omnipkg v1.0.5: NumPy + SciPy Version Mixing Without Crashes. Every Time.
Release Date: August 8, 2025
Version: 1.0.5
We are thrilled to announce omnipkg v1.0.5, a landmark release that solves a critical, long-standing challenge in Python dependency management. While omnipkg has always provided fluid, in-runtime version swapping, this release is a true "mission accomplished" milestone. We have successfully implemented a nuclear-grade stress test that proves our runtime can consistently and reliably mix the versions of Python's most notorious C-extension libraries, NumPy and SciPy, without crashes or errors.
✨ What's New in 1.0.5
Validated Nuclear Stress Test: The Holy Grail of Dependency Management
This release marks a major breakthrough. We have validated that omnipkg can perform the previously unstable: using different, conflicting major versions of C-extension packages like numpy and scipy within the same Python script runtime.

• The Problem: The complexity of C-extensions often makes reliable combo-swapping impossible. While omnipkg could sometimes perform these swaps in the past, inconsistencies—especially with NumPy's intricate C-API—made it unreliable for production use.
• The omnipkg Solution: Our stress test has been validated to pass 100% of the time. It proves that omnipkg can:
  • Dynamically Juggle: Fluidly activate and deactivate isolated package "bubbles" in a running Python environment.
  • Perform C-Extension Combos: Run a single script that successfully uses numpy==1.24.3 and scipy==1.12.0 simultaneously.
  • Guarantee Consistency: This fix provides a new benchmark for reliability, making multi-version C-extension environments a practical reality.
    This functionality is a game-changer for scientific computing, machine learning, and data analysis, where projects often require a complex mix of libraries with strict version requirements.
    Improved omnipkgLoader for C-Extension Robustness
• We've enhanced the omnipkgLoader with more sophisticated logic to handle C-extension modules. By intelligently determining when to perform a full process-level isolation (the "nuclear option"), we ensure that package juggling remains fast and efficient for pure-Python libraries while being rock-solid and reliable for complex binary packages.
  Updated Build & Packaging
• Minor fixes and optimizations to the build and bubble creation process.
• The project now includes a validated stress test script (omnipkg/stress_test.py) that users can run to see this unique capability in action.
  This release is a testament to the power of a new approach to dependency management. With omnipkg, developers can say goodbye to the overhead of multiple environments and hello to unprecedented flexibility and control.
  Thank you for your support,
  The omnipkg Team