TL;DR key findings:
- bug fixes were >50% of the workload in many repos
- Cycle times varied greatly; some repos like Microsoft VSCode were < 3 hours while others like NodeJS were > 70 hours
- Higher technical debt unsurprisingly correlates with longer cycle and merge times
- Repos with high maintenance activity still have steady PR counts despite longer cycle times
TL;DR key findings: - bug fixes were >50% of the workload in many repos - Cycle times varied greatly; some repos like Microsoft VSCode were < 3 hours while others like NodeJS were > 70 hours - Higher technical debt unsurprisingly correlates with longer cycle and merge times - Repos with high maintenance activity still have steady PR counts despite longer cycle times