Hub & integrations: what can actually be connected
Part of the «Waze для рыбалки» wiki · Research stream · 2026-07-02 · Status: complete
Related: Latvia market · Roadmap
==== TOPIC: official-lv ====
SUMMARY: Latvia's official data surface splits cleanly in two. The environmental backbone (LVĢMC weather, water levels, water temperature, 14-day hydro forecasts, warnings, plus geo layers like priority fish waters and bathing-water quality) is genuinely open: CC0-licensed, machine-readable via the data.gov.lv CKAN API or WFS/WMS, updated continuously — a fishing hub can legally build on it today at zero cost. The fishing-specific layer (licensed waterbody registry, license sales, mandatory e-catch reporting) has consolidated into a de facto monopoly at state-owned LLKC's Mana Cope platform, which has no public API, no affiliate program, and only informal deep-link access; the old e-Loms service and its OData registry on epakalpojumi.lv are dead, and the data.gov.lv 'licensed fishing waterbodies' dataset has been delisted. BIOR fish-stocking and stock-assessment data exists only as HTML news tables and PDF reports; violation reporting (VVD/Vides SOS) has no documented API. So the hub thesis is technically viable for the weather/hydro/map layers, but the license-purchase and catch-report loop requires one negotiated partnership with LLKC, and the gaps (real-time ice, catch data, stocking feeds) are exactly where crowdsourcing adds unique value.
INTEGRABLE:
- LVĢMC Hidrometeoroloģiskie novērojumi (data.gov.lv) — CKAN datastore API + CSV, CC0-1.0, continuous updates: meteo + hydro operational data (last 48h) and 365-day archive; hydro parameters confirmed in reference file: LIMEN water level (cm, hourly), WTEMD water temperature (°C, hourly), SEHL/SEDUT sea level/temp; station metadata CSVs included; SQL filtering via data.gov.lv/dati/api/3/action/datastore_search_sql. Source: https://data.gov.lv/dati/dataset/hidrometeorologiskie-noverojumi
- LVĢMC Hidroloģiskās prognozes (data.gov.lv) — CSV via CKAN, CC0-1.0: 14-day daily forecasts of water level (LAS-2000.5) and water temperature per monitoring station, with ensemble median and 25–75%/5–95% probability bands. Directly usable for a 'bite forecast' feature. Source: https://data.gov.lv/dati/api/3/action/package_show?id=hidrologiskas-prognozes (resource hidro_forecast.csv)
- LVĢMC Meteorological forecasts for populated places + spatial forecasts + marine forecasts + hydrometeorological warnings (data.gov.lv) — CC0, CKAN datastore: 24h hourly + 7-day point forecasts, 9-day spatial forecasts updated ≥4x/day, marine forecasts 2x/day, warnings feed. Source: https://data.gov.lv/dati/organization/lvgmc
- Prioritārie zivju ūdeņi (priority salmonid/cyprinid waters) via ĢEOLatvija on data.gov.lv — Shapefile + WFS + WMS + Atom, CC0-1.0, per Cabinet Reg. No. 118; ready-made map layer of legally designated fish waters. Source: https://data.gov.lv/dati (slugs prioritrie-zivju-dei, -wfs, -wms)
- Peldvietu ūdens kvalitāte un monitoringa stacijas (Health Inspectorate via ĢEOLatvija) — WFS/WMS/Atom, CC0, continuous updates during season: bathing-water monitoring stations, operational and long-term quality assessments (proxy for water condition at popular shores). Source: data.gov.lv slug peldvietu-dens-kvalitte-un-monitoringa-stacijas-dati
- WFD Bioloģisko elementu kvalitātes novērtējums (incl. fish quality element, 2006–2021) — WFS/SHP, CC0; plus Metāli Baltijas jūras un Rīgas līča zivīs (contaminants in perch/herring) — CSV/JSON, CC0. Ecological context layers. Source: data.gov.lv slugs bioloisko-elementu-kvalittes-novrtjums-virszemes-dei1, metali-baltijas-juras-un-rigas-lica-zivis
- Zivju pirmo pircēju saraksti (Ministry of Agriculture) — CSV, CC0 on data.gov.lv; marginal but shows ZM does publish some fisheries registers. Source: https://data.gov.lv/dati/dataset/zivju-pirmo-pirceju-saraksti
