With Chrome OS, Google will blur the line between the browser and the OS completely, says Ramesh Iyer, head of worldwide business development for mobile computing at Texas Instruments, which is one of Google's partners on the project. "The browser is your operating system," Iyer says. "The browser is your user interface. The browser is the mechanism from which you launch applications."

Streamlining the OS to focus on the Web, Iyer says, will allow devices to run more powerful programs with less powerful processors. By keeping processor requirements low, new devices could use less battery power and stay lighter. Texas Instruments, for example, is working with Google to integrate the Chrome OS software with its OMAP 3 multimedia applications processors, creating a system that could be easily installed in netbooks and other devices.

Giving Web applications deeper access to the underlying kernel could make it easier for Web developers to provide better functionality and a better user experience, says Jared Spool, founding principal of User Interface Engineering, a consulting firm based in North Andover, MA. When Web applications such as Gmail and Google Maps first appeared, Spool says, the software engineers who built them had to do a lot of hacking to create the appropriate levels of interaction. "When we went from the desktop to the browser, we took a huge step backward," Spool says.

With an OS tied closely to the Web, Google can introduce sophisticated resource management tools that will allow Web applications to run much more smoothly. A major role for the OS is allocating memory to applications and adjusting it as their needs change. A big problem with interactive Web applications to date has been that browsers didn't have efficient ways to adjust the memory assigned to different Web pages. The Chrome browser has already improved the situation, Spool says, and he expects the OS to go even farther. He says that this will allow more powerful Web applications that run more smoothly on the new OS.

But building the Chrome OS won't be as simple as sticking a browser on top of the Linux kernel, Spool says. The browser version of Chrome relies on the underlying OS's user interface, for example. Features such as scrollbars come from the OS, not the browser, so Google will need to build all of this from scratch, and even simple things will require significant time and effort.

The Chrome browser also lacks the drivers needed to power any external devices, such as printers or iPods. Texas Instruments' Iyer envisions a new way that Chrome OS could address this problem. "Wouldn't you rather have a printer connected in the cloud?" he says. As more devices, including cameras, printers, GPSes, and so on, become able to connect to the Internet in their own right, the concept of a Web interface between a user's computer and the device comes closer to reality. "This is the holy grail of the Internet," Iyer says.

Pichai and Upson have also said that Chrome OS will support all Web-based applications automatically, and that new applications written for Chrome OS will run "on any standards-based browser on Windows, Mac, and Linux." While Web application development has exploded in recent years, this might also introduce limitations, preventing the user from accessing interesting applications developed in programming languages not intended for the Web.

However, Google may have a solution for that too. The company is working on an experimental project called Google Native Client that would allow code written in non-Web languages such as C and C++ to run securely in the browser.

Chris Rohlf, a senior security consultant for Matasano Security, which has been involved in testing the implementation of Native Client, says, "It could be Google's secret weapon when it comes to Chrome OS, because it would allow developers to extend that platform with things like video and graphics without having to wait for Google to implement any of that."