First real end-to-end auto-update on a live install failed like this:
Automatically updating trivy (a recovery snapshot is taken first)
Snapshotting trivy before update…
Pulling new image(s) for …
Update of failed — rolling back…
Could not roll back automatically
The app name went empty after the snapshot. Cause: bash is dynamically
scoped, so a callee assigning an undeclared variable writes the CALLER's
local of that name — and a `while read app` loop leaves it EMPTY at EOF.
webuiBackupAppStatus's dashboard generator runs at the end of every backup
and did exactly that to updaterApplyApp's `app`.
Nothing was damaged: the pull ran against an empty name, failed before
touching the image, and the rollback was a no-op on a nonexistent app.
Fixed both ends. The generator (and three gluetun loops with the same
latent leak) now declare `local app`. updaterApplyApp/updaterRollbackApp
hold the name in `_upd_app` so they no longer depend on every callee's
hygiene, and updaterApplyAll stops leaking its own loop var.
This is exactly the untested path the roadmap flagged: "apply/revert not
yet exercised end-to-end on a live install with a pending update."
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Adds the decision half of the app updater. Detection (P2) and the
snapshot-first apply/revert (P3) were already real, but nothing ever
pressed the button — every update waited for a click.
CFG_<APP>_UPDATE_TYPE=auto|manual per app, default auto (33 templates)
CFG_UPDATER_AUTO=true|false master switch, default true
updaterAppPolicy resolves the two the way backupResolveStrategy already
resolves backup strategy: the global switch can only make things more
manual. updaterApplyAuto runs at the end of `updater check` and enqueues
the ordinary updater_apply task for each auto app that has an update —
never applies inline, so an automatic update is the same code path, task
log, History entry and Roll back button as a manual one.
Safety: each attempt stamps its target digest under generated/auto/, so a
build that fails is rolled back and then left alone rather than retried on
every scan; in-flight updater tasks are skipped so scans can't stack.
Tracked end to end: updates.json carries each app's resolved update_type,
History entries carry trigger=manual|auto. The WebUI says whether updates
install themselves, chips only the apps that opted out, labels automatic
history, and — since an auto app's pending update needs no decision — keeps
it off the Overview board's "Needs action" view.
Also fixes artifactApplyAuto enqueueing without --detach: it runs inside
the single-threaded task processor's own poll, so following the new task in
the foreground waits for a task that cannot start until it returns.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The board landed on "Everything", so the rows that want a decision were
mixed in with the healthy one-liners on arrival. Default to the "Needs
action" chip instead, falling back to "Everything" when nothing is
pending (that view would otherwise be empty). An explicit chip click
still sticks for the session.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Cut the 36-word blurb to 23 without losing either fact that matters: it's
DNS-based, and blocked domains resolve to a local blackhole.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Tags and Paths were stacked full-width blocks, so a single tag chip and a
single path each burned a whole row and left the panel mostly empty. Wrap
them in a .bsm-blocks auto-fit grid that seats them side by side and falls
back to stacking under ~460px. The wrapper now owns the divider, so it
renders once for the pair instead of once per block, and is omitted
entirely when neither block has content.
Applied to both renderers of this markup: the global Backups view and the
per-app backup card.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The App Center blurb ran three sentences, two of which restated things
the Security view already makes obvious. Keep the what, drop the rest.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The Security triage implied 'fix available · update the app to clear these'
even when the app was already on the newest image — where updating does
nothing and the patch only lands when the maintainer rebuilds. That read as a
false to-do. Reframe honestly:
- Status line keyed on update_available, not Trivy's fixed_in: up to date =>
'nothing to apply; clears when the maintainer ships a rebuilt image'; update
available => 'updating may pull in patched packages'.
- Groups relabeled to describe reality: 'Patch released upstream' (hint: lands
on rebuild / may be cleared by updating) and 'No patch yet'. Counts go
neutral so a big number doesn't read as either alarm or all-clear.
- Row 'no fix yet' -> 'no patch'; use getAppDisplayName so it's 'Trivy' not
'trivy'.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Consistency pass across the App Center Overview area: no tab has a top-level
manual "Check" button anymore — scans are automatic.
- Overview tab: remove the header "Check now"; the hero already reports last-scan
time, and the unscanned sub-copy no longer points at a button that's gone.
