Hello,
I am developing psycopg3 <https://www.psycopg.org/psycopg3/>, a major
release of the psycopg2 PostgreSQL driver for Python. This release
finally uses the advanced query protocol and server-side parameters
binding, so adaptation of Python values to PostgreSQL data types has,
at times, a different behaviour compared to the past. 100%
compatibility is not a target: that's the whole point of creating a
new package, but I am studying what is the sanest behaviour adaptation
can have.
I am exploring the best trade-offs in the choice of the mapping across
types: I have a write-up at
<https://www.varrazzo.com/blog/2020/11/07/psycopg3-adaptation/> and
would like to get input from PostgreSQL people regards whether the
choices we are making are the best possible ones.
Looking forward to get your feedback, thank you very much!
-- Daniele