Nothing is more annoying when you are trying to load a web page on your smartphone–and it’s taking forever. It’s a situation that is increasingly common as websites become more complex with various features that slow down load time. This “web bloat” can hurt traffic as those users get annoyed and go elsewhere. But now a team of researchers from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed a tool that cuts down on load times by as much as 34%, reports MIT News.
The new tool is a server-based framework called Polaris, and it works by pre-mapping various connections between different objects on a page in order to figure out the most efficient order in which to load the objects. This is different from how browsers typically load web pages, in which they need to fetch “objects” like HTML files, images, and JavaScript source code, and then evaluate the “dependencies” between them. “As an example, a browser might have to execute a file’s JavaScript code in order to discover more images to fetch and render,” says MIT News.
The problem lies in the fact that browsers can’t see all the dependencies beforehand, which means that once one object is downloaded it might reveal that object is dependent on another, so the browser has to go back out to the web and grab that next object and so on and so forth. These increased back-and-forth trips are what slows down page load times.
But thanks to Polaris’s ability to pre-map a web page, a person’s browser can now know exactly what objects are dependent on each other and download those concurrently. Harvard professor James Mickens, a researcher on the project, analogizes the process to that of a salesperson visiting different cities. If you don’t have a list of all the cities you need to go to ahead of time you can’t plan the most optimal route, meaning you’ll need to zigzag back and forth between distant ones, lengthening your trip time.
“For a web browser, loading all of a page’s objects is like visiting all of the cities,” Mickens says. “Polaris effectively gives you a list of all the cities before your trip actually begins. It’s what allows the browser to load a web page more quickly.”
The team behind Polaris tested the framework on 200 of Alexia’s top ranked websites and the median reduction in load times was 34%. That reduction could have a profound impact on site profits. As MIT News points out, for major sites like Amazon, every 100-millisecond delay cuts its profits by 1%.
For now, Polaris will be open sourced, but research team member PhD student Ravi Netravali says “our end goal is absolutely to have one of these major browsers adopt it. And of course web servers doing the same.”
This story has been updated to clarify the role of the Polaris system.