- Improvements tab: remove the header "Check"; lead its body with the same
auto-check line (right-side Check-now nudge) as the Updates tab.
- Remove the now-unused checkBtn helper.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A user rightly noted the Security section read as a wall of unrelated
dependency CVEs against an 'Up to date' app — no cue for what, if anything,
to do. Make it answer 'is this my problem, and will updating fix it?':
- Scanner (trivy_scan.sh): stop discarding Trivy's Class/Type/Status at the
jq flatten — bind them onto each vuln so the UI can tell an OS package from
the app's own bundled dependency, and a real fix from a won't-fix.
- Security section (updater-page.js): explain these are vulnerabilities in the
packages bundled in the image (not the app version), tally 'N with a fix ·
M no fix yet', then split the list into a 'Fix available' group (worst-first,
each row tagged OS/dependency and showing installed -> fixed) and a dimmed
'No fix yet' group. No fabricated 'this update fixes N' claim — fixed_in vs
the image tag isn't a reliable join, so we only state fix availability.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Extends the per-app Updates treatment to the fleet App Center → Overview →
Updates list, and folds the manual check into the status line so it's the one
canonical (secondary) affordance rather than a top-level tab button:
- updater-page.js: renderAutoCheckLine() now ends with a right-aligned "↻ Check
now" button (data-updater-action="check" — both surfaces already wire it).
- overview-manager.js: drop the "Check"/"Check now" button from the Updates tab
header (keep "Update all", only when updates exist); lead renderUpdates() with
the auto-check line.
- overview.css: .updater-autocheck wraps on narrow widths; .updater-autocheck-btn
sits right (margin-left:auto), smaller.
- app-tabbed-manager.js: friendlier no-data copy ("You're all caught up — no
updates found for this app.").
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The per-app Updates tab pushed a manual "Check" button (header + empty state)
even though scans run automatically on CFG_UPDATER_SCAN_INTERVAL — so an app with
nothing to update read like an empty/actionable page. Replace the manual Check
with a calm status line inside the panel: "Checked automatically · last checked X
· next check ~Y", backup-schedule style. The genuine Apply/Roll back actions stay
(applying is still manual and safe). No auto-apply.
- webui_updater_scan.sh: stamp scan_interval_minutes alongside generated_at in
updates.json so the display needs no separate config fetch (0 = auto off).
- updater-page.js: renderAutoCheckLine() + fmtRelFuture().
- app-tabbed-manager.js: drop the header/empty-state Check buttons; render the
auto-check line; friendlier no-data copy.
- overview.css: style .updater-autocheck (green dot live / muted when off).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The expanded-row deep-link pushed /overview/<tab>?app=<app> — wrong prefix
and a ?query the SPA's path-based router drops on a cold load, so the row
never reopened from a shared URL. Switch to /apps/overview/updates/<app>,
matching the Migrate/Backups sub-tab path pattern, and parse the app from
that path segment in _honorAppDeepLink so the row expands on cold load too.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The Security section in the per-app detail panel dropped the CVE list flush
against the panel edges (no side padding, no visible frame). Wrap it in an
inset dark rounded .updater-cve-box (side padding + border) so it reads as a
contained block matching the app rows, and make the scroll thumb more present
(wider, higher-contrast, padded track) so a long list clearly scrolls.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Reduce .ov-row-head vertical padding (12px -> 5px) and shrink the app icon
tile (32px -> 24px) so each Updates row is roughly half as tall.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Redesign the Fleet Overview Updates list rows to match the Tasks page:
each app now shows its icon tile, a status pill ('✓ Up to date' /
'↑ Update available' / '• Unscanned'), and a dedicated 'Details' toggle
button (chevron) in the row actions instead of the bare leading chevron.
