“We view the Internet as the fourth desktop operating system we have to support after Windows, MacOS, and DOS.” That quote was from an executive at McAfee, and DOS gives it away that it was spoken back in 1996.

With the announcement that Google will develop a quick-start operating system by next year for instant-on netbooks, I thought it might be interesting to take a trip down memory lane and remind us how we have gotten to the point where the browser has become the next OS, and probably now moving into first place rather than fourth.

Of course, the smarmy retort to Google’s announcement is that we already have a quick-start, ultra-reliable Web OS, it is called OS X and my MacBook takes about five seconds from when I open the lid to when I can be surfing the Web. Unlike many Windows PCs, I don’t have to have a degree in advanced power management techniques with a minor in spam and virus prevention to get this to work.

But let’s go into the WayBack Machine to the early 1990s and see the context of that McAfee quote.

The first collection of Web browsers literally weren’t much to look at, because they only displayed characters and basically just a page of hyperlinked text. This was the then-popular Lynx that was initially designed back in 1992 for Unix and VMS terminal users (that was back when we called them that). Think about this for a moment: this was pre-iporn, pre-IPO Netscape, pre-real Windows — when the number of Web servers was less than a few hundred. Not very exciting by today’s standards.

Then Microsoft got into the game, and things started changing. With the introduction of Windows 95 we had the beginnings of a graphical Internet Explorer, which ironically was licensed from the same code that Netscape would use to create their browser (and eventually Firefox). Windows 95 came with both IE and Windows Explorer, and the two were similarly named for a reason: browsing pages of the Web was the beginnings of something similar to browsing files on your desktop. Things didn’t really get integrated until IE v4, which came out about the same time as Windows 98, and they were so integrated that they begat a lawsuit by the Justice Department. At the end of 2002, Microsoft was legally declared a monopolist and had to offer ways to extract IE from Windows going forward for users who wanted to install a different browser.

During the middle 1990s, we began to see better support for TCP/IP protocols inside the Windows OS, although it really wasn’t until the second edition of Windows 98 that we saw Microsoft improve upon the browser enough that they could include it as part of their Office 2000 product. Before then, we had separate drivers and add-on utilities that required all sorts of care and feeding to get online, in addition to using AOL and Compuserve dial-up programs.

As an example of how carefully integrated IE was with Windows, when Microsoft released IE v7 along with Vista, initially you needed to verify your license of Windows was legit before you could install the latest version of IE on earlier operating systems. That restriction was later removed.

And lately Microsoft has announced its next version of Office 2010 will have even further Web integration and the ability to create online documents similar to the way Google Docs works. Google Docs is an interesting development of itself, because now documents are stored outside of the desktop and managed through a Web browser. As long as I have an Internet connection, I don’t need any software on my local machine to edit a document or calculate a spreadsheet.

So what is the real purpose of an operating system? Originally, it was to manage the various pieces of your PC so that your applications could talk to your printer or your hard drive or display characters on your screen without having to write low-level programs to do these tasks. Three things have happened since the early PC era:

First, as the Web and cloud computing became more powerful, we stopped caring where our information is located. In some sense, having documents in the cloud makes it easier to share them across the planet, and not have to worry about VPNs, local area network file shares, and other things that will get in the way. And we even have cellphones like the Palm Pre that have a Web OS built in, so that applications don’t have to be downloaded to the phone but can run in the cloud. At least, when developers will finally get their kits to build these Pre apps later this summer.

Second, as the desktop OS matures, we don’t have to worry about the underlying hardware as much because that hardware has gotten more generic and the OS has taken on a bigger role (to match their bigger footprints too). Although printer drivers are still scarce for Vista, and 64-bit apps aren’t as plentiful, for the most part we don’t need a “thick” desktop OS. Yes, there are enterprise apps that need the OS, and some that need a specific version of Windows too, but most of our computing can be done without really touching much of the OS.

Finally, the browser is the de facto Windows user interface. Perhaps I should say the browser plus Ajax or the browser plus Flash. But most applications that were formerly client/server now just use browser clients, or run inside a browser with minimal desktop downloads. This has been long in coming, but now Rich Internet Applications can be considered on par with local Windows and Mac ones.

So here we are, at the dawn of the new Google OS. We have come full circle: from the green-screen character mode terminals of the mainframe and Unix era to the browser-based Webtops of the modern era. This doesn’t mean that Windows 7 or 8 or whatever will become obsolete. Just less important. And given the multiple billions of dollars that Microsoft has made over the years from Windows (and let’s not forget dear old DOS), you can imagine that there are some nervous folks up in Redmond these days.