Show a detail line below the alert table for the selected row
with untruncated config, health status, last sent time, and
send/fail counts. Solves the CONFIG column truncation issue
without requiring the full-screen detail view.
Redesign the inline detail panel from a single left-aligned column
to a two-column layout that uses the full panel width.
Left column (55%): latency chart, min/avg/max stats, 30-day
uptime timeline.
Right column (45%): status/latency/last-check summary, type-
specific metadata (HTTP response code + SSL expiry, ping hostname,
port target, push interval/last heartbeat, DNS resolver), uptime
percentage + retry count, and up to 5 state changes listed
vertically with error reasons.
Narrow terminals (<80 cols) fall back to the original stacked
single-column layout.
Thread context.Context from callers instead of creating
context.Background() at every call site. Engine stores its
lifecycle ctx for use by triggerAlert goroutines. TUI Model
carries ctx for store operations in Cmd closures. CLI commands
create a root ctx and thread it through openStore, seed
functions, and init methods.
Only two intentional context.Background() remain in non-root
positions: engine constructor default (overridden by Start)
and drainWrites (parent already cancelled at shutdown).
Split monitor.go (1131→350 lines) into alerts.go, checks.go,
sites.go, and maintenance.go by concern. Split sqlstore.go
(833→470 lines) into sqlstore_alerts.go, sqlstore_history.go,
and sqlstore_maintenance.go by domain. Tests move with their
implementation. Pure reorganization — no behavioral changes.
Replace inline numeric literals with named constants across 14 files:
server timeouts/rate limits, cluster thresholds/intervals, DB pool
sizes, alert/dial timeouts, TUI uptime thresholds, node status
thresholds, and state history limits.
Warn via log when creating a maintenance window that overlaps with an
existing one for the same monitor, parent group, or global scope.
Handles both timed and indefinite windows.
Bottom section: graphs (latency + histogram) 70% left, state changes
30% right. Top section rearranged: endpoint details left, status +
timing + HTTP right. Bottom section labels use bold accent without
trailing rules. Spacing between top and bottom halves.
Removed unused renderSparkTooltip (zone marks not wired in column
layout — can restore later).
Section labels (STATE CHANGES, LATENCY, RESPONSE TIME DISTRIBUTION,
PROBE RESULTS) now render in bold accent color with a trailing ───
rule that fills the width. Two-column info block separated by a full
divider below. Creates clear visual boundaries between sections.
Refactored full-screen detail panel: status+endpoint on left column,
timing+config on right column. Halves vertical height of the info
section. Content wraps in a viewport when it exceeds terminal height
— scroll with j/k/arrows/pgup/pgdn.
Restored renderSparkTooltip lost in rewrite. Fixed lint warnings.
Horizontal bar chart showing latency distribution across 5 buckets:
0-50ms, 50-100ms, 100-200ms, 200-500ms, 500ms+. Color-coded green
(fast), yellow (medium), red (slow). Shows count per bucket.
Appears in the full-screen detail panel (Enter) after the sparkline
stats. Skipped for push monitors (no latency data).
Statuspage-style colored bar — one █ per day, colored by uptime:
green (>= 99%), yellow (>= 95%), red (< 95%), gray (no data).
30-day percentage shown to the right.
Daily breakdown computed via ComputeDailyBreakdown from state_changes
and loaded alongside detail data in loadDetailCmd. Auto-updates on
cursor move.
Settings tab warn count now includes failed alert channels — alerts
where the last send failed (LastSendOK=false, LastSendAt not zero).
Shows as a red badge number alongside offline node count. Surfaces
broken alert channels without navigating to Settings.
Press < / > to cycle sort column (Status, Name, Latency). Press r to
reverse direction. Sorted column shows ▲/▼ arrow in the header.
Groups always float to top. Sort applies to ungrouped monitors and
group children independently. Default: Status descending (DOWN first).
Replace ntcharts with custom multi-row sparkline. Each bar is color-
coded by latency threshold (green < 200ms, yellow < 500ms, red > 500ms)
and DOWN checks render in red. 3 rows tall with 24 discrete levels.
Y-axis labels (max/min ms) on the left, Min/Avg/Max stats below.
Zero external dependencies — removed ntcharts.
Replaced streamline chart with ntcharts sparkline (block elements,
auto-scaling). Height=2, Y-axis labels (max/min ms) on the left,
Min/Avg/Max stats below. Denser and more readable than the line chart.
Replace block-element sparkline with ntcharts streamline chart in the
inline detail panel. Renders a 4-row line chart with thin line style
using the theme accent color. Auto-scales Y axis to latency range.
Added github.com/NimbleMarkets/ntcharts v0.5.1 dependency (lipgloss v1
compatible). Min/Avg/Max stats rendered below the chart.
Introduced models.LogEntry{Message, CreatedAt} to replace raw strings
in the log pipeline. Timestamps are now formatted at render time, not
baked into stored messages.
- Engine: appendLog stores LogEntry with time.Now()
- Store: LoadLogs returns []LogEntry, selects created_at from DB
- Store: strips legacy [HH:MM] prefix from pre-refactor DB entries
- TUI: sidebar shows "MM/DD HH:MM" from CreatedAt
- TUI: full log view shows "MM/DD HH:MM" from CreatedAt
- SaveLog still receives plain message string (DB handles timestamp)
Detail open/closed state saved via SetPreference on toggle and
restored on session start. Same pattern as theme persistence —
survives restarts and works per-user over SSH.
Click any panel (Monitors, Logs, Detail) to focus it — accent border
follows focus. Mouse wheel scrolls the focused panel.
