MIT-Developed Method ‘Polaris’ Will Make the Web 34% Faster

The internet is becoming a rat race. As internet connections get faster and new techniques emerge to decrease website loading times, web pages are becoming more and more complex, negating those improvements in loading times. Tech companies like Google, Opera and Facebook have been striving to introduce new technologies which help make their specific services and apps faster. These improvements have so far failed to have a major effect on all internet.

MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) started on a project, named Polaris, to solve the problem. It is a browser agnostic method which will improve page loading speeds by up to 34 percent. MIT researchers say that one of the major bottlenecks in webpage loading is how browsers fetch data to assemble all of the page’s elements. A browser typically needs to know beforehand what it is downloading, like images, HTML, Javascript, etc. The browser, then, has to get at least one piece before it knows what else has to be downloaded. It repeats the procedure again and again, multiple times for a single page. This sums up to an inefficient and slow process.

Polaris is a ‘polar opposite’, since it builds a map of sorts and details all of a page’s dependencies. The data can then be used by the browser to download as many files as indicated on the page. This reduces the number of times a browser needs to ping the source before acquiring a webpage piece. The project, Polaris, is built using Javascript which has to be executed at the server side. In other words, it won’t require browser support, calibrations or updates to work. Any browser can make use of Polaris.

Researchers are saying that it’s a different approach to what others have tried before. PhD student Ravi Netravali said:

What prior tools have done with their dependency graphs between the objects on the page is make them with respect to how browsers today load the pages.

What we were able to do with Polaris is track, at a finer granularity, how these objects interact. So is one object writing some data that another then reads? Ok well then it’s a dependency. But if they’re totally doing separate things, and they don’t have any shared state, then you should be able to fetch them and handle them in parallel because they don’t depend on one another.

The team tested the browser with their tech on top 200 websites on Alexa’s list. The results were promising, with a 34 percent decreased loading times. Page loads are even faster for heavier and complex websites, while simpler websites aren’t affected that much. Polaris could indeed catch on, as all it requires is a simple implementation.