Last Updated Nov 19, 2009 9:09 AM EST
Larry quotes John Gruber at Daring Fireball when he says that the "biggest argument for having 35 operating systems instead of three big ones (Windows, Mac, and Linux), is the incompatibility argument":
The argument goes like this: It was a mess when there were a bunch of operating systems. Gruber's reply:Sure, "fun" -- and a complete pain in the rear if you're a user or a developer. Will your phone OS and the applications you may come to depend on be around tomorrow? Not that your current choice of operating system disappears. All that need happen is that you decide on a different device with a different OS and different set of apps. Poof! You're starting from scratch, because the more the number of OSs, the fewer devices available per OS. User choice effectively shrinks and doesn't grow.First, it may have been a mess, but it was a beautiful mess. It was glorious. It was fun. The Apple II, the IBM PC and DOS, Commodore, Atari, Acorn. The TI-99/4A.Gruber argues that we need a beautiful mess again in the PC market. I agree. After all, it's been fun watching the smartphone industry, which is arguably an operating system mess. There's Android, iPhone, Research in Motion, Palm, Windows Mobile and Symbian. There's no clear winner yet--and it's fun. The OS scrum in the mobile industry is a beautiful mess.
The phone developers I've spoken with talk about the time and energy they must put into porting, getting their software onto multiple platforms. This is the equivalent of an overly complex supply chain, taking 15 steps to get goods to the customer when you should only have to take three. As a result, more time and money goes into busy work, leaving fewer resources for what's really important: innovation and development. Even now there are many software packages that will run either under Windows, say, or Mac, but not both. That means there's an extra user cost in moving from one platform to another.
But I think Larry's argument really falls apart when he gives what should be its ultimate bit of support:
Gruber's argument to counter all the hand-wringing over a world with a bunch of operating systems: The Web is the glue that will bridge these operating systems. Many of the incompatibility root causes--file formats, various CPUs and storage set-ups--have been solved. Simply put, the industry is better equipped to allow 1,000 operating systems to bloom.Here's a slightly different way of phrasing the same thing. Instead of having multiple operating systems, he's depending on everything running on the Internet and using web interfaces. But that's an argument for the browser as a single super-OS. In other words, freedom comes when you can use any operating system and still get what you want, because all the important stuff starts running independently of the operating system. But then, where's the OS innovation? Does it matter? No, what matters is innovation in browsers, and you can see that already in how IE on Windows loses market share to Firefox and other entries. Effectively, you're down to one OS.
If anything, welcoming Google's Chrome OS is actually welcoming the advent of the anti-OS, where the services exist to let you run the browser and to let web-based applications get whatever local hardware support they might need. In other words, the tech industry must learn to think small, with PCs needing to abstract support through APIs rather than operating systems. In that light, perhaps what is necessary is not the multiplicity of OSs, but their destruction and dispersal. What will be important is not the OS, or even the browser, but the delivery of services. So forget the distraction of OS diversity and get it where it really counts.
Image via stock.xchng user speculator, site standard license.