2007 Circumvention landscape report: methods, uses, and tools
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
As the Internet has exploded over the past fifteen years, recently reaching over a billion users, dozens of national governments from China to Saudi Arabia have tried to control the network by filtering out content objectionable to the countries for any of a number of reasons. A large variety of different projects have developed tools that can be used to circumvent this filtering, allowing people in filtered countries access to otherwise filtered content. In this report, we describe the mechanisms of filtering and circumvention and evaluate ten projects that develop tools that can be used to circumvent filtering: Anonymizer, Ultrareach, DynaWeb Freegate, Circumventor/CGIProxy, Psiphon, Tor, JAP, Coral, and Hamachi. We evaluated these tools in 2007 -- using both tests from within filtered countries and tests within a lab environment -- for their utility, usability, security, promotion, sustainability, and openness. We find that all of the tools use the same basic mechanisms of proxying and encryption but that they differ in their models of hosting proxies. Some tools use proxies that are centrally hosted, others use proxies that are peer hosted, and others use re-routing methods that use a combination of the two. We find that, in general, the tools work in the sense that they allow users to access pages that are otherwise blocked by filtering countries but that performance of the tools is generally poor and that many tools have significant, unreported security vulnerabilities. The report was completed in 2007 and released to a group of private sponsors. Many of the findings of the report are now out of date, but we present them now, as is, because we think that the broad conclusions of the report about these tools remain valid and because we hope that other researchers will benefit from access to the methods used to test the tools. Responses from developers of the tools in question are included in the report.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.005 | 0.013 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it