3 ms·
Depends on the program. In some cases you understand what it is to be done and it is about building and not discovery. Sometimes the discovery is "what are th
by PaulHoule 2y ago
Depends on the program. In some cases you understand what it is to be done and it is about building and not discovery. Sometimes the discovery is "what are the requirements?" Some programs require a lot of "discovering" the truth about APIs you use. Some require discovering about how machine learning algorithms perform.
I have a few side projects that depend on arangodb, which like most databases from the 2010s, has a terrible license -- so I am getting these ported to run with Postgresql using JSONB.
I've been building an adaption layer that looks a lot like
https://docs.python-arango.com/en/main/ https://docs.python-arango.com/en/main/
but it's turning out to be an interesting act of triangulation between
(1) The parts of the arangodb API I really use
(2) The requirements of my applications
(3) Affordances provided by SQLAlchemy and Postgres, and
(4) The database API I really want
So there has been a lot of discovery there.