Tag

Legacy Code

All articles tagged with #legacy code

Voyager’s legacy code still steers interstellar ships as aging engineers guard fading memory
space10 days ago

Voyager’s legacy code still steers interstellar ships as aging engineers guard fading memory

NASA continues to run the Voyager probes with 1970s assembly-language software on purpose-built hardware, with total onboard memory around 64–70 KB. The operating team is small and aging, and much original documentation remains lost or scattered on paper, making it difficult to replace the few engineers who truly understand the code. The real challenge is preserving institutional memory and expertise as generations of specialists retire, not merely deciphering unreadable code, even as Voyager data transmission could continue into the mid-2030s.

technology18 days ago

Windows 11's backbone: Win32 from Windows 95 still powers today

Microsoft CTO Mark Russinovich confirms Windows 11 still relies on the Win32 API dating back to Windows 95 for core operations, with a vast ecosystem built on it. Efforts like WinRT to modernize the API didn’t replace Win32, so much of today’s OS and applications remain grounded in decades-old code, even as Microsoft markets Windows 11 as the most modern and secure Windows yet.

IBM's GenAI: Revolutionizing COBOL Code Refactoring and Translation to Java
technology2 years ago

IBM's GenAI: Revolutionizing COBOL Code Refactoring and Translation to Java

IBM is developing the watsonx Code Assistant, an AI tool aimed at refactoring ancient COBOL code for its mainframe systems. The tool will analyze, refactor, and test the code, with humans still involved in the process. IBM hopes that the generative AI will help decouple individual services from monolithic COBOL apps, making the code more modular. The company plans to offer the tool in three steps: refactoring, transforming the code into mainframe-friendly Java or COBOL, and validation with AI-generated test cases. While AI assistance could help modernize COBOL code, concerns remain about the loss of institutional memory and the potential for AI-generated code to introduce errors.