When Linuxconf receives an HTTP request, it will start to process the request and will stay running for some time. After a period of inactivity, equal to the timeout, linuxconf will exit. While this has little impact on its functionality (the HTTP protocol is logically stateless -- each transaction is conceptually complete), some complex HTML dialogs exist which require linuxconf to remember information about pages it has generated. If you are too slow to fill a screen and accept it, linuxconf may very well quit in the meantime. This has two drawbacks:
Using this feature you can raise the timeout so linuxconf will wait longer before exiting. The timeout is expressed in minutes.