3.4 KiB
Curls
(Notice in this demo how each color has their own set of curls)
(Also notice the $ picker to switch between curls you own)
What is this?
This is a text status host.
Users claim a curl and get a key to be able to update it.
For instance https://curls.merkoba.com/funone would return the current text of the funone curl.
The response is always pure text, there is no json or anything.
There is no history, or any other features except hosting the current text.
To view and change a curl there's some html pages available but you are encouraged to use your own tools to view and change.
Change one:
curl -X POST -d "curl=mycurl&key=mykey&text=mytext" https://curls.merkoba.com/change
If the change was successful it will return "ok".
Get current status of one:
curl https://curls.merkoba.com/[curl]
Get data of one or more (json response):
curl -X POST -d "curl=somecurl&curl=othercurl" https://curls.merkoba.com/curls
Intended Use
How I think this can be used is by users sharing their curls to others and then they add them to a program. The program would update every added curl every 5 minutes to check for changes. So a user gets an overview of the status of many people.
The status of the people you choose can serve as calls of action or to know about interesting activities happening in real time.
Yes it's similar to Twitter except there's no history or other kind of privacy invading mechanisms in place. And there's no CORS restrictions, reading and changing the curl status should be as easy as possible, from within any application.
It can also serve for software to point to pieces of dynamic text.
Accounts
There are no accounts. You claim a curl once, you are given the key once.
There are no mechanisms to recover a lost key.
Dashboard
I have my own implementation of a program that uses curls.
It is located at: https://curls.merkoba.com/dashboard
The html template is found here.
The code is found here.
Colors
Each color has their own set of curls. You can use them as profiles.
Picker
When you make a successful change to a curl, the curl is saved to your picker list.
So you can easily switch between them by clicking the $ button.
Installing
Everything happens inside /server
Create a virtual env and install dependencies using requirements.txt
Run init_db.sh to create the sqlite3 database.
Create captcha_key.txt and fill it with a secret key for the captcha.
Development
To run a dev server use venv/bin/python -m flask run --debug
To debug quickly create captcha_cheat.txt and fill it to use as the captcha.
Production
Use gunicorn to run it in production.
Set up a reverse proxy on apache.
Make a systemd service to start gunicorn with a non-root user.
Bundle
The js files in static/dashboard/js get bundled into bundle.js when the server starts.
There is a bundle.py file that can be called directly to force a bundle.
Rate Limiter
There is a rate limiter to limit all API endpoints to x requests per minute.
This depends on redis.
Schema
You can see the current schema here.
It's supposed to remain very simple.