dstandish commented on PR #30727: URL: https://github.com/apache/airflow/pull/30727#issuecomment-1535200593
> And even recently we had a problem where --preload flag implemented to fix problems in webserver caused something else [to break] Yeah exactly, every change, even a bug fix, has the potential to create more bugs. So even when we try our best to be consistent with semver policy and only introduce bugfixes in patch release, we can still break things. But we do so because, that's the tradeoff -- you try to fix something in exchange for small "new bug" risk. So from user perspective it's a fair trade. I am not trying necessarily to weigh in on what should happen in this specific case because I think I'm a little more removed from what's at stake. But just trying to help articulate the rationale for why patch releases, _as a general rule,_ should be governed more by "is bugfix" than "is not feature". Of course we can always make exceptions e.g. if it makes big impact re developer convenience with little likelihood of breakage. -- 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]
