Skip to content

Refactoring: Remove params_glob Global State

Date: February 22, 2026
Status: Complete
Scope: SIMPLE execution contexts

Executive Summary

Eliminated module-level global variable params_glob from the library by refactoring procedures to explicitly pass class(parameters) instances as arguments. This improves code safety, testability, and eliminates hidden dependencies that complicate distributed execution and error diagnostics.

Global Dependency Issues

  • params_glob was a module-level global in simple_parameters module, used everywhere
  • Hidden dependency made code difficult to trace and refactor
  • Violated encapsulation principles
  • Complicated distributed execution where multiple parameter contexts exist simultaneously
  • Created implicit couplings between procedures

Validation Checklist

  • [x] All procedure signatures updated consistently
  • [x] All call sites passing params argument explicitly
  • [x] No remaining params_glob references

Benefits

Safety

✅ Eliminates hidden global state — All parameter dependencies explicit and traceable
✅ Prevents race conditions — Multiple concurrent parameter contexts won't interfere
✅ Improves testability — Procedures can be tested with specific parameter configurations

Maintainability

✅ Code clarity — Dependencies visible in function signatures
✅ Refactoring safety — Grep/IDE tools can identify all affected call sites
✅ Future extensibility — New parameter versions can coexist during migration

For Future Development

  • All new procedures should follow the explicit parameter-passing pattern
  • Do not introduce new global state for parameter management
  • Test with multiple parameter contexts to detect hidden dependencies

Conclusion

This refactoring eliminates a significant source of hidden complexity in the SIMPLE library. By threading parameters explicitly, we gain: - Transparency — All dependencies visible in code - Safety — No implicit global state pollution - Debuggability — Better error traces and diagnostic output The fix enables safer concurrent execution contexts and provides a foundation for future distributed execution improvements.