Friendship ended with Deno, now Node is my best friend

Commenters worried that Deno is sliding toward obscurity: one former contractor said layoffs left no roadmap or comms and still hopes it recovers, while others noted practical failures (OpenSSL in child processes, compatibility headaches) and argued Deno’s custom APIs force you to build a “Deno app,” reducing portability. Several people prefer Bun for performance and fewer dependencies - praising riscv64 support and simpler developer experience - and some see Node as the pragmatic safe bet that will ultimately absorb or outlast alternatives. Others framed the trend around LLMs and agents, saying rapid context shifts make investing in niche runtimes riskier and that teams now pick specialized stacks (Go, React, C#) rather than a single shared language.

Discussion split on TypeScript and protocol correctness. One commenter framed Node’s restriction on shipping raw TS as philosophical (Microsoft ownership); others rebutted that the real issue is tsconfig mismatches and pointed to declaration files as the right solution. A smaller argument revolved around GET requests with bodies: some said sending a body on GET is wrong or unsupported (citing RFC/Fetch semantics), while others asked whether web fetch allows it, with an anecdote about GET-body ambiguity in standards teaching. Overall positions ranged from nostalgia for unified toolchains to pragmatic prioritization of compatibility, performance, and ecosystem stability.