Distribute the Luftdaten and Madavi databases on a content network

[LuchtwachtersDelft]

I don't know about Luftdaten, but Madavi is hosted on a single private server. When Madavi goes down the measurements aren't logged and the graphs and archive of past measurements can't be accessed.

Can we increase the availability and reliability of the project by distributing the database, for example on a CDN or distributed peer-to-peer sync network?

And can we set up failover servers for the main APIs that are enabled by default, Luftdaten and Madavi?

[ricki-z]

Before we can set up failover servers for Luftdaten we have to separate personal informations from sensor data. And if you read the GDPR you will see all the things we need to do, if we save data on systems not "owned" by us.

[LuchtwachtersDelft]

I specifically brought up the case of Luftdaten during a meeting with an organization whose goal is to both legally and technically address the issues with data sharing and privacy, to give individuals more control over the data they share through a platform called a DataCommons. We could figure out a solution together.

The EU claims GDPR in fact supports the concept of Open Data.

https://www.europeandataportal.eu/en/highlights/protecting-data-and-opening-data

In practice it's more complex.

Location data is an example of personal data protected under GDPR. Even without GDPR I understand the desire of air quality data providers to protect their identity. However, location data is necessary to render an air quality map. Several platforms attempt to anonymize that data by adding a random offset, however it's not fully effective because the resulting location can still end up on the original location https://github.com/opendata-stuttgart/meta/issues/83

What other (potentially) personal data is being gathered now? Can we scrub the unnecessary data and anonymize the data we need?