dosubot[bot] commented on issue #37925: URL: https://github.com/apache/superset/issues/37925#issuecomment-3948918643
<!-- Answer --> Thanks for the additional context! With UTC offset set, no time comparison enabled, and second-level granularity, the bug becomes clearer. **Key insight**: The [normalize_dttm_col function](https://github.com/apache/superset/blob/829e4d92d91ceae4b43b1ed3b063ffe45377799c/superset/utils/core.py#L1668-L1711) applies the datasource UTC offset unconditionally (regardless of time comparison settings) by parsing timestamps as naive datetimes and then adding offset hours. This logic appears to be misapplying offsets inconsistently across rows, which explains the non-uniform shifts you're seeing. This aligns with a [related regression in Bar Charts on master](https://github.com/apache/superset/issues/37947#issuecomment-3894691320) where UTC offsets are being incorrectly applied during backend normalization. There's also an [open PR #37014](https://github.com/apache/superset/pull/37014) aimed at fixing timezone offset handling, but it hasn't been merged yet. **Regarding priority labeling**: I don't have the ability to change issue labels or priority settingsāthat requires a maintainer with triage permissions. However, you've made a strong case for high priority given that: - It's a regression on master - It affects core time-based visualization reliability - The behavior is non-deterministic and breaks trust in chart output I'd suggest pinging a maintainer (e.g., @betodealmeida, @villebro, or @dpgaspar) directly to escalate this for triage. **Temporary workaround**: You might try setting the dataset offset to 0 (no offset) to see if that eliminates the inconsistent shifts, though this may not be viable for your use case. <!-- Dosu Comment Footer --> *To reply, just mention [@dosu](https://go.dosu.dev/dosubot).* --- <sup>How did I do? [Good](https://app.dosu.dev/response-feedback/47a271d1-c1d0-4043-b88e-16ebb8dc56c2?feedback_type=great_response) | [Irrelevant](https://app.dosu.dev/response-feedback/47a271d1-c1d0-4043-b88e-16ebb8dc56c2?feedback_type=irrelevant_answer) | [Incorrect](https://app.dosu.dev/response-feedback/47a271d1-c1d0-4043-b88e-16ebb8dc56c2?feedback_type=incorrect_sources) | [Verbose](https://app.dosu.dev/response-feedback/47a271d1-c1d0-4043-b88e-16ebb8dc56c2?feedback_type=too_verbose) | [Hallucination](https://app.dosu.dev/response-feedback/47a271d1-c1d0-4043-b88e-16ebb8dc56c2?feedback_type=hallucination) | [Report š](https://app.dosu.dev/response-feedback/47a271d1-c1d0-4043-b88e-16ebb8dc56c2?feedback_type=bug_report) | [Other](https://app.dosu.dev/response-feedback/47a271d1-c1d0-4043-b88e-16ebb8dc56c2?feedback_type=other)</sup> [](https://app.dosu.dev/a28d3c7e-a9d3-459e-9fb6-3a6f9ff4f357/ask?utm_source=github) [](https://cursor.com/link/prompt?text=Thanks%20for%20the%20additional%20context%21%20With%20UTC%20offset%20set%2C%20no%20time%20comparison%20enabled%2C%20and%20second-level%20granularity%2C%20the%20bug%20becomes%20clearer.%0A%0A%2A%2AKey%20insight%2A%2A%3A%20The%20%5Bnormalize_dttm_col%20function%5D%28https%3A//github.com/apache/superset/blob/829e4d92d91ceae4b43b1ed3b063ffe45377799c/superset/utils/core.py%23L1668-L1711%29%20applies%20the%20datasource%20UTC%20offset%20unconditionally%20%28regardless%20of%20time%20comparison%20settings%29%20by%20parsing%20timestamps%20as%20naive%20datetimes%20and%20then%20adding%20offset%20hours.%20This%20logic%20appears%20to%20be%20misapplying%20offsets%20inconsistently%20across%20rows%2C%20which%20explains%20the%20non-uniform%20shifts%20you%27re%20seei ng.%0A%0AThis%20aligns%20with%20a%20%5Brelated%20regression%20in%20Bar%20Charts%20on%20master%5D%28https%3A//github.com/apache/superset/issues/37947%23issuecomment-3894691320%29%20where%20UTC%20offsets%20are%20being%20incorrectly%20applied%20during%20backend%20normalization.%20There%27s%20also%20an%20%5Bopen%20PR%20%2337014%5D%28https%3A//github.com/apache/superset/pull/37014%29%20aimed%20at%20fixing%20timezone%20offset%20handling%2C%20but%20it%20hasn%27t%20been%20merged%20yet.%0A%0A%2A%2ARegarding%20priority%20labeling%2A%2A%3A%20I%20don%27t%20have%20the%20ability%20to%20change%20issue%20labels%20or%20priority%20settings%E2%80%94that%20requires%20a%20maintainer%20with%20triage%20permissions.%20However%2C%20you%27ve%20made%20a%20strong%20case%20for%20high%20priority%20given%20that%3A%0A-%20It%27s%20a%20regression%20on%20master%0A-%20It%20affects%20core%20time-based%20visualization%20reliability%0A-%20The%20behavior%20is%20non-deterministic%20and%20breaks%20trust%20in%20chart%20outpu t%0A%0AI%27d%20suggest%20pinging%20a%20maintainer%20%28e.g.%2C%20%40betodealmeida%2C%20%40villebro%2C%20or%20%40dpgaspar%29%20directly%20to%20escalate%20this%20for%20triage.%0A%0A%2A%2ATemporary%20workaround%2A%2A%3A%20You%20might%20try%20setting%20the%20dataset%20offset%20to%200%20%28no%20offset%29%20to%20see%20if%20that%20eliminates%20the%20inconsistent%20shifts%2C%20though%20this%20may%20not%20be%20viable%20for%20your%20use%20case.) [](https://go.dosu.dev/discord-bot) [](https://twitter.com/intent/tweet?text=%40dosu_ai%20helped%20me%20solve%20this%20issue!&url=https%3A//github.com/apache/superset/issues/37925) -- 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]