Resolve the app icon from the slug (/core/icons/apps/<slug>.svg, hidden on
error) and prefer window.getAppDisplayName for a prettier title.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The task-log "Loading logs..." state rendered as a bare left-aligned text
line (or a tiny ad-hoc 16px spinner overlay), which read as unfinished next
to the rest of the UI. Swap all three log-loading spots (initial placeholder,
toggle-open fetch, running-task stream placeholder) to the canonical
window.lpLoadingBox('Loading logs…'), and scope .lp-loading inside the log
terminal box to fill it and drop its own card chrome so the spinner sits dead
centre over the terminal surface instead of a box-in-a-box. Widen the stream
overlay-removal selector to also clear .lp-loading.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The per-app Updates tab's recessed panel used padding: 4px 16px, so the first
section (VERSION) sat 4px from the top edge — looking unpadded — while the last
section ended 20px from the bottom. Match the sibling .backup-snapshots-container
idiom (uniform 16px) and zero the first/last sections' outer padding so the
container's inset sets an even top/bottom margin.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A fresh install ran webuiLibrePortalUpdate twice back-to-back: once in
installLibrePortal step 11, then again in startScan at the end of preinstall.
The 30s time-debounce meant to collapse them is fragile (it never fired on a
recent install — >30s elapsed between the two), and debouncing is the wrong
lever anyway: startScan's pass runs AFTER scanConfigsForRandomPassword
finalises app passwords, so it — not the step-11 pass — is authoritative.
Defer the step-11 generation deterministically during a bootstrap install
(libreportal_bootstrap_install=true), leaving startScan's single pass to do the
work. Standalone reinstalls (no bootstrap flag) still generate in step 11. The
time-debounce stays as a general back-to-back backstop.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The authoritative icon source containers/trivy/trivy.svg held a generic
hand-drawn shield/magnifier placeholder, while the frontend copy had the
official brand mark. The icon sync (webuiSyncAppIcons) treats the source as
canonical, so it was clobbering the official logo with the placeholder on
every full WebUI update/install — leaving the setup wizard's app tile
showing the generic icon. Replace the source with the official logo so it
sticks.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Trivy had no /core/icons/apps/trivy.svg (the canonical app-icon path used by
task rows, tiles, tools, services, routing), so it fell back to the default.
Add the official Trivy mark, white on its #1904DA brand tile so it stays legible
on the dark, near-transparent icon tiles across every theme.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
The type glyph (🩺/🔄/✅…) sat bare next to the framed .task-app-icon tile. Give
it the same 32px rounded tile (border + surface-elevated bg) so the two read as
a matched pair; spacing now comes from .task-info's gap like the app icon.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
Previously retry silently created a new task, left the failed one in place with
no link, and didn't take you to the retry. Now:
- The new task is linked to the failed one via `retryOf`; the server stamps the
original `retriedBy` and KEEPS it (its log is the failure record — never
deleted), so history survives and there's no confusing bare duplicate.
- The failed row shows a muted "↻ Retried" pill, hides its now-stale Retry
button, and offers "View retry" to jump to the new run. The new run shows a
"Retry of:" backlink to the original.
- After retrying, the UI selects the new task and opens its live log so you
follow the retry instead of hunting for it.
retryOf rides through the existing POST /api/tasks (no new endpoint); createTask
gained an optional extra-body arg.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
Add a `#LIBREPORTAL|<APP>_VERSION_TAG|<current-tag>` sentinel to each app's
anchor (`<slug>-service`) image line, so a `CFG_<APP>_VERSION` config drives
that tag through the existing tagging system — the config becomes the source,
the compose tag is derived (like ports/IPs/domains already are). This is the
"install/pin a specific version" knob.
- Placeholder is the literal current tag (e.g. `31-fpm-alpine`), NOT a `*_DATA`
token: an unset var leaves the line untouched (inert comment, keeps the real
tag) and never trips the up_app stale-tag scanner.
- Only the anchor line is tagged (sidecars mariadb/redis/nginx stay
tracked-by-digest, not user-version-picked). ollama correctly targets
`ollama/ollama`, not the companion open-webui.
- Untagged anchors normalized to `:latest` (semantic no-op) so they're
templatable too. The manager (libreportal) is skipped — it updates via its
release channel, not a docker tag.
32 apps wired. Verified: CFG_NEXTCLOUD_VERSION=32 → `nextcloud:32` (prefix
preserved, idempotent, sidecars untouched); all compose files still valid YAML.
