Internet Filtering: Client-based vs. Server-based

The original filtering solutions were client-based -- programs that ran on individual workstations. They typically use keyword filtering or small URL databases, or combinations of both. They tend to have high support costs, are difficult to update, are often easy to disable, and generally aren't very effective.

Server-based solutions started being used in about 1994. These are usually implemented as servers that run on the network between workstations and the Internet. They are easier to maintain and update than client-based solutions and are difficult to disable or circumvent. Their primary disadvantage is that they don't offer as much end-user control over the filtering parameters.

Hybrid solutions are just starting to come into use recently. These combine the flexibility of client software with the ease-of-maintenance and power of servers. When implemented properly they can partially overcome the disadvantages of client software because they are simpler and need less maintenance, and can be integrated into existing client software.