstboot.{service,sh} — runs once per device
before betterframe-kiosk starts. Generates a 24-char unambiguous-glyph
password, applies via chpasswd, stores at /etc/betterframe/admin-password
(0400 root), and prints a banner to tty1 so an HDMI-attached operator
can transcribe it during the boot window before cage takes over the
screen. Marker at /var/lib/betterframe/.firstboot-complete prevents
re-run on subsequent boots. Without this, every kiosk built from the
same image shipped with bfadmin:betterframe — a single password leak
compromises the entire fleet.
Future follow-up: post the rotated password (encrypted with cluster_key)
to the BF server via heartbeat so admin UI can surface it. Not in this
commit; the local file + tty banner are the only retrieval paths today.
Phase 2b. Bake the runtime side of RAUC into the curated image so a
freshly-flashed kiosk can accept .raucb bundles immediately:
- Add `rauc` + `dosfstools` to the apt package list.
- Drop deploy/rauc/system.conf to /etc/rauc/system.conf (already declares
the A/B slot layout that repartition-image.sh produces).
- Drop deploy/rauc/betterframe-rauc-boot.sh to
/usr/local/sbin/betterframe-rauc-boot.sh — the custom bootloader
backend that flips the BF_BOOTSEL autoboot.txt to point at the
freshly-installed slot via Pi 5 tryboot.
- Drop deploy/rauc/ca-cert.pem (operator-supplied, committed) to
/etc/rauc/keyring.pem so rauc can verify CMS signatures. If the cert
isn't committed yet, image build emits a workflow warning and the
kiosk image installs but refuses every bundle — image still flashes,
just no OS OTA until the cert is committed.
- Enable BF_ENABLE_OS_OTA=1 in /etc/default/betterframe-kiosk so the
kiosk Rust consumer actually polls for bundles. Set to 0 to pin OS
version for a specific kiosk.
mark-good was already wired (deploy/systemd/betterframe-rauc-mark-good.{service,sh}).
The kiosk's heartbeat loop also calls `rauc status mark-good` as a
belt+suspenders backup; both are idempotent.
Phase 2a of OS OTA: post-process pi-gen output into a RAUC-compatible
A/B layout. New deploy/rauc/repartition-image.sh:
- Decompresses the stock pi-gen 2-partition image
- Extracts bootfs (vfat) + rootfs (ext4) blobs
- Compacts rootfs with resize2fs -M and grows back with 25% headroom
- Patches /etc/fstab inside rootfs to use LABEL=BF_BOOT_A /
LABEL=BF_ROOT_A / LABEL=BF_DATA (slot-agnostic; RAUC re-labels per
slot on install)
- Stamps /etc/betterframe/{os-version,os-compatibility} for the kiosk's
os_update.rs to read at runtime
- Builds two bootfs copies, each with cmdline.txt root= rewritten to
the matching ROOT slot
- Lays out 6 GPT partitions: BF_BOOTSEL (autoboot.txt with tryboot
pointing at boot_partition=2 / [tryboot] boot_partition=3), BF_BOOT_A,
BF_BOOT_B, BF_ROOT_A (populated), BF_ROOT_B (empty, RAUC fills on
first install), BF_DATA
- Recompresses with xz -T0
build-bundle.sh now takes the already-extracted slot images so the
.raucb bundle re-uses the exact same blobs that ship inside the A/B
initial-flash image — no duplication, no drift.
CI wires the repartition step between pi-gen output and the GitHub
Release upload. Ships the A/B image (not the stock pi-gen one).
Also: bump Blacksmith binary builders from 2/4 vCPU to 8 vCPU each.
Image job stays on GitHub's ubuntu-24.04-arm (Blacksmith arm kernel
6.5 doesn't ship binfmt_misc as a loadable module, which pi-gen-action's
defensive modprobe step still requires).
What's still pending:
- In-image RAUC install (rauc package + drop system.conf + CA cert
at /etc/rauc/keyring.pem). Without this, the image boots A/B-laid-
out but rauc install commands have no daemon to talk to.
- Admin UI for OS releases + rollouts (task #4).
