IdeaWell

Project ideas from Hacker News discussions.

Make your first edit to OpenStreetMap

Original Article

Hacker News Discussion

📝 Discussion Summary (Click to expand)

Theme 1: Beginners should avoid JOSM and use simpler editors
Many commenters warn that JOSM’s steep learning curve makes it a poor first tool, recommending the web‑based iD editor, StreetComplete, or Every Door instead.
- “As a beginner use the editor built in on the website instead (called ID). JOSM is a great power user tool, but this article's recommendation to use it for your first edit is misguided.” — VorpalWay
- “I would honestly not recommend beginners use JOSM - the web based iD editor [...] is much easier to learn and will hold your hand a lot more.” — 0l
- “JOSM is great when you get really into, but if it's really your first edit, just use the editor on openstreetmap.org.” — mikeocool

Theme 2: Clarity needed on what counts as a valid website vs. social‑media tags
The discussion shows confusion over the tutorial’s instruction to avoid social‑media profiles as a “website,” with users pointing out that separate contact: tags exist and that guidance should be clearer.
- “Determine if any of the search results represent the official website for your place. Do NOT use search results for social media profiles, review sites, or other business aggregators.” — ssl-3
- “For social media profiles, a different tag like contact:facebook, contact:instagram, etc. are recommended!” — juliantigler
- “There's separate tags for the social media profiles. So those can still be added.” — boredpudding*

Theme 3: Technical limits on data downloads create frustration for new editors
Users repeatedly hit the API’s 50 000‑node limit when trying to download large areas, receiving vague error messages with no hint about an appropriate box size.
- “The area you tried to downlaod is too big or your request was too large. And no indication whatsoever of how big the area should be, I'm at my fifth attempt so far.” — corecoder
- “The limit is imposed by the OSM API itself… a max of 50,000 nodes according the API docs. [...] Editing software could be generally aware of high and low density locales and advise accordingly.” — jmkb
- “Sorry you're having such a frustrating first experience. I'd recommend trying just 1-3 city blocks max.” — juliantigler

🚀 Project Ideas

OSM Bounding Box Advisor

Summary

  • A web tool that estimates the number of OSM nodes in a user‑defined bounding box before downloading, warns if the area exceeds the API limit, and automatically splits large requests into safe tiles.
  • Core value proposition: eliminates the frustrating “area too big” error by giving beginners clear guidance and automated chunking for bulk edits.

Details

Key Value
Target Audience New contributors using JOSM or desktop editors who hit the 50 k node limit
Core Feature Real‑time node count estimate via Overpass count, dynamic tiling suggestions, one‑click download of safe chunks
Tech Stack React + Leaflet frontend, Node.js/Express backend, Overpass API for counts, optional Python tiling service
Difficulty Medium
Monetization Hobby

Notes

  • Addresses corecoder’s pain: “The area you tried to download is too big … no indication whatsoever of how big the area should be.” (HN comment)
  • Could be packaged as a browser extension or a companion site linked from learnosm.org, encouraging discussion on optimal chunk sizes across urban vs. rural areas.

Beginner OSM Edit Wizard

Summary

  • An interactive, step‑by‑step overlay for the iD editor that walks first‑time users through a simple edit (e.g., adding a missing business) with contextual help and validation before submission.
  • Core value proposition: lowers the barrier to entry by providing a guided, task‑based experience directly in the official web editor.

Details

Key Value
Target Audience Absolute beginners who find iD overwhelming or prefer a “StreetComplete‑like” flow on desktop
Core Feature Wizard UI that highlights map features, prompts for required tags (name, contact:phone, opening_hours, etc.), checks for common mistakes, and submits the changeset
Tech Stack iD plugin built with JavaScript/React, Mapbox GL for highlighting, OSM OAuth for auth, optional localization via i18next
Difficulty Medium
Monetization Hobby

Notes

  • Responds to Juliantigler’s advice: “As a beginner use the editor built in on the website instead (called ID). JOSM is a great power user tool, but this article's recommendation to use it for your first edit is misguided.”
  • Provides a concrete way to implement the “quick mod using the web page” suggestion from nutjob2, likely sparking HN discussion about the best onboarding flow for OSM.

AI‑Powered Street‑Level Business Detector

Summary

  • A mobile/web app that ingests street‑level imagery (Mapillary, user photos) and uses OCR + LLMs to detect business names, phone numbers, opening hours, and social‑media contacts, then proposes ready‑to‑review OSM tags (name, contact:phone, opening_hours, contact:website, contact:facebook, etc.).
  • Core value proposition: automates the tedious manual entry of business details, letting validators focus on approval rather than data collection.

Details

Key Value
Target Audience Mobile mappers (Every Door, StreetComplete users) and remote contributors wanting to enrich business data
Core Feature Image upload → OCR/text extraction → LLM‑based entity recognition → auto‑filled OSM tag form → one‑click upload via Overpass
Tech Stack Frontend: React Native (or web with React); Backend: Python (FastAPI) with OpenCV/Tesseract OCR, HuggingFace LLMs (e.g., Llama‑3 or Mistral), Overpass/OSM API for writes
Difficulty High
- Monetization Revenue-ready: Freemium (free tier = 10 scans/day, paid plan = $9/mo for unlimited scans & priority processing)

Notes

  • Directly tackles KolmogorovComp’s request: “Is there any software that can take raw footage from 360deg cameras … identify businesses without websites, opening hours etc, and pre-fill them? With the user only having to validate.”
  • Leverages existing enthusiasm for AI‑assisted mapping (rapideditor.org) and would likely generate lively HN debate about accuracy, privacy, and the balance between automation and human validation in OSM.

Read Later