When connected, you will see a webpage that can be either the standard atomsnet entrypage or a page created by the server's administrator. In the general case you will find both a search and multiple browse options. Searching the network works similar to other searchengines. Browsing is modeled to so called directories, such as Yahoo! (http://www.yahoo.com/) and the Open Directory Project (http://www.dmoz.org/ and http://directory.google.com/). Anyone accustomed to searching the web will be able to use the atomsnet. Because many computers have to be accessed for a request the waiting time can be longer than with regular searchengines, however. How long you have to wait before results are displayed is linked directly to the speed of the server, the size of the network and the bandwidth between the participating computers.
Although the program's functionality is extendible the basic browsing options are the following. Firstly, you can find files according to their type, this means all files in the network are grouped according to their nature. Searching by type (or more precisely MIME-type) uses a two level structure. You start by selecting the broad type, for instance video. After that a selection has to be made according to the precise filetype, for instance MPEG. It is, however, also possible to simply select video files and from there start a search, so that all video files regardless of their subtype that are of interest will be presented.
Another browse option is 'by category'. Categories are directly copied from the Open Directory Project format and display files not according to their type, but according to their content. This way it is for instance possible to find all files that concern football. Thousands of subcategories exits, making this a very finegrained browsetree. It is important to note, however, that this feature is experimental and not all files are categorized correctly. Hopefully the successrate will go up when new and improved plug-ins become available, but currently it is mainly useful for files where the filename contains the subject, or for mp3 audio (browse by artist).
The last browse option is 'by directory'. This option mimics the physical directory structure of the server you are accessing. It is added with the idea that people normally create their directory structure in a logical way. Using this option means relying solely on the archiving techniques of another person. It can be useful as an alternative to the browsing options discussed earlier, since many people index their files either by filetype (for instance a 'song' directory containing all mp3 files) or by category (for instance a directory containing all files belonging to a certain project).
When a file is selected, all characteristics can be shown by clicking on the filename. This way the mimetype location, the category and the directory location of the item are exposed, allowing you to browse similar files from this starting point. Furthermore, file size, date and additional information is made visible.
Since the network user interface is modeled closely to existing web resources I believe using the system is straightforward for most internet users. It is however possible for anyone that administers a server to alter the look and the working of his server to his personal liking. This means that certain features described above can be turned off, others can be added and the overall navigation can be rewritten completely. The usability of the network is completely in the hands of the server administrator. If you do not like a particular layout you can choose to 'hop' to another server until you find one that suits you. As a last resort, it is possible to create a personal entrypoint by installing the application on your local computer and connecting to the existing network.