Ujjwaljain16 commented on issue #44174: URL: https://github.com/apache/superset/issues/44174#issuecomment-5831540580
one thing am curious about regarding the execution convergence: does SIP-227 intend GTF to be the convergence point only for task lifecycle/execution or also for the underlying query-execution path and its cross-cutting invariants? i was looking through the current SQL Lab/GTF work and noticed an interesting split SQL Lab and MCP’s `execute_sql` now converge on `Database.execute()` → `SQLExecutor` → `apply_rls()`, where RLS rules are resolved from the tables referenced in the SQL. Chart-data (and MCP’s `query_dataset`) still take the `QueryContext`/`ChartDataCommand` path, where RLS is applied directly to the resolved dataset before reaching `Database.get_df()` so there are currently two different mechanisms for resolving/applying the same underlying RLS rules, with some explicit handling for their interaction around virtual datasets since #43928 already lists moving chart-data onto `Database.execute()` as future direction, am wondering: when SIP-227 says GTF becomes the shared execution contract, is the intention eventually to converge these underlying execution/security semantics as well or is GTF deliberately only the orchestration/lifecycle layer while those remain separate? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
