There's a much more active current thread[1].

[1] Qwen3.6-27B: Flagship-Level Coding in a 27B Dense Model https://news.ycombinator.com/item?id=47863217