n:point

npoint.io is a lightweight data store for your app or prototype.

Save FAQ answers, customer stories, configuration data, or anything else that will fit in a JSON blob. Then access your data directly via API.

Once your app is live, come back later to edit your saved JSON without having to redeploy. Or share a login with a friend so they can help you experiment! Features like schema validation and locking mean you can make these changes confidently, without breaking your app.

Contributing

Contributions are welcome!

Please open an issue to discuss proposed changes, rather than opening a pull request directly.

Development

Setup

It's an old project. To get things running nicely:

# Something like this
asdf use nodejs v25.2.1

# Might need this?
# Python 3.11 introduced an issue that node-gyp hits when building. Use older version
brew install python@3.10
export NODE_GYP_FORCE_PYTHON=/opt/homebrew/bin/python3.10

bundle

Note that I'm not using yarn now for local builds; can't seem to get it to reliably install.

Yarn is still used in deploys.

Running locally

rails s -p 3001
node scripts/start.js

Testing

Setup:

1. Install Chrome
2. Install chromedriver (brew install chromedriver on mac)

rspec
node jest  # no jest tests yet

Important note: Rspec integration tests run against the compiled version of the app in app/assets. Build with node scripts/build.js first, or set up capybara to run against your live webpack version (I haven't done this yet, but have ideas in spec_helper.rb).

Maintaining

Production build

# 1. Build files
node scripts/build.js

# 2. Make an "Add build files" commit
git commit -a -m "Add build files"

Deploying

Push to main. This deploys staging.

Deploy prod manually via render UI.

Bandwidth tracking (Admin)

To monitor which API documents are using the most bandwidth, first set credentials:

export ADMIN_USERNAME=your_username
export ADMIN_PASSWORD=your_secure_password

Then access the endpoint with HTTP Basic Auth:

# View top 20 documents by bandwidth (last 24 hours)
curl -u $ADMIN_USERNAME:$ADMIN_PASSWORD http://localhost:3001/admin/bandwidth

# Custom time window and limit
curl -u $ADMIN_USERNAME:$ADMIN_PASSWORD http://localhost:3001/admin/bandwidth?hours=12&limit=50

# Clear tracking data
curl -u $ADMIN_USERNAME:$ADMIN_PASSWORD -X POST http://localhost:3001/admin/bandwidth/clear

Tracking runs in-memory and automatically cleans up data older than 24 hours.

Similar Tools

• JSONbin.io
• Firebase
• AirTable
• JSON Schema Validator

Bookmarks

• JSON Schema
• JSON in Postgres

Self-hosting

Want to run your own instance of n:point? Go right ahead!

I'd recommend setting it up on render.com, which is what I use for npoint.io.

1. Set up a hosted Potsgres DB in render, and make sure DATABASE_URL points there
2. Configure environment variables. You'll at least need:
   • HOST (e.g. npoint.io)
   • CLOUDFLARE_API_TOKEN (optional, for cache purging)
   • CLOUDFLARE_ZONE_ID (optional, for cache purging)
3. Use these render settings:

# build command
./build.sh

# start command
./start.sh

The one piece of the app that won't work is password-reset emails, which go through a Sendgrid account. I recommend deleting the code for this and handling it yourself.

Codebase TODOs / Wishlist

• Add sentry or similar error collection service (search: TODO(sentry))
• Self-host and use privacy-respecting analytics (search: TODO(self-host))
• More testing (search: TODO(test))

Copyright & License

Copyright (c) 2017-2018 Alexander Zirbel - Code released under the MIT license.
n:point and the n:point logo are the property of Alexander Zirbel.