OODA-HTTP / Protocol & research

Results

Understand which experimental records are available and what they can support.

Experimental records

The project reports a 150-request proof of concept and a separate three-request functional check. The records below have different scopes and are not combined into one validated experiment.

Reported evidence; consolidated reproduction pending
RecordWhat it addressesEvidence limit
Reported 150-request experimentSelected-action visibility and reconstruction in a controlled HTTP path.Earlier summaries require reconciliation with the matching logs and analyzer. A stripping problem was reported for this run; it is not evidence of successful stripping across 150 requests.
Separate preliminary three-request checkOne implicit, one explicit and one strip-mode request. The reported strip observation shows the token at the gateway and absent at the edge.A small functional check. It does not establish a latency distribution or a general overhead benchmark.

How to read a result

Observation point

Identify whether the measurement describes the application, gateway or edge. A token can be visible at one point and absent at another.

Calculation model

State the reconstruction rule and candidate-set assumptions. Four equally likely candidates yield 2 bits under that model; this is not automatically an empirical entropy estimate.

Reproduction

A complete result needs the run-specific logs, analyzer version, request joins, exclusions and mode counts. A consolidated reproduction package is not published here.

Read the measurement method →