Blog

CodeScene 3.3: Change Coupling Trends and Delivery Performance Metrics

We’re proud to announce a new release of CodeScene, version 3.3!

CodeScene 3.3 detects dependency trends on both file- and architectural level. This is useful to reason about the change impact, e.g. in microservice architectures. This new version of CodeScene also includes a Delivery Performance module that focuses on throughput and lead time metrics that are known to correlate with business values.

New Features

  • Visualize Change Coupling Trends: Detect trends in change coupling to highlight dependencies that increase or decrease. We have found this feature particularly useful for large, complex architectures like microservices where the emerging system behavior makes it hard to get a holistic picture of how the building blocks co-evolve.
Visualize change coupling dependency trends between microservices and/or other architectural building blocks. Example from Spinnaker.

Visualize change coupling dependency trends between microservices and/or other architectural building blocks. Example from Spinnaker.

  • Measure Change Coupling Trends between Code in separate Git Repositories: The change coupling trends are available for the file-level metrics too. We have found it most valuable for measuring coupling between code located in separate Git repositories and/or code implemented in different programming languages. CodeScene handles both cases.
Change coupling between files implemented in different programming languages and located in separate Git repositories.

Change coupling between files implemented in different programming languages and located in separate Git repositories.

  • Measure Delivery Performance: CodeScene’s Delivery Performance module focuses on throughput and lead time metrics that are known to correlate with business values like time to market, customer satisfaction, and profitability (see the excellent research in Accelerate: The Science of Lean Software and DevOps by Forsgren, Humble, and Kim for a deeper read on this topic).
Measure lead times to visualize the effects of any quality issues that CodeScene detects.

Measure lead times to visualize the effects of any quality issues that CodeScene detects.

  • Run Multiple Analyses in Parallel: Many CodeScene users are part of larger organizations that might have many analysis projects. You can now run concurrent analyses to shorten the overall time it takes to analyze all codebases.

Improvements

  • Introduce an Active Authors view in case the license is exceeded.
  • Add Gerrit support for CodeScene’s Jenkins plugin.
  • Correct sorting of branch duration in the feature branch analysis.
  • Minor bug fixes and improvements.

Get your license for CodeScene here.

Adam Tornhill avatar.
,
comments powered by Disqus