Splice's row editing feature currently relies on database cursors and positioned updates. This allowed it to edit very large result sets efficiently, but has the downside that a forward-only resultset cannot be updated. We should fallback on doing updates using: UPDATE TABLE SET COLX=VALX WHERE COL1=VAL1, COL2=VAL2, …
We may want to only enable this when a resultset contains the primary key(s) for the table so we don't accidentally update more rows than we intend. Alternatively, we can use the database's native rowid feature (if any) to perform safe & reliable updates.
Splice's row editing feature currently relies on database cursors and positioned updates. This allowed it to edit very large result sets efficiently, but has the downside that a forward-only resultset cannot be updated. We should fallback on doing updates using:
UPDATE TABLE SET COLX=VALX WHERE COL1=VAL1, COL2=VAL2, …
We may want to only enable this when a resultset contains the primary key(s) for the table so we don't accidentally update more rows than we intend. Alternatively, we can use the database's native rowid feature (if any) to perform safe & reliable updates.