- Mana Cope (manacope.lv, operator SIA LLKC reg. 40003347699, state-owned) — DEEP-LINK path only: stable public URLs per waterbody license product (manacope.lv/waterbody/licence/{id}/view, /licenses/{id}/view, /products lists ~80 licensed waterbodies) and makskeresanaskarte.lv/product/{id}/buy for the state fishing card; a hub can route users to purchase without any agreement. Formal partnership route exists in principle: single counterparty LLKC Fisheries Division (info@manacope.lv, info@makskeresanaskarte.lv); precedent for third-party card resale exists (Viada fuel stations sell 3-month/annual makšķerēšanas karte physically). Sources: https://www.manacope.lv/products, https://www.makskeresanaskarte.lv/, https://www.viada.lv/degvielasuzpildesstacijas/produkti-un-pakalpojumi/makskernieka-karte/
- BIOR fish migration barriers database — downloadable database (updated Feb 2026) of dams/obstacles on Latvian rivers; usable as a map layer with attribution. Stocking data (species, counts, waterbody, dates — e.g. 93,134 salmon smolts to Daugava Apr–May 2025) is published but only as HTML news tables on zm.gov.lv and annual PDF reports — integrable by scraping/manual ETL under public-sector information reuse rules, not by API. Sources: https://bior.lv/lv/par-mums/jaunumi/izstradata-zivju-migracijas-iespeju-atjaunosanas-datu-baze, https://www.zm.gov.lv/lv/bior-zivju-audzetavas-izaudzeto-zivju-mazulu-ielaisana-dabiskajas-udenstilpnes
- VVD / Vides SOS — integrable only as OUTBOUND deep link: hub can link users to the Vides SOS mobile app (iOS/Android) or vvd.gov.lv 'Ziņo par vides pārkāpumiem' channels for poaching reports; public report map exists at videssos.lv/karte (JS app, no documented data export). Partnership timing is favorable: VVD has an active EU-funded project to modernize Vides SOS functionality (2026 app update shipped). Sources: https://www.videssos.lv/, https://www.vvd.gov.lv/lv/projekts/mobilas-lietotnes-vides-sos-funkcionalitates-izvertesana-un-pilnveidosana, https://www.vvd.gov.lv/lv/zino-par-vides-parkapumiem
BLOCKED:
- Mana Cope / makskeresanaskarte.lv public API — does not exist; no developer docs, no affiliate/embed program anywhere on either site; license purchase and catch-report submission happen only inside their web/app; scraping their waterbody/product pages is technically trivial but has no stated reuse license (state-owned operator, ToS page not publicly resolvable) — needs LLKC agreement
- e-Loms on epakalpojumi.lv — DEAD: /Services/LOMS/ and per-lake pages (e.g. Lake.aspx?id=45702 Usmas ezers) now redirect to a generic municipal-services portal with no fishing content; function migrated to manacope.lv (organizer admin environment was visvaris.lv); verified live 2026-07
- epakalpojumi.lv OData endpoint /odata/Home/BodyOfWaters ('Licencētās makšķerēšanas ūdenstilpes') — HTTP 404; the once-open machine-readable registry of licensed fishing waterbodies is gone
- data.gov.lv dataset 'licencetas-makskeresanas-udenstilpes' — DELISTED: dataset URL and CKAN package_show both 404, portal search for 'makšķerēšana' returns 0 results; Latvia currently has NO official open-data registry of licensed fishing waters — the only current list is Mana Cope's own product catalog and individual municipal binding regulations on likumi.lv
- Vides SOS API — no public API, no export, no embed documented; report data visible only through their own JS map; integration beyond deep-linking requires a VVD agreement
- BIOR structured data feeds — BIOR has no organization/datasets on data.gov.lv; stock assessments and stocking plans exist only as PDFs (e.g. annual 'Atskaite par zivju resursu izpēti' on zm.gov.lv) and news pages; EU data-collection outputs go to the European Commission portal, not a national API
- Official real-time ice data — LVĢMC observes ice phenomena at ~65 stations but the open-data hydro feed contains NO ice parameter (confirmed from hidro_parametri.csv: only LIMEN/WTEMD/SEHL/SEDUT); ice thickness reaches the public only as winter news bulletins/PDF — no machine-readable source exists
IMPLICATIONS:
- The 'hub' claim is immediately defensible for environmental context: CC0 + real APIs mean the app can legally fuse water level, water temp, 14-day hydro forecasts, weather, warnings, and fish-water map layers with crowdsourced reports on day one, no negotiation, no fees, no attribution burden (CC0).