Keyboard: l toggles log panel focus. Arrow keys scroll logs when
focused, navigate monitors when not. Esc returns focus to monitors.
Log sidebar now supports scroll offset — tracks position across
renders without a viewport. Mouse wheel scrolls 3 lines, keyboard
scrolls 1.
All panels wrapped in titled rounded borders (╭─ Title ──╮). Focused
panel gets accent-colored border, unfocused panels get muted border.
- Monitors panel: titled "Monitors", focused when detail is closed
- Logs panel: titled "Logs", always unfocused (passive display)
- Detail panel: titled with monitor name, focused when open
Table's own RoundedBorder replaced with HiddenBorder — the titled
panel border provides the visual frame, table uses space-separated
columns internally. Consistent chrome across all panels.
Log sidebar wrapped in rounded border (no left/bottom edge — shared
with monitors table). Creates visual separation between panels.
Enter on a monitor opens the full-screen detail view (existing
stateDetail) for deep dive — history, SLA, probe results, connection
chain. i stays as inline detail toggle.
Footer key hints now context-sensitive: show h/s/Enter when detail
is open, show full keybindings when closed.
Log sidebar was rendering all lines regardless of table height. When
detail panel was open and table shrank, the sidebar stayed tall, pushing
the detail panel past MaxHeight (clipped to empty). Now sidebar accepts
a maxLines parameter capped to table row count.
Press i to toggle a compact detail panel below the monitors+logs
split. Shows status, latency, uptime, state changes, sparkline, and
key hints in ~6 lines. Auto-updates when cursor moves between
monitors. h/s/e keys work from the inline detail for history, SLA,
and edit. Escape closes the panel.
No more full-screen detail takeover for the common case. The old
stateDetail path remains for h/s sub-views which still go full-screen.
Log lines now hard-clamped to panel width via lipgloss MaxWidth.
Stripped "Monitor " and "Push " prefixes from sidebar messages —
redundant in a monitoring app, saves 8 chars per line. Improved
prefix width calculation to prevent line wrapping at narrow widths.
Table columns were computed from terminal width, causing row wrapping
when the monitors panel only gets 70% of the space. Introduced
contentWidth field set per-tab in viewDashboard. computeLayout,
isWide, and renderTable now use contentWidth for column visibility,
available space, and max table width calculations.
Columns gracefully hide (SSL, RETRIES, TYPE, UPTIME) when the panel
is narrower, matching the existing responsive breakpoint behavior.
Replace full viewLogsTab with compact sidebar renderer for the 70/30
monitors split. Single-char severity icons (▼▲◆●·), truncated messages,
no header chrome. Renders directly from engine logs without viewport.
Tab bar: Monitors | Maint | Settings (was 6 tabs).
Settings tab merges Alerts, Nodes, Users as sub-sections with
left/right arrow navigation. Each section keeps its own cursor,
keybindings, and CRUD operations.
Monitors tab now shows a log sidebar at >= 120 cols (70/30 split).
Under 120 cols, monitors render full-width without logs.
- Introduced tab constants (tabMonitors, tabMaint, tabSettings)
- Introduced section constants (sectionAlerts, sectionNodes, sectionUsers)
- Removed stateLogs and stateUsers states
- All magic tab numbers replaced with named constants
Full-width horizontal rules above and below the content area. Tab bar
sits above the top divider, status/keys bar sits below the bottom
divider. Creates three clear visual zones: navigation, content, status.
Revert upper-third centering — inconsistent start positions across tabs
felt jumpy. Back to standard top-align with consistent table placement.
Demo GIF now pauses on Nodes tab (cluster view is a selling point) and
skips Maint/Users quickly instead of sprinting through all three.
Tables on tabs with few rows (Alerts, Nodes, Maint, Users) now sit in
the upper third of the viewport instead of flush against the tab bar.
Dense tabs like Monitors and Logs fill naturally and are unaffected.
Strip Go module pseudo-version suffix (timestamp+hash+dirty) from the
footer version string — shows "v0.1.0" instead of the full build
metadata. Rename "Sites" tab and breadcrumbs to "Monitors" for
consistency with README, CLI help, and user-facing docs.
Docker releases now dual-push to Docker Hub and ghcr.io/lerkolabs/uptop.
GoReleaser brews section generates a Homebrew formula and pushes to
lerkolabs/homebrew-tap on GitHub.
Requires new Gitea secrets:
- GHCR_USERNAME / GHCR_TOKEN for GHCR login
- HOMEBREW_TAP_GITHUB_TOKEN for tap repo push
- GitHub repo lerkolabs/homebrew-tap must exist
Re-captured all 6 TUI screenshots via VHS with realistic demo data.
Added animated demo GIF as README hero image and a 5-theme montage
strip showing Flexoki Dark, Tokyo Night, Catppuccin, Nord, Gruvbox.
LABEL_IDS was pretty-printed multi-line JSON, which breaks
GITHUB_OUTPUT's single-line format. Also compact GH_LABELS
before passing as --argjson to avoid multi-line toJSON expansion.
GitHub mirror previously redirected issue reporters to Gitea, which
requires login. Now GitHub Issues are accepted directly via form
templates (bug report + feature request) and a workflow forwards
new issues to Gitea with label mapping and provenance header.
- Disable healthcheck on probe compose services (no HTTP server)
- Remove stale "(Phase 4)" comment from dev compose
- Add data/ to .gitignore (compose volume creates deploy/data)
- Clarify -db-type flag help text (sqlite or postgres)
lerko
