A lot is possible with a zipfile of data and just the programs that are either already installed or a quick
apt installaway. I remember how impressed I was when I was first shown this
eurofxref-hist.zipby an old hand from foreign exchange when I worked in a bank. It was so simple: the simplest cross-organisation data interchange protocol I had then seen (and probably since).
A mere zipfile with a csv in it seems so diminutive, but in fact an enormous mass of financial applications use this particular zipfile every day. I’m pretty sure that’s why they’ve left those commas in – if they removed them now they’d break a lot of code.
When open data is made really easily available, it also functions double duty as an open API. After all, for the largeish fraction of APIs in which are less about calling remote functions than about exchanging data, what is the functional difference?
I wonder how many of these types of simple, but extremely powerful open datasets that are so relatively easy to use exist.