- The license-purchase loop runs through exactly one counterparty: state-owned LLKC controls both the state fishing card (makskeresanaskarte.lv) and licensed-water sales + mandatory e-catch reporting (manacope.lv). Plan A: undocumented but stable deep links (works today, zero permission). Plan B: a single partnership negotiation with LLKC unlocks referral/embed status — the Viada retail precedent shows they do use distribution partners. This is also the strategy's biggest single point of failure.
- Latvia has no official machine-readable registry of licensed fishing waterbodies anymore (the open dataset and OData feed are both dead). Building and maintaining that registry (from Mana Cope's catalog + likumi.lv municipal regulations) is itself a hub-worthy asset the state currently doesn't offer.
- Catch data flows one way into Mana Cope and never comes back out as open data — so the app's crowdsourced catch/bite layer would be a dataset nobody else has, including BIOR. That's leverage: offer aggregated anonymized catch/ice data to BIOR/ZM/LVĢMC as the carrot in partnership talks.
- Crowdsourced ice reports fill a verified hard gap: the official open-data feed has zero ice parameters, and LVĢMC's 65-station ice observations reach the public only as prose bulletins. Same for 'recently stocked' alerts — stocking info is HTML/PDF only, so even a scraped weekly ETL beats the status quo.
- Violation reporting should be out-linked, not rebuilt: route users to Vides SOS / VVD channels now, and pursue a VVD integration while their Vides SOS modernization project is active — a 'report poacher' button feeding VVD is a cheap credibility win with the regulator.
- Legal posture: everything on data.gov.lv/ĢEOLatvija is CC0 or CC-BY (reuse-safe, GDPR-irrelevant); the risky zone is scraping manacope.lv and videssos.lv (no reuse license). Keep scraping limited to factual catalog data (prices, waterbody lists) pending agreements; facts aren't copyrightable but database rights could be asserted by a state-owned operator you also want as a partner.
==== TOPIC: community-legal ====
SUMMARY: The Latvian community fishing surface splits cleanly: two of four legacy sites are alive (copeslietas.lv — the dominant LV-language portal; ribak.lv — an active Russian-language phpBB forum with a working public Atom feed), while mansloms.lv (dead since ~2018) and fishing.lv (dormant personal site) are gone. Facebook — where the largest actual communities live (e.g. groups/fishinglv) — is fully closed: Meta removed the Groups API in April 2024 and logged-in scraping is an enforceable ToS breach (Meta v BrandTotal) with GDPR exposure. YouTube (official Data API + embeds) and Telegram (MTProto/TDLib reads of public channels) are technically open, though Latvian fishing Telegram content is thin. Legally, an EU/Latvian aggregator is on solid ground with a link + very-short-snippet + attribution model (hyperlinks and very short extracts are excluded from DSM art. 15, which doesn't cover forums/UGC anyway), but full ingestion of a Latvian site's forum database risks the sui generis database right — the controlling CJEU precedent (CV-Online Latvia v Melons, C-762/19) is literally a Latvian aggregator case.
