'Remarkable' new algorithm could dramatically speed up web browsing
A new algorithm could significantly speed up web browsing by making caching more effective. The open-source program, called " SIEVE ," introduces a new way to handle web caching — the process of storing and retrieving objects from a computer's long-term storage as you encounter them while surfing the internet. These objects — tiny files stored on your hard drive — include images, logos or entire copies of webpages. When you encounter these elements for the first time, you retrieve them from the server, but they are stored on your hard drive for reuse. The second time you encounter these objects, your browser can retrieve them from your computer's memory rather than from the server, which saves time and consumes less energy. But because local storage is limited, cache-eviction algorithms work to decide how long to store objects for, and when to replace older ones less frequently accessed by a user, with newer or more popular ones. Although many such al...