One ruler to measure them all: Benchmarking multilingual long-context LLMs | Not Hacker News!