His initial comment describing it: https://news.ycombinator.com/item?id=28504740
GET /:city/weather.svg
|> fetch("https://geocoding-api.open-meteo.com/v1/search?name=" + .params.city + "&count=10&language=en&format=json", {method: "GET"})
|> jq: `{
latitude: .data.response.results[0].latitude | tostring,
longitude: .data.response.results[0].longitude | tostring
}`
|> fetch("https://api.open-meteo.com/v1/forecast?latitude=" + .latitude + "&longitude=" + .longitude + "&hourly=temperature_2m", {method: "GET"})
|> jq: `
.data.response.hourly as $h |
[$h.time, $h.temperature_2m] | transpose | map({time: .[0], temp: .[1]})
`
|> gg({ "type": "svg", "width": 800, "height": 400} ): `
aes(x: time, y: temp)
| line()
| point()
`
I think what they mean is that the data and code itself is libre/free/open [0] [1] but the API access is essentially rate limited for non-paying customers?
[0] https://open-meteo.com/en/licence (CC-BY)
[1] https://github.com/open-meteo/open-meteo/blob/main/LICENSE (AGPL)
This is not ideal for shared hosting services like cloudflare workers, but is the easiest and privacy-friendly way to limit access to fair-use.
Additionally, weather data is uploaded to a AWS S3 open-data sponsorship and you can run your own API instances (even commercially). The only draw back is, that a lot of data needs to transferred. I am working on a S3 cloud-native approach, but it is still in testing.
The free tier is cross-financed by commercial customers that use the service for energy forecasting, agriculture planing or wild fire prevention. There is no external funding, VCs, or whatsoever, the code is build in public on GitHub and I intent to continue running the free API service as is.
I think what they're trying to say is "Our service is completely open source. Our code is open source and our data is open source. We provide reasonable rate limits to our API access for non-paying customers. See our pricing plan if you'd like to become a commercial user and increase your rate limits".
Also, I looked into omfiles for a project recently, I was very impressed with their speed and blown away by the compression ratios!
It's the best weather data and presentation I ever had. The only solution that is clean and where long term predictions are worth something. Partially because of the data source, partially because of the UI that doesn't bullshit me.