PR Review: NUnit3TestAdapter 6.0.0 Update

Summary

This PR updates NUnit3TestAdapter from v5.2.0 to v6.0.0 (major version bump). The change is minimal - a single line in Directory.Packages.props.

Analysis

Scope of Impact

The NUnit3TestAdapter is only used in the speed-comparison benchmarking project (tools/speed-comparison/UnifiedTests/UnifiedTests.csproj:57), specifically when running benchmarks with TestFramework=NUNIT. This is not part of TUnit's core functionality or test suite.

Code Quality ✅

• Clean, minimal change
• Follows centralized package management pattern
• No code changes required

Potential Issues ⚠️

Major Version Upgrade Concerns:

1. Breaking Changes: NUnit3TestAdapter v6.0.0 is a major version bump and may contain breaking changes. Common breaking changes in test adapters include:

   • Minimum .NET version requirements
   • Changes to test discovery behavior
   • Deprecated configuration options
   • Performance characteristics changes

2. Testing Recommendation: Since this affects the speed-comparison benchmarks, I recommend:

   • Running the benchmarks with TestFramework=NUNIT to verify compatibility
   • Checking that benchmark results remain comparable (no unexpected performance changes)
   • Example command: dotnet build tools/speed-comparison/UnifiedTests -p:TestFramework=NUNIT

Performance Considerations ✅

• No performance impact on TUnit itself (only affects NUnit benchmark comparisons)
• May affect benchmark comparison accuracy if adapter behavior changed significantly

Security Concerns ✅

• Dependency update from trusted source (NUnit project)
• No security implications for TUnit

Test Coverage ✅

• Change is isolated to dependency version
• Speed comparison tests should validate this works correctly

Recommendation

LGTM with testing caveat ✅

The change is low-risk since it only affects benchmarking/comparison tooling, not TUnit's core functionality. However, I recommend:

1. Before merging: Run the NUnit speed comparison benchmarks to ensure v6.0.0 works correctly
2. Verify: The benchmarks complete successfully without errors
3. Document: If there are any notable behavior changes in the benchmark results

Pre-Commit Checklist (from CLAUDE.md)

• ✅ No core TUnit code changes
• ✅ No source generator changes (snapshots not needed)
• ✅ No public API changes (snapshots not needed)
• ✅ No dual-mode implementation needed
• ⚠️ Should verify speed comparison benchmarks still run with NUnit v6.0.0
• ✅ Code follows project conventions (dependency update only)

Automated review by Claude Code