It’s no secret that China tries to filter the Internet content its citizens can see. But a new US study has shown that the firewall the government has thrown up is not as effective as theymight have hoped. The study, which was undertaken by grad student Earl Barr with colleagues at UC Davis and the Universityof New Mexico, investigated the effectiveness of the Chinese firewall, and discovered that it’s least effective when many Chinese are online at once, and also that the idea of the firewallis often more potent than the firewall itself. The Chinese technology doesn’t simply block Web pages that discuss forbidden topics. It scans data for banned words or addresses. When itdiscovers one, it instructs the source server and the PC receiving information to stop the flow of data. The US group manipulated this to determine how far messages with banned terms couldpenetrate before being shut off. They loaded data streams with words from the Chinese version of Wikipedia then sent them into the Chinese network. If a data stream was stopped, they used a techniqueknown as latent semantic analysis to find related words and see if they were also blocked. The blocking tended to happen only when the data was deep in the Chinese network, and in over aquarter of the cases blocking failed completely. This would suggest that China’s firewall isn’t as all powerful as has been previously suggested, and blocking tended to be erratic whenmany users were online. The full results of the study will be presented at the ACM Computer and Communications Security Conference next month.