DigiNews

Summary

The Lone Lisp Heap offers a technical look at how Lone’s Lisp-like runtime evolves its memory model: from a handcrafted memory allocator with a simple free-list to a flat, index-based heap of Lisp values and a page-based memory strategy using mmap and mremap. It also discusses the pitfalls of simple allocators (fragmentation, metadata overhead) and the rationale behind moving to a tag-based, position-independent value representation and a centralized heap.