When I see the term "Open Data" I instantly think of open data portals - mostly run by governments around the world. These things have never been healthier: ten years ago they hardly existed, today you can get civic data from local governments all over the place (last time I saw an attempt to count there were over 4,000 of these portals, and that was a few years ago).

My favourite example is still this CSV of all 190,000+ trees in San Francisco, which is updated most business days with details of the latest tree changes: https://data.sfgov.org/City-Infrastructure/Street-Tree-List/... - I track changes to it here: https://github.com/simonw/sf-tree-history/

This article is about something different: it's about what I guess you could call the "Open APIs" movement. Back in the days of Web 2.0 every service was launching an open API, hoping to harness developer attention to help make the platforms more sticky. Facebook and Twitter both did incredibly well out of this strategy, at least at first.

THOSE APIs are mostly on the way out now. Companies realized that giving away their data for free has a lot of disadvantages.

Open Data is doing great. Open APIs are not.

https://data.gov/

Wikidata and many other things "open". It's my own (partially true) opinion that open data is downloadable as a data dump, anything involving an API is like breathing through a straw.

...

Ah, it's a veiled advert for "Sort".