Inert until a version config is set, so no behaviour change on existing installs.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
"Control plane" is jargon. Rename the self-heal task and de-jargon all the
user-facing copy: task titles ("LibrePortal - Fix System Issues" / "System
Health Check"), the action label, the badge/banner/panel text and "Fix now"
button, the status summaries, and the heal task-log messages. Behaviour
unchanged; code comments keep the technical term where accurate.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
The task-list row only paints the LibrePortal app-logo for commands matched by
isLibrePortalSystemTask's whitelist (appless tasks otherwise get just the type
icon, so arbitrary custom commands don't get a spurious logo). `libreportal
system health/network heal|check` weren't in it — only `system reclaim|image` —
so the heal rows showed a bare function icon instead of the logo like "Check for
Updates". Add health|network to the whitelist.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
Crash-loop detection keyed on `docker ps --filter status=restarting`, but a
backed-off loop sits "exited" between restarts once docker's backoff grows to
tens of seconds — so a slowed loop is missed. Detect via RestartCount CLIMBING
between scans (what a crash loop actually is), unioned with the instantaneous
restarting signal for fast loops. Baseline counts persist in
.health_restart_counts, written only by the throttled check so the heal's
re-scans don't disturb the delta.
Also give the system_health_heal / system_network_heal tasks proper display
(they fell through to the raw command + generic ⚙️): friendly titles in
formatCommandForUser, type icons (🩺 / 🌐) in getTaskTypeIcon, and action
labels in formatActionTitle — so they read as "LibrePortal - Repair Control
Plane" etc. with an icon, like install tasks.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
An offline trivy install crash-looped (server FATALs when it can't fetch the
vuln DB), and on rootless docker the restart storm churned the shared network's
port-forwarder until the WebUI's own published host port was torn down — the
WebUI stayed healthy INSIDE its container but was unreachable from the host, with
nothing detecting or healing it.
Three fixes, in the house self-healing style (mirrors the network-drift trio):
1. Control-plane health checker wired into the existing task-processor idle poll
(maybeRegenPoll), no new daemon. dockerHealthScan (read-only) detects daemon
down, a WebUI running-but-host-port-unreachable (the port-forward corruption),
and crash-looping containers. webuiSystemHealthCheck writes
frontend/data/system/health_status.json + self-dispatches a heal — the user
can't click a button on a dead WebUI, so the poll drives the fix. Frontend
health-notifier surfaces a topbar badge + dashboard banner + details panel.
2. Failure cap, enforced centrally by dockerHealthHeal (task-gated): stops
crash-loopers (removing the churn), restarts the WebUI to re-publish a lost
port forward, and — only if that fails — recycles the rootless daemon and
restarts the core container. Caps every app immediately, no template churn.
3. Trivy no longer crash-loops offline: the server runs in a shell retry-loop so
the container stays Up and quietly retries on a backoff instead of exiting
FATAL. Verified: container stays Up across repeated DB-download failures.
Core WebUI compose gains restart: unless-stopped so it self-recovers after a
reboot / daemon recycle instead of staying down.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
Follow-up to the backup-refresh throttle/dedupe, cutting the cost of the
remote pulls that do still happen.
* restic stats now runs in --mode raw-data (restic_check.sh). The default
restore-size mode walks every snapshot's tree to sum logical file sizes —
the slowest restic op — just to fill a size readout. raw-data reads the
index only and reports the repository's actual deduplicated on-disk size,
which is exactly what the dashboard already labels "deduplicated,
encrypted". raw-data omits total_file_count, so the per-location card now
shows that location's snapshot count (already loaded client-side, and more
useful for a backup repo) instead of a file count.
* engineLocationStats now shares the same per-refresh memoiser as
engineSnapshotsJson (engine_dispatch.sh). Both the locations and dashboard
generators call it per location, so repo stats went from two restic calls
per location per refresh to one. Factored the cache into _engineCachedPull.
