v3.6 — Set a response-time budget per endpoint, catch "slow but still up" before it turns into an outage, and a faster, lighter cloud refresh. PER-ENDPOINT TIMEOUTS • Give each endpoint its own request timeout (1–60s) instead of the one-size-fits-all 15s. Tighten it for snappy APIs, loosen it for backends that are slow but fine. • Works the same for local and cloud checks — cloud endpoints honor your timeout on the server too. NEW "SLOW" WARNING • Set a response-time threshold and PingOps flags an endpoint that's still up but slower than your budget — a yellow "Slow" badge instead of plain green. • Softer than a failure: the endpoint is up, it's just over your latency budget. Catch creeping slowness before it becomes a real outage. • History tags the exact checks that breached, so you can see when it started. FASTER CLOUD REFRESH • Cloud sync now pulls your whole fleet's history in a single request instead of one per endpoint — noticeably quicker and lighter, especially on large fleets. What v3.5 shipped, still here: • Pull-to-refresh is local only; cloud sync has its own button. • Small fresh-install pulls — last hour on cold start, lazy 24h per endpoint on open. • Actionable cloud-sync error banners + "Force resync from cloud". What v3.3 shipped, still here: • Public status pages (Pro & Archon) — never exposes endpoint URLs, headers, or bodies. • Full localization to 日本語 / Español / Deutsch alongside English.

v3.5.1 — Pull-to-refresh is local only. Cloud sync gets its own button. Fresh-install pulls are dramatically smaller. PULL-TO-REFRESH IS LOCAL ONLY • Drag down to recheck local endpoints — that's it. No more spinner sitting through a cloud round-trip on every pull. • Cloud endpoints have a dedicated iCloud button in the dashboard toolbar. Tap it to sync server-side results on demand. • Foreground polling + scene activation still keep the dashboard fresh in the background without you doing anything. FRESH-INSTALL CLOUD SYNC IS WAY SMALLER • First launch (or returning after a long absence) now pulls only the last hour of cloud check results — sub-second on a fleet that used to download tens of thousands of rows on cold start. • Opening any endpoint detail view lazy-loads the last 24h for that one endpoint — sparkline + uptime band light up immediately on first open. CLOUD SYNC ERRORS NO LONGER SILENT • Cloud-side endpoints that fail to land in your dashboard now surface an actionable error banner instead of disappearing without trace. • New "Force resync from cloud" option in Cloud Settings — clears every cloud endpoint's sync cursor and re-pulls. Use it if a reinstall left the dashboard missing endpoints that exist on another device. What v3.5 shipped, still here: • Background catch-up + atomic refresh during long cloud sync. • Absolute time labels (14:23 / May 24, 23:59) instead of stale "5 minutes ago". • Memory-safe streaming writer path for multi-day offline catch-ups. What v3.3 shipped, still here: • Public status pages (Pro & Archon) — never exposes endpoint URLs, headers, or bodies. • Full localization to 日本語 / Español / Deutsch alongside English. • Bidirectional cloud sync with reinstall recovery.

v3.5 — Background catch-up, atomic refresh, absolute time labels. CLOUD CATCH-UP IS SMOOTH NOW • Foreground after hours offline? The dashboard stays responsive while the catch-up downloads in the background, then snaps to the new state in one update at the end — no more "is it frozen?" pause while it syncs. • Pull-to-refresh on a 30-endpoint fleet feels atomic: the whole list updates together instead of rippling row by row. • The loading banner now stays pinned to the top of the dashboard while a long catch-up runs — you can scroll your endpoints and still see progress. ABSOLUTE TIME LABELS • Goodbye stale "5 minutes ago" — those snapshots never actually ticked, so they got stale the moment the row drew. • Same day → local 24-hour time (e.g. 14:23). Older → date + time (e.g. May 24, 23:59). Always honest about WHEN something happened, in your device's timezone. UNDER THE HOOD • One SQLite save per cloud pull instead of one per endpoint — eliminates the multi-hang "MainActor frozen" feel during long catch-ups. • Streaming, memory-safe cloud-pull writer path for the worst case (30-endpoint Archon, multi-day offline). • Live Activity + notification body construction no longer fault entire endpoint histories — constant-time SQLite limit-1 fetches instead. What v3.4 shipped, still here in v3.5: • Dashboard responsiveness during cloud syncs (background-actor SwiftData writes, denormalised hot-path reads, predicate-based history pruning). • Parallel pull-to-refresh fan-out (capped to 6 concurrent), 60-second server-cadence poll, smarter notification/Live Activity gating. • Always-visible subscription plans, persistent history filter, orbital splash matching the icon. What v3.3 shipped, still here: • Public status pages (Pro & Archon) — never exposes endpoint URLs, headers, or bodies. • Full localization to 日本語 / Español / Deutsch alongside English. • Bidirectional cloud sync with reinstall recovery.

