System Design and Core Capabilities
System Design and Core Capabilities
Hello to all readers, I’m starting this project to build a free software and open-source operating system tailored for essential infrastructure, embedded devices, and scientific applications. A fitting example is when such an OS powers a Mars rover or satellite, managing real-time data processing... or runs the machines behind cars and power plants. Essentially, I aim to create a Linux alternative that truly serves community needs, especially since Linux already has a vast codebase making major enhancements challenging. Your feedback would be invaluable—please share your ideas or updates whenever you have them. Thank you, and stay tuned as I set up the server with the upcoming repository and website.
Trying to build a Linux competitor from scratch is extremely challenging—especially if you're passionate about it. Making it functional is already a significant achievement. Key enhancements like...? Many useful but currently absent features depend heavily on existing code, making progress difficult. Why not focus on completing HURD? It's been around for decades, and finally someone should invest the effort to make it a solid release?
Thank you for your feedback so far. I want to be clear that my aim isn't to replace Linux, but rather to develop a competing solution in the scientific computing space—specifically targeting markets like satellite control or high-performance computing environments. My secondary focus will be on infrastructure systems, which is a less competitive area. Your insights are valuable, and I truly appreciate your time and perspective.
In my view, the most useful suggestion is to adopt a microkernel or hybrid design, embracing a modern "monolithic Unix" style. Following today’s Nix OS trends seems outdated; Linux reflects this shift clearly. Plan 9 is no longer supported, making it a potential successor. Keep a stable ABI and consider a C-compatible interface to simplify development for FOSS projects, encouraging broader participation. (Speaking of this, early on it was just a thought—whatever you do) C remains timeless. For compatibility, favor x86_64 over ARM, ensuring smooth support for both current and future systems.
99.99999999999999999999............% chance you will fail, much less so for competing against something as robust as Linux.
Linux works well for most of these applications, and it's currently the standard for many projects. It offers strong performance, scalability, and a lightweight design. If you have a particular use case that could be improved, consider developing a patch for Linux rather than building from scratch. Also, it's considered a specialized option.