* SSH connection reuse for SFTP locations (backup_ssh.sh): ControlMaster=auto
with a self-reaping ControlPersist master, so the several restic
subprocesses a refresh/backup spawns against one location share a single
authenticated connection instead of a fresh handshake each — the dominant
per-call cost on a high-latency link. Toggle via CFG_BACKUP_SSH_MULTIPLEX.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
The prior .ov-row override tied .updater-row on specificity, and updater.css
loads later, so `display:grid` won and the detail stayed boxed. Use the
two-class selector .updater-row.ov-row to win regardless of load order.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The fleet Updates expander row inherited .updater-row's 3-column grid, so an
open row's detail body (CVEs, recovery, history) landed in a narrow right-hand
`auto` column with dead space to its left. Make .ov-row a block: the head stacks
on top and the detail body uses the full width.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
An app with many CVEs (e.g. Trivy's 28) rendered every row full-height, pushing
the page down. Extract one renderCveList() shared by the standalone Security tab
and the per-app expander: sort worst-severity-first, and once past ~6 rows cap
the height (260px) with an internal scroll + themed scrollbar and a bottom fade
hint. Add a count pill to the expander's "Security" heading. Per-app sections
stay stacked and independently collapsible as before.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The client scan resolved images by pulling from Docker Hub, which fails on an
offline/privacy box ("index.docker.io ... network is unreachable") — yet every
installed app's image is already present locally. Point Trivy at the docker
socket that's already bind-mounted into the container (--image-src docker, plus
DOCKER_HOST=unix:///run/user/<uid>/docker.sock for rootless, derived from the
install user — rooted's default path is found automatically). Scans local
images with zero network. Verified: aquasec/trivy:latest -> 28 CVEs.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Standalone `docker exec trivy-service trivy image ...` opens the vuln-DB cache
the long-lived server already holds, failing with "cache may be in use by
another process: timeout" — so every scan silently returned [] (apps: 0 even on
vulnerable images). Point the exec'd client at the server (--server
http://localhost:4954, the container's fixed --listen port); the server owns the
DB, the client just submits the image. Verified against alpine:3.10.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Trivy runs as a server whose vulnerability DB downloads on first boot; until it
lands no scan can produce results. Previously the updater generator wrote an
empty-but-valid cves.json the moment the file was missing, so installing Trivy
painted a green "no known vulnerabilities" all-clear that was actually a lie —
the DB hadn't even downloaded, and the Updates/Security view gave no signal.
Add an honest scanner state the WebUI branches on:
- containers/trivy/scripts/trivy_scan.sh — trivyScannerState (absent |
db_updating | ready) via `trivy version -f json`, trivyDbUpdatedAt, and
trivyScanImageCves (per-image scan normalized to {id,severity,package,
installed,fixed_in,url}, deduped). All degrade safely on error.
- webui_updater_scan.sh — stamp cves.json with scanner.state; only run real
per-image scans once the DB is ready. Always rewritten so state tracks live.
- updater-page.js — Security tab shows a loading box while the DB updates, an
install nudge when absent, and the genuine 🎉 only when ready+empty; Overview
CVE card sub + hint reflect the state.
- overview-manager.js — fleet Security row surfaces the "building CVE database"
pending state instead of silently omitting.
- function_manifest.sh — regenerated for the new trivy_scan.sh functions.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
New containers/trivy app (aquasec/trivy in server mode, private API port,
docker socket + cache volume) in the security,recommended categories, plus
placement in the setup wizard recommended step, the server-side install
tier after crowdsec, and the CLI first-install prompt. The updater's CVE
scan (design doc P4) will gate on this app being installed; §5 of the
design doc updated to record the app-based decision.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
With no apps tracked the sub line read '0 apps tracked · 0/0 backed up ·
last scan …' — pure noise. Show only the scan time in that case.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
.backup-retention-preset-block carried a permanent border-bottom (+ padding),
but the Keep-* fields it separates the dropdown from are hidden for every
preset except Custom — so on Self-hosting/etc. it was an orphaned line with
empty space beneath it. Moved the divider + spacing onto an .is-custom state
that applyVisibility() toggles alongside the fields.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A background task like the updater check surfaces its result in the page it was
launched from (the Updates/Overview tab repaints itself), yet it still fired the
full "View Task" pair — "Check task started!" then "Check task completed!" —
on top of the launcher's own small "Checking apps for updates…" line. Three
notifications for something you have no reason to open the Tasks page for.
Add a small, generic classifier (LP_BACKGROUND_TASKS) keyed by task action/type
(with a command regex that also catches the backend `updater check auto` run).
For a classified task:
- executeTask() skips the standard "task started!" toast and, when the run is
hand-launched, records its id as pending.
- the taskCompleted handler skips the standard "task completed!" toast; if the
id was pending (user launched it) it shows one small plain result line
("Apps checked for updates & vulnerabilities."), otherwise stays silent so
the periodic auto-scan makes no noise at all.
