code wiki / _hdl_build / nx_uat_run.nx
nx_uat_run.nx
buildroot/runtime/_hdl_build/nx_uat_run.nx
about
nx_uat_run.nx -- THE CONSUMER UAT RUNNER (operator doctrine 2026-08-02: digital-twin, see it ->
log it -> work it -> UAT it, on cadence -- the loop a team of people AND consumers would run).
Walks knowledge/uat_journeys.conf (id|url|connect|latency_ms_max|must_contain), fetches each
journey through the REAL sovereign edge by forking the promoted nx_https_get.elf (the same bytes a
user's request traverses), measures WALL latency as a human feels it, checks HTTP 200 + the content
the human came for, and journals ONE frame per journey to knowledge/status/uat.jrnl (append-only).
A journey slower than its bar FAILS even if the bytes are perfect: the bar is the consumer's
patience, not the backend's throughput (seeded by the 8845ms SERP finding, 2026-08-02).
Bars/journeys live in DATA (rule 11): expanding UAT = adding a conf row, never a code change.
license_tier: ORIGINAL expect_exit: 0 when every journey passes
dependencies 3 imports · 0 importers
imports: nx_syscalls.nxnx_estate_path.nxnx_gate_verdict.nx
imported by: nobody (leaf or entry point)
call flow from main pre-order; caps 40 nodes / depth 6 declared; ↻ = already shown
structs
| none |
consts
| 17 | const UA_FORK_FAILED: i64 = 0 - 1 |
| 35 | const UA_FETCHER: *u8 = "nx_https_get_cli2.elf" as *u8 |
functions
| 37 | func ua_slen(s: *u8) -> i64 { var n: i64=0; while s[n]!=(0 as u8){n=n+1} return n } called by 1: ua_find |
| 39 | func ua_find(hay: *u8, n: i64, needle: *u8) -> i64 |
| 53 | func ua_atoi(s: *u8) -> i64 called by 1: main |
| 82 | func ua_fetch(url: *u8, conn: *u8, out_path: *u8) -> i64 |
| 106 | func ua_journal(id: *u8, ok: i64, ms: i64) -> i64 |
| 135 | func main() -> i64 |