INTEGRABLE:
- ribak.lv ('Латвийский форум любителей рыбалки') — ALIVE July 2026 (last post 2026-07-02; 346,019 messages, 7,797 users). Mechanism: public forum-wide Atom feed at https://www.ribak.lv/app.php/feed (phpBB also supports per-forum/per-topic feeds). Data: latest forum posts incl. regulations Q&A, catch chatter (Russian). No API, but no scraping-hostile ToS observed (standard phpBB rules/privacy links). Small hobby forum — direct partnership realistic. Source: https://www.ribak.lv/
- copeslietas.lv (NEWS + partnership only) — ALIVE 2026 and the dominant Latvian-language portal: 2026 competition calendar (https://www.copeslietas.lv/site/sacensibas/kalendars.htm?year=2026), forum active (topics with March 2026 posts), video section that itself aggregates YouTube fishing channels (https://www.copeslietas.lv/site/video/kanali/). Mechanism: official RSS feed for its news section (confirmed in its own materials) + partnership/deep-linking; ToS at https://www.copeslietas.lv/lv/noteikumi/ (binding on all users, grounded in Latvian Copyright/Data Protection law). Note: server returns HTTP 403 to non-browser clients, so integration must go through the offered RSS or a deal, not scraping. That it aggregates YouTube itself sets a cooperation precedent in the niche.
- Telegram public channels — Mechanism: MTProto client API / TDLib reads public channel history; Bot API works if a bot is added as admin; t.me/s/
- YouTube Data API v3 — both target channels confirmed live: Cope Latvijā TV (https://www.youtube.com/@Cope_Latvija_TV, Latvian, carp/all-species) and FISHING LV – Рыбалка с Виталием (https://www.youtube.com/channel/UC4olgZHBqgNlqWAxyp51yLg, Russian). Mechanism: official API (search, channel/video metadata, playlists) + iframe embeds for playback. Compliance rails: non-authorized data storable max 30 days then refresh/delete, must display current data, mandatory YouTube branding/attribution next to content (https://developers.google.com/youtube/terms/developer-policies), default 10k units/day quota. Lowest-risk content surface — the API exists precisely for aggregators.
- EU-legal 'link + snippet + attribution' layer (applies to ALL living sources) — hyperlinking is expressly outside DSM Directive art. 15 and outside communication-to-public (CJEU Svensson); 'individual words or very short extracts' are excluded from art. 15; and art. 15 only covers PRESS publications — forums, blogs, UGC communities like copeslietas/ribak are outside its scope entirely (https://legalblogs.wolterskluwer.com/copyright-blog/taking-freedom-of-information-seriously-the-very-short-extracts-limitation-in-article-15-cdsm-directive-and-how-not-to-implement-it-part-1/, https://communia-association.org/2024/02/19/the-post-dsm-copyright-report-the-press-publishers-right/). TDM exceptions (DSM arts. 3–4, transposed into Latvia's Autortiesību likums) allow mining lawfully accessible content for analytics unless the site posts a machine-readable opt-out.
BLOCKED:
- Facebook Groups — NO API path: Meta deprecated the entire Groups API in Graph v19.0 and removed it from all versions on 2024-04-22, including groups_access_member_info and publish_to_groups; in 2026 nothing remains for reading (or posting to) group content — native tools only (https://developers.facebook.com/docs/graph-api/changelog/version19.0/, https://www.sprinklr.com/help/articles/getting-started/meta-deprecates-facebook-groups-api/66229eb25f9dd9599d632712). Scraping is legally hostile: groups sit behind login, and Meta v BrandTotal (2022) held logged-in automated collection an enforceable ToS breach; Meta's Automated Data Collection Terms require separate written permission; GDPR adds exposure for group members' personal data (Meta's own €265M fine over scraped user data shows regulator sensitivity). Meta v Bright Data (2024) only shields LOGGED-OFF scraping of public pages — inapplicable to groups. This blocks the single biggest LV community surface, e.g. 'Fishing LV 🎏 Рыбалка в Латвии' (https://www.facebook.com/groups/fishinglv/).
