A task is a request to retrieve web pages or files (resources) from the Internet and make them available offline. The downloaded resources are permanently stored in the project database until you explicitly delete them. You can view the downloaded pages and browse the downloaded websites using the built-in Browser.
You can think of each task as a separate spider. The spider starts by retrieving the starting webpage of a task along with the embedded resources (graphics, scripts, stylesheets, sounds, movies, frames, applets). It then searches the page for links and retrieves each linked page and its embedded resources, then searches the page for new links, and so on. The spider remembers each link it has retrieved and never retrieves the same link again.
Note that every time you use the built-in browser you browse the entire project file, not just a single task. Resources (webpages, images, and other files downloaded from the Internet) you browse with the built-in browser remain to be linked even if they are downloaded with different tasks. When you click on a link, the browser first searches for the page in the list of files retrieved by the task you browse. If the page is not found among the files of the task the browser searches the entire project database. If there are different versions of the page associated with the URL the version that has a timestamp nearest to the browsed task is retrieved.
You can save a single task or entire project file to the file system to create a duplicate of a website on your hard-drive. Alternatively, you can export a task or entire project file to a HTML Help file (.CHM).