Non-background tasks are unaffected — the standard task-notification style still
applies. Extend byAction to quiet another self-surfacing action.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
In the config grid (e.g. Backups > Configuration > Engine) toggle pills sat
~6px below and 2px taller than the text/select controls beside them, so they
read as misaligned. Two causes, both scoped to .config-fields:
- .field-group carried a vestigial margin-bottom:6px (the grid's own 16px gap
already spaces rows). It inflated the grid track, so the align-self:end
toggle landed 6px below the input controls. Zeroed inside the grid.
- The .checkbox-label pill's 24px switch + 10px vertical padding rendered 46px
vs the inputs' 44px. Trimmed to 9px padding so the pill is 44px.
Toggles now share the exact vertical band (top + height) as the inputs in
their row. Applies to every config-grid toggle, not just the backup engine.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
On /admin/config/general the Basic section's fields sit in the darker inset
".config-fields" box, but the Mail Enabled toggle and Installation Mode select
did not — their control sits in its own .mail-master-toggle / .git-master-toggle
wrapper, a sibling of .config-fields, which the boxing rule never covered. So a
gated section's enable row read as loose, unboxed content above the (hidden)
options.
Extend the boxed-panel rule to the master-toggle wrappers (mail / install-mode /
generic / backup-remote). Element-qualified to the wrapper <div> because
renderToggleSection stamps the same class on the inner .checkbox-label pill too,
which must not get its own nested box.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
When the official catalog is hidden, its "Re-enable" action rendered as a
flat, greyed-out ~27px pill pinned to the top-left of the row — shorter than
the 44px URL field and not obviously clickable (the generic .btn-secondary
loses its nebula-theme polish outside .config-actions, so it fell back to an
unstyled base button).
- The action column now align-self:stretch, so its control matches the URL
field's height and sits inline (also vertically centres the compact × in the
enabled state).
- The Re-enable button gets a dedicated .catalog-reenable-btn treatment: accent
fill + accent border + hover state + an eye icon, filling the row height so it
lines up with the input and clearly reads as a button.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The Backups tab's Configuration sub-tab showed only the warning banner and
an orphaned divider -- the actual config form (retention presets, engine,
keep-policies, Save/Reset) never appeared.
Root cause: a duplicate id. The Fleet Overview, the app grid and per-app
detail all render the shared apps-unified-layout, which keeps a hidden
#config-tab > #config-section in the DOM. The Backups Configuration sub-tab
created its OWN #config-section, and configManager.renderConfig() resolved
the target with getElementById('config-section') -- which returns the FIRST
(hidden) node in document order. So the 21KB form painted into the invisible
per-app container while the backup tab's container stayed empty. Because the
/backup routes now redirect into /apps/overview/backups, this fired every
time -- the form had effectively never been visible.
Fix:
- renderConfig(category, target) now accepts an explicit container (element
or id) and falls back to #config-section; the error-retry button re-targets
via configSection.id.
- The Backups Configuration sub-tab uses a unique id (#backup-config-section)
and passes it to renderConfig; retention-preset and engine-details
selectors are scoped to it.
Also: the Refresh button now spins + locks while refreshAll()+render() run,
so the click has visible feedback instead of appearing to do nothing.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The global Backups tab rendered its snapshot list (and the "No backups
yet." empty state) as a bare flex column directly on the tab panel, so
the list section had no visual container -- unlike the Dashboard tab's
cards and the per-app Backups tab's recessed panel. Wrap it in a
.backup-snapshot-panel box (translucent fill + hairline border, matching
.backup-card) so the section reads as a bounded surface.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The embedded backup center deep-links *in* fine (mountBackupCenter parses the
sub-segment), but clicking a sub-tab never wrote the URL: BackupPage disables
pushTabToUrl() when embedded and nothing on the fleet side stepped in, so the
address bar stayed stale and the sub-tabs weren't shareable/deep-linkable —
unlike the sibling Migrate sub-tabs.
Give BackupPage an onTabChange hook fired on every switch; the OverviewManager
supplies it to replaceState /apps/overview/backups/<sub> (replaceState, matching
Migrate, so the on-entry deep-link re-sync can't stack a duplicate history
entry). Standalone /backup keeps using pushTabToUrl() unchanged.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
The badge centres on the input box (top:50%), but the single-line input renders
its text a hair low, so the pill read as slightly above the URL. +2px lands it
optically inline.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Signed-off-by: librelad <librelad@digitalangels.vip>
