Google's speed push more best practices than revolution

Google's attempt at rallying developers, ISPs and others to increase the speed at which the Internet operates focuses less on rip-and-replace and more on a set of best practices that includes compression techniques, optimized JavaScript code and Web graphics, and reductions in HTML file sizes.

The company kicked off a campaign called “Let's make the Web faster” aimed to develop changes that will speed up the Web. The company suggested that needed changes could potentially include replacement of cornerstone protocols such as TCP/IP and HTTP.

Google co-founder Larry Page wants the Web to be comparable to flipping the pages of a magazine and the company is trying to lead a need-for-speed charge.

“The belief is that human beings around the 100th millisecond, the 10th-of-a-second range, can start to perceive change in time,” says Bill Coughran, senior vice president of engineering at Google. The company says that time can add up.

That result is that it doesn't take much for users to think the Web is slowing down. But on the other hand, a few alterations here and there can save precious time and increase speed.

The changes to the Internet including Web page development and protocol tweaks will take some time, Coughran says.

“We will have to advocate for it, educate users, other companies, and work toward [a faster Web],” he says. “This will be a series of difficult advocacy steps over a long period of time. ”

Initially Google is focusing on a core set of best practices the company has developed over the years.

The list includes 12 suggestions, including the use of HTTP features such as caching to speed the loading of Web pages.

Google advises users to tactically set HTTP caching headers, which are sent by Web servers to specify how long a resource is valid and when it last changed. Google recommends setting controls for expiration and maximum age for a Web page resource to a year in the future.

The company also recommends that Web developers explore gzip compression, which works by finding similar strings within a text file, and replacing those strings temporarily to make the overall file size smaller. Google says gzip compression works well on the Web because HTML and Cascading Style Sheets files usually contain plenty of repeated strings.