- mansloms.lv — DEAD: last gallery/forum content 2017–2018, copyright notice frozen at 2017; custom platform, no RSS/API. Nothing to integrate.
- fishing.lv — DORMANT: now a personal info site (Jānis Stikuts, 'Makšķerēšana Latvijā'), no fresh content, no RSS/API/ToS. Not a community surface anymore.
- copeslietas.lv FORUM full ingestion — blocked on three layers: (1) infrastructure bot-blocking (HTTP 403 to non-browser agents on every page incl. ToS); (2) ToS binding all users, anchored in Latvian Copyright Law; (3) EU Database Directive 96/9 sui generis right — CJEU C-762/19 CV-Online Latvia v Melons (a Latvian case) holds that an aggregator's extraction/reindexing infringes where it risks depriving the database maker of income redeeming its investment (https://www.twobirds.com/en/insights/2021/uk/cv-online-latvia-cjeu-complicates-the-enforcement-of-database-rights). News-RSS use and deep links stay fine; wholesale forum mirroring does not.
IMPLICATIONS:
- Architect the hub as a router, not a mirror: deep links + very short snippets + source attribution is legally safe in Latvia/EU (art. 15 doesn't cover forums; hyperlinks and very short extracts are excluded even for press), while full ingestion triggers the sui generis database right under the exact precedent a Latvian rightsholder would cite (CV-Online v Melons). The CJEU's own test — does the aggregator threaten the source's income? — is satisfied in your favor if the hub demonstrably sends traffic TO the sources.
- Facebook cannot be a data source, only a distribution channel. The largest live communities (FB groups) are API-dead and scraping-toxic, which is strategically GOOD for a new hub: the fragmentation has no incumbent aggregator and cannot get one via FB. Play it in reverse — shareable catch-report cards from your app into FB groups as the acquisition loop.
- None of the integrable sources carry structured real-time bite/ice data — ribak/copeslietas are conversational, YouTube is lagged video, Telegram is thin. Aggregation buys you CONTEXT and content gravity (news, competitions, video, forum chatter), but the core real-time layer must be your own crowdsourced data. That means the moat is the crowd, and the aggregated shell is the cold-start content that makes the app worth opening before the crowd exists.
- Prioritize by friction: YouTube API (official, designed for this — just respect 30-day storage cap and branding) > ribak.lv Atom feed (works today, zero negotiation) > copeslietas.lv partnership (bot-blocked but offers news RSS and itself aggregates YouTube, so aggregation norms are accepted; it's the must-have LV-language partner) > Telegram (easy API, little Latvian content yet — cheap to monitor, could seed your own channel).
- Two ML-specific legal rails: if bite-prediction models train on aggregated text, DSM art. 4 TDM exception covers lawfully accessed content UNLESS a machine-readable opt-out exists (check robots.txt/ToS per source, and note copeslietas' 403 arguably signals reservation); Telegram's API ToS flatly bans AI/ML training on its data — keep Telegram content out of training corpora regardless.
- A partnership-first posture converts the two legal risk points (copeslietas ToS+database right, small-site goodwill) into distribution deals: these are hobby-run operations with no API budget, and 'we send you traffic + embed your competitions calendar/news with attribution' is a cheap yes — likely cheaper than any scraping workaround and it forecloses the CV-Online-style claim entirely.
==== TOPIC: commercial-apis ====
SUMMARY: The commercial fishing-app world splits cleanly: weather/solunar data is a cheap commodity (free-to-€990/yr APIs), bathymetry is gatekept but reachable via Garmin/Navionics' application-based API and user-owned Deeper CSV exports, and the social incumbents (Fishbrain, FISHSURFING) are walled gardens you cannot legally aggregate. Latvian commercial ponds all run on phone/email booking with zero online reservation infrastructure — a genuine whitespace for a hub. Successful precedents (onX Fish, Angler's Atlas/MyCatch) aggregated public/government data and partnered with agencies; nobody has succeeded by aggregating other private apps.
