Google Engineers Argue for Changing Fundamental Web Architecture

For most of us, using the web, online email and other connected services goes on all day long, but how often do you step back and consider the subterranean technology that powers the web. Transmission Control Protocol (TCP) is one of the core components of the Internet Protocol Suite, and is fundamental to how data, media and communications travel across the web. TCP is also a core component in many messaging platforms. Now, Google engineers are calling for an upgrade to TCP, in order to speed up the web at the most fundamental level.

In case you didn't know, Google actually has a "Make the Web Faster" team, and you can get some thoughts on deficincies in TCP in this Google Code post.  In it, Yuchung Cheng writes:

"To deliver content effectively, Web browsers typically open several dozen parallel TCP connections ahead of making actual requests. This strategy overcomes inherent TCP limitations but results in high latency in many situations and is not scalable. Our research shows that the key to reducing latency is saving round trips. We’re experimenting with several improvements to TCP."

Could Google engineers boost everyone's web experience by solving a few simple latency issues? Hey, I'm all for it if they can.

Cheng summarizes a number of very specific recommendations for changes to the TCP standard, and some of them seem quite simple to implement. For example, he calls for reducing the length of connection timeouts, and you can read a detailed rationale for this via his post. He says:

"Reduce the initial timeout from 3 seconds to 1 second. An RTT of 3 seconds was appropriate a couple of decades ago, but today’s Internet requires a much smaller timeout. Our rationale for this change is well documented here."

Now, there will be naysayers who will argue against Google have any dictatorial rule over the fundamental architecture of the web. After all, that could provide an evil inroad to invasion of our privacy, they'll argue. But if you actually look at Cheng's recommendations, they seem extremely logical and worthy of consideration by Internet standards bodies. His fundamental argument is that the web we all work and play on is old in its architecture.