v3.4 — performance overhaul + UX polish. PERFORMANCE (the big one) • Dashboard stays responsive while cloud history loads in the background — writes happen off the main thread via a dedicated background actor, so scroll, taps, and navigation keep working during a poll tick. • Pull-to-refresh is snappy now. Local checks fan out in parallel (capped to 6 concurrent); cloud sync runs in the background and surfaces its own progress banner — the gesture spinner no longer waits on the slow path. • Pro/Archon foreground polling now matches the 1-minute server cadence (was 30s) — half the work, same freshness. • Notification + Live Activity evaluators only fire when something failure-relevant lands. Most polls land successful rows only and now skip the SLO walk entirely. • Major reduction in main-thread work: denormalised endpoint state for the dashboard rows, predicate-based history pruning, and an in-actor session-token cache that stops thrashing the keychain. UX • Subscriptions screen always shows every plan — even Archon. Your current plan is tagged "Current"; upgrade options stay one tap away. • History filter ("All" / "Failures only") persists across launches and navigation. • New animated splash background — orbital glow that matches the app icon. • Live Activity sparkline cleanup on endpoint delete (no orphan tail entries between recreates). EARLIER IN THIS UPDATE (v3.3) • Public status pages (Pro/Archon) — shareable read-only view of your cloud endpoints' uptime; URLs/headers/body never exposed. • Full localization to 日本語 / Español / Deutsch on top of English, with a first-launch language picker and a Settings → Language row. • Bidirectional cloud sync — cloud is the source of truth; cloud rows removed on another device flip back to Local locally (history preserved, never silently deleted). • Reliability fixes: StoreKit → cloud tier sync the moment Pro/Archon is recognised, "Clear history" wipes cloud too, first cloud check fires within ~30s of endpoint create, SwiftData cold-launch crash fixed.

v3.3 — public status pages + full localization. PUBLIC STATUS PAGE (Pro/Archon) • New read-only status page you can share with your team or customers — shows live state of your cloud-monitored endpoints. • Publish from Cloud → Public status page. URL format: arkynate.com/pingops/status/{token}. • Surfaces only the friendly name, status, uptime % (24h / 7d / 30d), 24h latency sparkline, and last incident. URLs, headers, and request bodies are intentionally NEVER exposed. • Polls every 60s; the proxy caches the response for 30s so a viral page can't hammer your D1. • Revoke any time from the same screen — the token becomes a 404 the moment you do. LOCALIZATION • PingOps is now translated to 日本語 / Español / Deutsch on top of English. Every screen, error message, push alert, and Live Activity string speaks your language. • First-launch language picker lets you pick the language before the onboarding carousel — so the rest of onboarding lands in your chosen language. • Settings → Language lets you switch later. iOS Language & Region remains the default; the in-app pick overrides it for PingOps only. • Localized App Store listing pages in those four languages — search, screenshots, and description all match. UNDER THE HOOD • Pricing localized to 175 storefronts via Apple's automatic pricing (¥600 / €3.99 / £3.49 / R$ 16.90 for Pro, etc.). • Proxy migration 0006 adds an indexed status_page_token column to users; no impact on existing rows. RELIABILITY (v3.3.1) • Pro and Archon subscriptions now sync from StoreKit to the cloud the moment they're recognised on the device — no more "you're at the Free tier limit" message right after upgrading. • Switching an endpoint from Cloud to Local in its settings now also removes it from the cloud (no more ghost-monitored rows). • "Clear history" now wipes cloud check results too, not just on-device. • First cloud check for a new endpoint fires within ~30 seconds of creation instead of waiting a full interval. WHAT'S NEW IN v3.3.2 • "Sync with cloud" — bidirectional reconciliation. Cloud is the source of truth: local rows tagged as Cloud whose cloud counterpart no longer exists are flipped back to Local (history preserved, never deleted). • Dashboard pull-to-refresh now also pulls fresh cloud history for cloud endpoints, using a per-device cursor — only new check results travel the wire. • Subscription page is always accessible. Pro users see Archon upgrade; Archon and Patron users see Manage Subscription (cancel, downgrade, view billing). • Duplicate now respects your tier cap up front — no more silent ghost copies when you're at the limit. • Fixes a SwiftData crash that could trip when evaluating failure notifications on a cold launch.

