Michael J. Sullivan
Michael J. Sullivan
Some of this is probably because `CREATE DATA BRANCH` is serialized; we could consider loosening that restriction a bit (based on size of conn pool?) or bringing back the old...
Currently connections die in silence even with debug printing turned on
See https://github.com/pganalyze/libpg_query The branch we are using is unmaintained now and getting hard to build, and we want up-to-date features anyway
The trickiness here is in supporting all of the various special edgeqlisms that affect writes: * Rewrites * Triggers * Access policies? (We'd probably want that to work on reads...
The tests take real long.
``` import functools class Foo: def f(self, x: int) -> int: return x class Bar(Foo): @functools.lru_cache() def f(self, x: int) -> int: return x + 1 ``` crashes with ```...
Currently, we will inject a LIMIT at any exposed SELECT, which results in us applying implicit LIMIT before OFFSET, which is wrong. Drop the injection inside of SelectStmt compilation and...
If constraints that have an inherited subjectexpr have a new base type added to the type they are on, we were losing 'subjectexpr' as an inherited field, because a code...
Currently we generate an "inheritance view" for every object type and multi pointer in the schema, named the same as the object but suffixed with "_t" and selecting from the...