In a news cycle dominated by conflict and endless scrolling, SunSettings returns attention to a universal, ever-present wonder: the sunset. Our team, swept up by political turmoil and displacement, has gained hard-earned clarity on mental health and deeply values these quiet, shared moments. With “third places” eroding in many cities, we use sunsets to create place and event ex nihilo - a reason to gather, meditate, sketch, play music, or share food in a park. The app turns that impulse into a simple, repeatable habit.

Many people are still at work when the sky performs (especially in winter), so sunsets rarely fit the calendar and are easily missed. SunSettings builds a doable micro-ritual into everyday life, reminding us that no two sunsets are the same. Community and personal sunset counters, plus gentle outdoor self-challenges, nudge participation and celebrate a new kind of luxury: time spent offline, outside, together.

What it solves / how it helps:

• Removes decision friction: a clear quality score, optimal time window, and nearby hotspots.
• Works with real schedules: timely nudges for narrow golden/blue-hour windows; quick check-ins you can actually make.
• Recreates community safely: opt-in check-ins, local Sunset Clubs, and light guidelines provide a friendly third place.
• Builds a healthy routine: streaks, mindful prompts, and private notes turn dusk into a low-effort daily ritual.

1. Forecast accuracy & trust
   Problem. Early prototypes overrated hazy evenings and underrated thin high-cloud events; users won’t forgive “go outside” when it’s mediocre.
   What we did. Switched to BrightSky/DWD for Europe with granular cloud layers (CLCT/CLCH/CLCM/CLCL) and humidity/precip features; OpenWeatherMap as a non-EU fallback. Built a prompt/scoring pipeline that ingests: cloud_total_pct, cloud_high_pct, cloud_mid_pct, low_cloud_pct, humidity_pct, precip_total_mm, precip_prob_max_pct and produces a quality score + confidence (conservative by design).
   Outcome (now). Fewer overconfident calls; the UI shows Go / Maybe / No-go with a confidence bar and a short “why” (e.g., “thin cirrus → vivid reds” vs “low stratus → flat light”). We want to be trusted.

2. Frictionless onboarding (hide the blockchain)
   Problem. Wallet setup killed adoption; we need ages 15–99 to sign up without crypto knowledge.
   What we want to do. We want to add a guest mode and email/passkey accounts; on-chain actions are batched by a relayer. Badges/NFTs are lazy-minted post-factum; users can export to a self-custodial wallet later.
   Outcome (now). Onboarding feels like any mainstream app; the blockchain is invisible until users are curious. Voting/rewards still settle on Base, but UX stays one-tap.

3. Photo integrity, NSFW, and “today vs. old”
   Problem. We must ensure contest posts are real sunsets from today, not spam or NSFW.
   What we did. Default in-app camera.
   What are we planning to do. If a gallery photo is older/missing EXIF, it’s accepted as Archive (not eligible for voting/prizes). Added a lightweight sunset classifier (sky/horizon/tonal pattern) + NSFW filter + perceptual-hash de-dupe.
   Outcome (now). Clean feed for pilots; suspicious posts are auto-queued for review without blocking normal users.

4. Privacy-first maps & anti-stalking community model
   Problem. We want great sunset spots on a map without turning them into stalking beacons.
   What we are planning to do.
   Three visibility levels:

• Private (journal only),
• Friends (mutual trust),
• Public (aggregated).

Public map uses jittered coords and k-anonymity (pins/heat only if ≥10 recent check-ins). Friendships are two-way trust; communities require a quorum (e.g., 5 members, each with ≥2 mutual trusts) before member locations become visible to that group. Emergency “hide past check-ins” switch.
Outcome (now). People can share safely; the map inspires exploration without exposing exact movements.

5. And we encountered several challenges during the development and deployment process:

Farcaster Integration Issues: Despite ensuring all fields in the manifest are correctly configured, Farcaster fails to index the app, preventing proper integration and visibility.

Base App Search Inconsistencies: The Base app experiences inconsistent search functionality, where some users can locate the app while others cannot, leading to accessibility issues.

Base Names Fetching Complications: Fetching Base names has proven problematic, as the names are not being retrieved successfully. Additionally, the process for fetching Base names is overly complex and lacks simplicity.

Location Restriction in Base App: The mobile Base app currently lacks access to the phone's location services, which limits its functionality compared to the web app, where location-based features work as expected.

https://github.com/alyonafuria/sunsettings_base

https://catch.sunsettings.app

SunSettings turns sunsets into a daily wellness habit, not a one-off photo opportunity. Our unique value is a habit loop (prediction → timely nudge → check-in → fair recognition) built for everyday people, plus a privacy-first community model that makes meeting outdoors safe and meaningful. Unlike photographer tools that only calculate, we combine:

1. a conservative, trust-oriented sunset quality score,
2. streaks, mindful prompts, and personal sunset counters,
3. on-chain, tamper-resistant voting and NFT badges for authentic participation, and 4) blockchain-hidden UX (guest accounts; export to self-custody later).

