Puerto Rico Proxies

Run a documented cold request, one immediate repeat, and one request after the chosen wait interval. Keep query strings and headers fixed unless the planned step changes exactly one cache key, and label rather than discard cache misses. The PRI worksheet needs a direct-control identifier, proxy configuration hash, observed country, exit IP, connection identifier, status or error class, safe content marker, and UTC timestamp. Store rejected outcomes instead of repeating until a preferred response appears. Use only an endpoint permitted for the Puerto Rico test, keep missing measurements explicit, and never rotate around a refusal or other mandatory stop.

Separate server output from browser translation, saved preferences, experiments, extensions, and account data. If those inputs cannot be held constant, describe the observation as mixed rather than as proof of country-specific presentation. For annual catalogue context, compare Puerto Rico with Aruba's 12,116 and Bermuda's 11,037 within North America. Do not infer relative availability, quality, latency, or buyer value from that stored-number sequence. Keep PRI comparator rows labelled and diagnose the requested Puerto Rico route from its own evidence rather than a neighbouring count. Record why each comparator belongs in the authorized review before using it.

Compare offers on the same workload, sample size, valid-result rule, metering basis, and minimum commitment. Reject a headline unit price that depends on hiding failed observations or assuming the historical pool is live inventory. Before purchasing Puerto Rico, make the PRI threshold reproducible: state product, country precision, protocol, session mode, representative workload, valid-result rule, retries, evidence, and retest trigger. Include every planned outcome when applying it. Reject any Puerto Rico conclusion built from an unverified exit, hidden failures, or a changed rule, and repeat PRI verification after a material configuration change.