MIT have developed a system that makes web pages load faster

34% faster to be exact

Young exhausted African American woman sitting and using laptop.

Over the years our internet connections have gotten faster, but thanks to these increases in speed we’ve been able to develop websites that are increasingly complex. As a result, we can still find ourselves tapping our fingers impatiently on the sides of our phones or across our keyboards as we wait for a page to load. There have been data compression techniques developed in order to try and reduce load times, but researchers at MIT Computer Science and Artificial Intelligence Laboratory and Harvard University have decided to go in a slightly different direction with their attempt that they’re saying is “more consistent and more substantive” with its results.

The researchers are calling their system Polaris, and they’re saying it can reduce page-load times by up to 34%. When you click a link or enter a URL that you want to visit, your browser has to fetch, apply, and load things like the HTML files, JavaScript source code, and images. The problem just now is that your browser is not a planner. It can’t fetch, evaluate and load all of these elements individually because they’re all dependent on each other and before it gets to work, the browser can’t actually see where these dependencies are to determine the best order of action before hand. As a result, your browser might try to load one part of a page and find it’s dependent on the loading of another, meaning it has to stop this process, perform the other one, and then come back again. These cross network trips all add up to result in a slow loading time.

The Polaris system essentially gives your browser a plan of action, logging all of a web page’s dependencies and compiling them into a graph that your browser can use to establish the most efficient loading path ahead of time. The team acknowledge that dependency tracking methods like this have been attempted before, but say these methods that focussed on comparing lexical relationships don’t”capture more subtle dependencies” that Polaris is able to.

The team have compare the process to that of a travelling salesman:

“When you visit one city, you sometimes discover more cities you have to visit before going home. If someone gave you the entire list of cities ahead of time, you could plan the fastest possible route. Without the list, though, you have to discover new cities as you go, which results in unnecessary zig-zagging between far-away cities.

For a Web browser, loading all of a page’s objects is like visiting all of the cities. Polaris effectively gives you a list of all the cities before your trip actually begins. It’s what allows the browser to load a webpage more quickly.”

Polaris has been written in JavaScript so it could be used by any website in unmodified browsers. The researchers say that Polaris will work particularly well with larger more complex sites, as well as mobile networks “since those tend to have larger delays than wired networks.”


Main Image © iStock/BraunS