Our alpha validates this by shipping the full loop: prediction + map hotspots, one-tap check-ins, streaks/counter, and on-chain voting/rewards with a relayer so users don’t see crypto friction. We’re piloting with city “Sunset Clubs” to measure alert CTR, D7/D30 retention, UGC/DAU, and forecast trust (Go/Maybe/No-go accuracy). Early signals we target: a majority of users enabling alerts, stable streaks over multiple weeks, and clean feeds via EXIF/time geofencing and NSFW filters - proof that SunSettings creates a safe, sticky third place in the real world.

• After-work resetters (22–40, knowledge workers). Want a low-effort way to decompress. Value: clear Go/Maybe/No-go, narrow time window, streaks.

• Park & waterfront walkers (25–55, runners/dog owners). Already outside at dusk. Value: hotspot map, one-tap check-ins, gentle prompts.

• Couples & small friend groups (18–35). Micro-dates / shared moments. Value: “great sunset” alerts, shared counter, group challenges.

• Students & newcomers/expats (18–30). Low-cost socializing, new city discovery. Value: local Sunset Clubs, aggregated safe maps.

• Mindfulness & mental-health seekers (20–45). Prefer private reflection to public feeds. Value: mindful prompts, private journal, streak badges.

• Parents with kids (5–12). Short outdoor rituals that fit dinner/bedtime. Value: kid-friendly windows, nearby safe spots.

• Casual creators (non-pro IG/TikTok, 18–40). Want authentic content without clout games. Value: fair, on-chain voting; anti-spam/NSFW; exportable badges.

• Digital nomads & weekend travelers (20–40). Plan city breaks. Value: quality score across cities, lightweight check-ins, streak continuity.

• Older walkers (50–70). Gentle routine + nature. Value: large-type UI, simple alerts, aggregated (non-precise) maps.

Secondary customers: city parks & tourism boards, eco-travel orgs, universities, community centers. They get UGC-driven campaigns and safe, aggregated heatmaps of activity.

Sunsettings operates in a nascent market of sunset forecasting tools, primarily catering to professional photographers. Key competitors include:

• Sunset Hue
  https://sunsethue.com/
  A web-based platform offering robust sunset predictions via weather data integration. Lacks mobile accessibility and features an information-dense interface that compromises user-friendliness. No in-app photo capture, tamper-proofing, social elements, or blockchain/NFT integration.

• Alpenglow
  https://apps.apple.com/us/app/alpenglow-sunset-predictions/id978589174
  Delivers multi-day sunrise/sunset quality forecasts with notifications and widgets. Predictions exhibit variability in accuracy; UI is functional yet cluttered for photo navigation, with minimal social integration beyond hashtag sharing. No in-app photo capture or tamper-proofing; targets photographers for planning shots, without blockchain features.

• Freyr
  https://apps.apple.com/de/app/freyr-sunset/id1605500146?l=en-GB
  Emphasizes a clean interface for basic sunset timing and viability forecasts. Accuracy is moderate, and functionality remains rudimentary, devoid of advanced analytics or community features. No mention of in-app photo capture, tamper-proofing, social elements, or blockchain; appeals to general users for casual sunset viewing.

Ember
https://apps.apple.com/de/app/ember-vivid-sunset-prediction/id6502885454?l=en-GB
A student-developed tool using ML for 9-day vivid sunset predictions with probability visualizations. Allows photo submissions to improve models but lacks tamper-proofing, social/NFT integration, or robust UI beyond auto light/dark mode; targets tech-savvy users worldwide.

Helio
https://apps.apple.com/de/app/helio-sunset-forecast/id1512790275?l=en-GB
Provides user-friendly daily forecasts with quality ratings for sunrises/sunsets, with recent accuracy enhancements. No photo capture, tamper-proofing, social, or blockchain elements; focuses on photographers and nature enthusiasts seeking optimal lighting.

Market Gap: No competitor offers in-app photo capture for tamper-proof integrity, on-chain metadata/NFT minting, streak-based engagement, or token-gated social communities. Existing tools are prediction-centric for photographers, lacking Web3 incentives and casual user appeal. Sunsettings uniquely blends predictive AI, secure blockchain capture, and social dynamics to democratize sunset appreciation.