50 shades of PR dark matter

on YouTube ↗14m

" summary

Data shows PRs consume 23% of cycle time and scale to huge cost; research from Linear B finds pick-up and review idle time are the main culprits. The claim is that a Continuous Merge pipeline that auto-categorizes PRs by risk, assigns dynamic code owners and applies dynamic restrictions plus CI tools like Circle CI GitHub actions drone and GitHub co-pilot cuts idle time. An implemented CM flow reports measurable cycle-time improvement.

AI-written summary · about

" chapters (7)

1. 0:00Introduction
2. 1:07Delivery cycle in software
3. 3:57Efficiency in development
4. 6:24PR bottleneck in cycle
5. 9:54Context switching in PR
6. 11:44Continuous Merge approach
7. 13:29Summary

" related talks

• 06m50 Haunting Shades of PR Dark MatterElad Kohavi · DevOpsDays Tel Aviv 2024
• 47mHow to Optimize Non-Coding TimeYishai Beeri · Devoxx Ukraine 2021
• 17m50 shades of PRAvishag Sahar · NDC Copenhagen 2025
• 42m7 Metrics to Accelerate Delivery without Damaging CultureDan Lines · Community Leadership Conference @ Open Source Summit North America 2022
• 38mHow to Optimize Non-Coding TimeYishai Beeri · Voxxed Days Bucharest 2022
• 29mBreaking the PR Review BottleneckPavel Oborin · PyCon AU 2026
• 26mA State of Continuous Merge: The Secret to More Productive Dev TeamsYishai Beeri · SREday London 2022
• 24mUnlocking CI Performance in the Age of Agentic DevelopmentScott McAllister · Commit Your Code 2026