INTEGRABLE:
- Windy Point Forecast API — REST API, €990/yr Professional (10,000 req/day, 100+ variables, models incl. AROME/ICON-EU/GFS; ECMWF excluded for licensing); free trial returns deliberately shuffled data, production forbidden. Separate Map Forecast API (Leaflet plugin) for embedding animated weather maps. https://api.windy.com/point-forecast/pricing
- OpenWeather — One Call 3.0/4.0: 1,000 calls/day free then pay-per-call (card on file); classic free tier 1M calls/month, 60/min for current+3h forecast. Cheapest brand-name option for a small app. https://openweathermap.org/price
- Open-Meteo — free non-commercial tier 10,000 calls/day (600/min), no key; commercial plans (1M–50M calls/mo, dedicated endpoint) at contact pricing; CC BY 4.0 attribution required. Best cost profile for MVP. https://open-meteo.com/en/pricing
- Meteoblue — Free Weather API 5,000 calls/yr; 1-year free trial with 10M credits; paid packages from ~€2,400/yr (40k–500k calls/mo). Works but priced above alternatives for a small app. https://business.meteoblue.com/pricing
- Windguru — Station JSON API for live station measurements (documented, free: stations.windguru.cz/json_api_stations.html) + Upload API for adding your own stations + free embeddable forecast widgets (help.php?sec=distr). No official forecast REST API — forecast integration = widget embed, not data. https://stations.windguru.cz/
- Solunar/bite-forecast — api.solunar.org: free, no auth, GET by lat/lon/date (majors/minors/moon). Stormglass API (marine weather+tides+astronomy): free 50 req/day, paid tiers above. Solunar math is also trivial to compute in-house — zero-cost feature. https://solunar.org/ (api.solunar.org)
- Garmin/Navionics Marine Charts — Mobile SDK (iOS/Android) + Web API with daily-updated depth data, HD contours, spot soundings; explicitly offered to 'app developer, fishing forum' audiences; access via developer request form, pricing/terms not public (negotiated). The only legitimate licensed-bathymetry path. https://developer.garmin.com/marine-charts/ + https://www.garmin.com/en-US/forms/navionics-web-api/
- Deeper / Fish Deeper (user-data path) — no public API/SDK, but users can export their OWN bathymetry scans as CSV (lat, lon, depth, timestamp, water temp) and raw sonar CSV, plus NMEA0183 over UDP from the device. Mechanism: user-initiated import of their exports into your app = legally clean crowdsourced depth layer. Deeper is Lithuanian (Vilnius) — realistic Baltic bizdev target via contact form (research/collab invited; formal partnership page is retail-only). https://support.deeper.eu/668053 + https://deepersonar.com/en-us/pages/research-and-surveying-with-deeper-sonar
- Latvian commercial ponds (copes dīķi) — 9+ with own websites and public price lists: kalnaspulles.lv, albertadiki.lv, sillakas.lv, pertnieki.lv, udensrozes.lv, plavnas.lv, vipedis.lv, dikis.lv (Draudziņu), bergadiki.lv; directory content exists at lielaisloms.eu and fishing.lv. ALL booking is phone/email — no online reservation system found anywhere. Mechanism: directory listing + direct partnership; offering them a booking widget is an open wedge. https://www.lielaisloms.eu/saraksts-interesantakie-copes-diki-latvija
- FISHSURFING (precedent, partial) — Czech aggregator: 35,000+ waters, 4,000+ verified private fisheries in Europe, business profiles for fisheries/shops/guides, freemium PRO tier; ~500k users, backed/acquired by RSBC Group (Prague). Proves the private-fishery-directory model is buildable via business-profile onboarding (not APIs), with modest scale after ~8 years. https://www.fishsurfing.com/en/ + https://www.rsbcgroup.com/en/investment-portfolio/fishsurfing-app-en
