Release notes

Release notes#

[2.8.0] - 2025-09-04#

This is a landmark release for EON, representing over five years of dedicated development. The entire framework has been modernized to be more powerful, easier to use, and more accessible to researchers across all major platforms. This release introduces cutting-edge simulation methods, a vast expansion of supported computational chemistry codes, and a complete overhaul of the build and installation process.

✨ Major New Features#

  • Advanced Transition State Finding with RO-NEB-CI: Implemented the novel RO-NEB-CI method, providing a powerful new tool for accurately locating complex transition states. (#239)

  • Machine-Learned Potentials with metatomic: Integrated full support for metatomic potentials, enabling high-performance simulations with state-of-the-art models from the metatensor ecosystem. (#201)

  • Expanded Potential Interfaces: Drastically increased interoperability by adding direct interfaces for a wide range of popular computational chemistry packages:

    • NWChem: High-performance, socket-based interface for large-scale calculations. (#244)

    • ORCA: Direct support for the versatile and efficient ORCA quantum chemistry program.

    • AMS: Interface for the Amsterdam Modeling Suite (ADF, BAND, DFTB).

    • XTB: Fast and reliable calculations using the GFN-xTB semi-empirical tight-binding methods.

    • ASE: General-purpose interface to any calculator supported by the Atomic Simulation Environment (ASE).

  • Input Validation and Schema: Implemented a comprehensive Pydantic schema for all configuration files. This provides automatic input validation, clearer error messages for users, and a robust foundation for auto-generated documentation.

  • ZBL Universal Potential: Added support for the Ziegler-Biersack-Littmark (ZBL) universal screening potential, ideal for simulating high-energy collision cascades and ion-implantation effects. (#241)

🚀 Enhancements & Improvements#

  • Nudged Elastic Band (NEB) Enhancements: The NEB module has received significant upgrades for flexibility and performance:

    • Endpoint Minimization: Added an option to pre-optimize the initial and final states of a NEB path, improving convergence and accuracy. This feature is fully compatible with restarts. (#221)

    • Custom Initial Paths: Users can now provide a custom series of intermediate structures to initialize a NEB calculation, offering greater control over the reaction pathway.

    • Dual Optimizers: Enabled the use of different optimizers for the climbing image versus the regular images in a CI-NEB calculation.

    • Energy-Weighted Springs: Implemented energy-weighted dynamic springs, improving stability and performance for reaction paths with high energy barriers.

🛠️ Build, Installation & Developer Experience#

  • Modernized Build System with Meson: The entire project has been ported to the Meson build system. This provides a significantly faster, more reliable, and truly cross-platform installation experience on Linux, macOS, and Windows. This work also lays the foundation for a future pure Python eon-server package. (#124)

  • Cross-Platform CI & Support:

    • Established a robust Continuous Integration (CI) pipeline, automatically testing builds and features across Linux, Windows, and macOS (Intel & Apple Silicon ARM).

    • Official support for Apple Silicon (M1/M2/M3) machines.

    • The command-line interface is now fully compatible with Windows environments.

  • Streamlined Dependency Management: Added official support for Conda and Mamba, allowing for simple, one-command environment setup.

  • High-Performance Logging: Replaced the internal logging system with spdlog for asynchronous, configurable, and more informative output with minimal performance overhead.

  • Code Modernization:

    • The C++ backend has been upgraded to the C++17 standard.

    • Adopted modern C++ features like smart pointers and the <filesystem> library for safer memory management and platform-independent file I/O.

    • Enforced consistent code style and formatting across the entire codebase.

📚 Documentation#

  • Added comprehensive documentation for the Nudged Elastic Band (NEB) module, covering theory, keywords, and practical examples.

Contents