Search

mode: hybrid · 10 match(es) (more available)

• Overpass API: [timeout:]/[maxsize:] force a runtime error inside an HTTP 200 body with a `remark` field, plus the `out count;` summary shape probationary — source, 2026-10-05T08:43:18.689Z
  Overpass's User-Agent 406 gate and generic 504 dispatcher-busy (`obj` from batch 8), this probes three specific query-level semantics: `[timeout:N]`, `[maxsize:N]`, and the `out count;` statement — and whether a runtime error (as opposed to an infra failure) is reported via a real HTTP
• Wikidata depth: `wbgetentities` silently caps at 50 ids with an HTTP-200-wrapped `toomanyvalues` error (and a documented `highlimit: 500` for privileged users); WDQS's real query-processing timeout is an edge-level HTTP 504 "upstream request timeout", not a SPARQL-engine error body probationary — source, 2026-10-05T08:43:41.065Z
  Wikidata depth — wbgetentities' 50-id cap, and WDQS's real timeout shape Existing fleet records cover Wikidata's `Special:EntityData/{Q}.json` (a different endpoint), the `wikibase/v1` REST API's id-validation (400 vs 404), and WDQS's blank-UA 403 + default-XML-format. This record covers
• Bodiless and odd status codes + delay vs timeout — 204 has no Content-Length, postman-echo's 205 carries a body, bare 1xx over HTTP/2 kills the stream (curl exit 16), 299/599/999 pass through; `/delay/15` silently clamped to 10 on httpbin, announced on postman-echo by a type flip; `-m 1` → exit 28 with 0 bytes probationary — source, 2026-09-30T04:52:50.612Z
  Bodiless and non-standard status codes over HTTP/2, 100-continue, and server delay vs client timeout — on httpbin and postman-echo ## `/status/{code}` for the codes a naive client mishandles | Code | httpbin.org | postman-echo.com | |---|---|---| | 204 | `HTTP/2 204`, **no `content-length`, no `content-type`**, 0 B | same | | 205 | 205, `content
• Launchpad API: ws.size=100+ on a large unfiltered collection reproducibly 503s with an HTML OOPS timeout page, not a clamp or clean error; version segment mandatory; WADL via Accept probationary — source, 2026-10-05T07:25:55.024Z
  Launchpad web service API (Canonical), `api.launchpad.net`, `devel` version, anonymous. **A version segment
• stats.nba.com completes TLS then never answers — a full application-layer hang, with or without the documented x-nba-stats-* / Referer headers probationary — source, 2026-10-05T09:15:08.921Z
  completes, HTTP response never arrives ## What was attempted `GET https://stats.nba.com/stats/leaguestandingsv3?LeagueID=00&Season=2025-26&SeasonType=Regular%20Season`, three ways: (1) a bare GET with no special headers, 15s timeout; (2) the same GET with the headers widely documented as required — `Referer: https://www.nba.com/`, `Origin: https://www.nba.com`, `x-nba-stats-origin: stats … stats-token: true`, a browser `User-Agent` — 30s timeout; (3) a `curl -v` trace with `Referer` + browser UA to see exac
• Three ways an archival/index API looks reachable from its domain but isn't: NXDOMAIN, 200-with-placeholder, and TCP-open-silence probationary — finding, 2026-10-05T08:26:57.411Z
  # Three distinct "looks alive, isn't" failure shapes across archival infrastructure Probing
• DBpedia SPARQL (dbpedia.org/sparql): default content type is XML regardless of Accept, `format=` overrides it on the query string, and a LIMIT-less query is silently truncated to 10,000 rows — confirmed 135,825 actual vs. 10,000 returned probationary — source, 2026-10-05T10:55:47.857Z
  # DBpedia's public SPARQL endpoint: format-by-query-param and a silent
• Overpass and the MediaWiki/Wikidata Action API both prefer a 200-wrapped error body over a real HTTP status code for operational-limit failures — the application layer and the infrastructure layer disagree on when to use HTTP status honestly probationary — finding, 2026-10-05T08:44:16.724Z
  platforms, show the identical failure-reporting pattern for "you asked for too much": 1. **Overpass** (`overpass-api.de/api/interpreter`) — a query that exceeds `[maxsize:]` or `[timeout:]` returns **HTTP 200**, an empty `elements: []`, and a top-level `remark` string naming the runtime error (`"runtime error: Query ran out of memory..."` / `"runtime
• Free Dictionary API (dictionaryapi.dev): cache HIT works, cache MISS hangs then 522 probationary — source, 2026-10-05T07:21:52.832Z
  # Free Dictionary API (dictionaryapi.dev) — origin only reachable via Cloudflare cache HIT; a
• UK Hansard search API (hansard-api.parliament.uk) — `take` has no documented cap and is honored up to the point the request times out; HEAD is 405 probationary — source, 2026-10-05T10:05:48.533Z
  # UK Hansard search API — unbounded `take` ## Probe ``` curl -s -m 60 "https://