- Angler's Atlas / MyCatch (precedent) — citizen-science hub since 2018: angler catch logs shared with government/university researchers under data-sharing agreements (creel-survey-compatible data, point data never public), agency-sponsored tournaments as engagement engine. Proves B2G data-broker positioning works for a small company. https://www.anglersatlas.com/ + https://www.ijc.org/en/mycatch-lets-anglers-help-scientists-gather-fish-data
- onX Fish (precedent) — launched free beta 2024 (Minnesota), paid $34.99/yr from 2025, now ~10 Midwest states; hub value built by aggregating PUBLIC data (state lake surveys, regulations, stocking, satellite imagery) plus tourism-board partnership (Visit Bemidji). Proves public-data aggregation + local depth beats social-network breadth. https://www.onxmaps.com/blog/onx-officially-introduces-onx-fish
BLOCKED:
- Fishbrain — no public API or embed exists (GitHub is internal tooling; maps are licensed FROM Mapbox); ToS explicitly bans scraping/automated extraction AND 'third-party applications that interact with the Services' without written consent; data flows only via negotiated research agreements (e.g., Univ. of Texas biodiversity project). Cannot be aggregated. https://fishbrain.com/terms-of-service
- Deeper cloud bathymetry (fishdeeper.com global maps / Lakebook) — the aggregated community depth-map layer has no API and no licensing program; only each user's own scans are exportable. The partnership page covers retail resellers only. https://deepersonar.com/en-us/pages/partnership
- Windguru forecast data — no official forecast API; existing integrations are scrapers/unofficial wrappers (GitHub, Apify), ToS-gray and fragile. Only station data and widgets are sanctioned. https://stations.windguru.cz/json_api_stations.html
- Windy ECMWF model — excluded from the API entirely at any price due to ECMWF licensing; and the free trial tier is legally unusable in production (deliberately corrupted data). https://api.windy.com/point-forecast/pricing
- Navionics/Garmin self-serve — no published pricing or terms; access is gated behind an application form and negotiation, so it cannot be assumed in a launch plan until Garmin approves (timeline/cost unknown). https://www.garmin.com/en-US/forms/navionics-web-api/
IMPLICATIONS:
- The conditions layer (weather, wind, pressure, solunar/bite windows) is a commodity costing €0–990/yr — build it on Open-Meteo + api.solunar.org for free at MVP, upgrade to Windy/OpenWeather paid tiers later. It is table stakes, not a moat.
- The only defensible data moat available is depth/bathymetry via two channels: apply for Navionics Web API (fishing apps are their stated audience) as licensed baseline, and build a Deeper-CSV upload pipeline so Latvian anglers crowdsource their own scans into the hub — legally clean because users own their exports, and Deeper being Lithuanian makes a formal Baltic partnership pitch plausible.
- You cannot be a hub OF the incumbent apps: Fishbrain and FISHSURFING are legally walled (ToS bans, no APIs). The viable hub position is aggregating public/state data + owned crowdsourced data + partner content — exactly the onX Fish and MyCatch playbook — and out-localizing global apps rather than integrating them.
- Latvian copes dīķi are a genuine greenfield: every pond found books by phone/email with public price lists and no online reservation anywhere. A directory + booking widget offered free to pond owners is a low-cost partnership wedge that no global player (Fishbrain, FISHSURFING) will bother building for a 1.9M-person market.
- Precedent check on 'aggregator hub' viability: FISHSURFING (pure directory/social aggregator) reached only ~500k users across all of Europe in 8 years; onX Fish and MyCatch succeeded faster by anchoring on exclusive public-data integration and government/tourism partnerships. For Latvia this argues the hub should lead with state data (licences, regulations, stocking) and B2G citizen-science credibility, with commercial integrations as garnish.