Hi Dave, Thank you for confirming that human judgment is still a key part of release validation. I agree with you. I also appreciate the update on how ATR is changing and what it is meant to support. I’ll look more closely at how it works.
To answer your other questions: - The analysis was based on parsing and classifying all [VOTE] and [RESULT] subject lines on [email protected] 2015 through September 2025. Threads were grouped by project and outcome. - I used AI tools to help create scripts for downloading and parsing the data, as well as for cleaning up and summarising it. I mostly did not use AI for making inferences or interpretations. I double-checked the data with lists.apache.org, and also used AI to verify any interpretations. •-I also checked the decrease in active podlings. The overall trends are similar, and I did not find any connection to the number of podlings. It seems likely that we now have more mentors available for new projects, and that some mentors have taken on less work. The piece is intended to highlight cultural and procedural change across the decade. Kind regards, Justin --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
