Brent Moran
Brent Moran
## Problem Currently, we have some tests in `db/tests/types/operations/test_cast.py` that run quite slowly, and are parameterized to run many times. We should split them up. For example, the test `alter_column_casts_data_gen`...
## Problem The casting functions get run a _lot_ (especially when doing type inference). We need them to be efficient. ## Proposed solution One potential option is to use `RETURNS...
## Problem For SQLAlchemy columns, `server_default` can only be a string (or a default clause). This implies needing to know the actual DB representation of values for columns where we...
## Problem As we near the end of the initial round of implementing database types, we should reassess our current setup and look for improvements. Especially since some GSoC projects...
## Problem Currently, casting from textual types (e.g., `TEXT`) to numbers relies on the default PostgreSQL number handling. This will be solved for the `NUMERIC` type once #1107 is solved,...
## Problem Currently, we're not exposing hints about how different grouping methods work in the API, so we'd need to hard code some of that information in the front end...
## Describe the bug The language footer autonyms (words by which languages refer to themselves) are sometimes not properly capitalized ('properly' here means proper within the grammar of the self-referring...
Fixes #1786 Now, we automatically append a last-step ordering by the primary key field(s) of a table when getting its records. **Technical details** This does _not_ handle the case where...
## Problem Mathesar currently doesn't support any kind of authentication to use the API, and has not users/permission control (at the Mathesar level; the underlying DB could). ## Proposed solution...
## Problem Currently, it's possible to have a connection to a database defined in Mathesar without that database having a working, or up-to-date Mathesar installation. ## Proposed solution We should...