The only thing that's certain to happen quite regularly on a web page are page requests. This is where pseudo-cron comes into play: With every page request it checks if any cron jobs should have been run since the previous request. If there are, they are run and logged. pseudo-cron uses a syntax very much like the Unix cron's one. All job definitions are made in a text file on the server with a user-definable name (usually crontab.txt). An extensive logging facility is also included for logging job results, and you may choose to receive emails with the job results.