v3.2 — Push notifications, Live Activity, and cloud sync all get a reliability + polish pass. PUSH NOTIFICATIONS • Re-notify while down — if you miss the first alert, you'll get a reminder every 15 minutes while an endpoint stays failing. The bubble replaces in place on your Lock Screen, not stacking. • APNs holds for up to an hour — pushes survive a phone briefly off Wi-Fi or in low-power mode. Before, Apple dropped them on first failed delivery. • "Notify on failure" now applies to Cloud endpoints too — the per-endpoint toggle in the form was iOS-only before, so the proxy alerted regardless. Same for the streak threshold and first-failure switches. • Friendlier alert text when a site rejects automated checks — "Rate-limited — this site is blocking our cloud checks" instead of the bare HTTP code. LIVE ACTIVITY • Slimmer Lock Screen card — roughly half the previous height. • Snooze 10m + Snooze 1h actions actually work (the previous build's intent dispatch was misrouted). Tapping a snooze dismisses the card immediately and recomputes the failing count. • Re-check removed — Cloud monitoring runs its own re-check schedule, so the manual button was duplicating work. CLOUD MONITORING • Pro and Archon now genuinely get the 1-minute check cadence the paywall promises. The old build was sending every cloud endpoint at 5 minutes regardless of tier. • Foreground sync polls every 30 seconds while the app is open — Cloud rows feel live without pull-to-refresh. • Self-heal — endpoints whose initial cloud mirror failed silently re-register on the next sync, so a one-shot network hiccup at sign-up no longer leaves an endpoint dark. • Bot-block detection — when a site returns 429 or 403 with anti-bot signatures, the failure reason explains it and suggests a probe-friendly endpoint instead of showing a raw status code. DASHBOARD + UX • Cloud cadence banner explains how server-side checks differ from pull-to-refresh. • Dashboard + Cloud tab badges show the unhealthy count and any session-expiring-soon warning. • iPad detail pane shows a fleet summary instead of an empty placeholder. • Search filter handles "no matches" with a clear-search button. • Sheet detents on Auto-refresh, Move to group, and cURL import so they stop covering the dashboard. • "Send debug bundle" in Settings → About exports a redacted JSON snapshot for support threads — endpoint names + URLs intentionally NOT included. If you're new to PingOps: CORE • On-device monitoring — local mode is free, no account, never phones home. • Cloud monitoring — Sign in with Apple, flip an endpoint to Cloud mode, and a Cloudflare-backed proxy checks it on a tier-appropriate cadence (Free 5 min, Pro & Archon 1 min). Push alerts when something breaks. • Validates status codes, response keywords, and JSON fields — not just "did the socket open." • Live Activity for active incidents — snooze without unlocking. PLANS • Free — 2 cloud endpoints, 5-min interval, 24h history. • Pro — 10 cloud endpoints, 1-min interval, 30-day history, per-endpoint pushes, weekly SLO digest, 3-device sync. From $3.99/mo or $34.99/yr. • Archon — 30 cloud endpoints, 90-day history, SSL expiry tracking, Family Sharing, 5-device sync, plus everything in Pro. From $9.99/mo or $84.99/yr. PRIVACY • Cloud monitoring is opt-in per endpoint. Local endpoints never leave your device. • Request bodies and custom headers are encrypted per-user. • Aggregate alerts contain only a count — never endpoint names or URLs. Use of PingOps and any in-app subscription is governed by Apple's Standard EULA: https://www.apple.com/legal/internet-services/itunes/dev/stdeula/

v3.1 — Cloud Pro and Archon now ship every promised feature. NEW • Per-endpoint push notifications (Pro / Archon) — get a dedicated alert the moment an individual endpoint goes down, in addition to the rolling aggregate count. Snooze, recheck, and Open jump straight into that endpoint's history view. • Weekly SLO digest (Pro / Archon) — Sunday morning push with the past week's uptime % and the endpoints that had the most trouble. • Multi-device sync — server-enforced per tier (Free 1 / Pro 3 / Archon 5). Reinstalling on the same phone no longer counts as a new device. You get a clear "limit reached" prompt with an upgrade path on Free and Pro. • SSL expiry tracking (Archon) — see the days remaining on every HTTPS endpoint's TLS certificate. The check runs on your device when you open an endpoint, so open it weekly (or tap Check now) to keep the date fresh. Continuous server-side SSL monitoring is in the works for a future update. CHANGED • Multi-region checks have been removed from the Archon feature list while we ship the proper multi-pop infrastructure. Archon pricing is unchanged — the 30-endpoint cap, 90-day history, SSL expiry tracking, Family Sharing, and 5-device sync remain the value prop. If you're new to PingOps: CORE • On-device monitoring — local mode is free, no account, never phones home. • Cloud monitoring — Sign in with Apple, flip an endpoint to Cloud mode, and a Cloudflare-backed proxy checks it every minute, every day. Push alerts when something breaks. • Validates status codes, response keywords, and JSON fields — not just "did the socket open." • Live Activity for active incidents — recheck or snooze without unlocking. PLANS • Free — 2 cloud endpoints, 5-min interval, 24h history. • Pro — 10 cloud endpoints, 1-min interval, 30-day history, per-endpoint pushes, weekly SLO digest, 3-device sync. From $3.99/mo or $34.99/yr. • Archon — 30 cloud endpoints, 90-day history, SSL expiry tracking, Family Sharing, 5-device sync, plus everything in Pro. From $9.99/mo or $84.99/yr. PRIVACY • Cloud monitoring is opt-in per endpoint. Local endpoints never leave your device. • Request bodies and custom headers are encrypted per-user. • Aggregate alerts contain only a count — never endpoint names or URLs. Use of PingOps and any in-app subscription is governed by Apple's Standard EULA: https://www.apple.com/legal/internet-services/itunes/dev/stdeula/

v2.0 — major design + reliability update. • Faster status picker and Test on Edit in the endpoint form • Cleaner detail layout — pills replace duplicated quick-action row • Refined preview row trailing layout and heatmap allocation • Unified snooze banner copy • Dual-task background scheduling (refresh + processing) with diagnostics • Crash fix when moving subfolders between folders • Improved light theme color contrast for WCAG AA • No more residual "cancelled" toasts on pull-to-refresh Privacy-first as always — all data stays on your device.