{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreibojk4q623n7a7nmlshwx7bfrdomhqdtwfaogqs27vv5q57k3qjki",
    "uri": "at://did:plc:pi6woz4d47bkuws673w2il2r/app.bsky.feed.post/3mi53bhyw7vp2"
  },
  "path": "/t/bootstrapping-xml-schema-definitions-with-claude-opus-4-6-a-case-study-the-good-the-bad-and-the-ugly/13853#post_4",
  "publishedAt": "2026-03-27T23:50:15.000Z",
  "site": "https://discourse.haskell.org",
  "tags": [
    "GitHub - dfinity-side-projects/winter: Haskell port of the WebAssembly OCaml reference interpreter · GitHub"
  ],
  "textContent": "jackdk:\n\n> Haskell is a remarkably stable language given the pace of development at the frontier, and it seems that you could keep that around without too much additional effort.\n\nI agree. If it works and isn’t going to be some big maintenance burden (which an implementation of some spec, protocol, format, etc is), it’s worth keeping for testing’s sake alone.\n\nI worked at a crypto company once and someone there wrote a Haskell implementation of a Wasm interpreter [1]. We were gonna use v8 in prod, but having this let us not worry about the complexities of using v8. And it let us write a lot more interesting tests.\n\n[1] Cribbed from the Ocaml one iirc. Claude would’ve been good at that. GitHub - dfinity-side-projects/winter: Haskell port of the WebAssembly OCaml reference interpreter · GitHub",
  "title": "Bootstrapping XML schema definitions with Claude Opus 4.6: A case study (the good, the bad, and the ugly)"
}