In his keynote speech at Macworld Expo in January 2001, Apple CEO Steve Jobs declared that the PC was not dead, just boring--and that Apple would cease to make their PCs boring by making them ``hubs'' for people's ``digital lifestyles''. To illustrate, he showed a PC circled by a posse of mostly portable devices that record and play back sound and video: CD player, MP3 player, DVD player, camcorder, etc. The idea was to turn the PC into a personal entertainment storage and production appliance. The term ``personal'' was his. The term ``appliance'' is mine. It's hard to reconcile Steve's ``art'' (and margins) with a word we associate with Hoover and Mr. Coffee.
But, it's beginning to become clear that appliance is to hardware what application is to software. Both tend to have singular purposes. I don't know about you, but ``hub'' seems like it has a rather singular purpose to me. Rhetoric aside, it's smart positioning on Steve's part. The more connected everything gets, the more purposeful every connected device becomes.
In the same keynote Steve reported on the progress of Apple's new operating system, OS X. The X stands both for the numeral ten and the last letter of UNIX. The OS itself is called Darwin, and it's basically BSD with a Mach kernel. There's a lot of proprietary UI and development software that Apple layers on top of Darwin for use in its boxes, but it's Darwin that makes OS X not only native to the Net but also to the guild of infrastructural masons we call the UNIX community. Apple opened the source code to Darwin almost two years ago, but the OS wasn't very interesting to the UNIX guild as long as Apple's source code license basically restricted all the useful benefits to Apple.
But then OS X Beta came out in the Fall of 2000, and lots of builders went at it. By the time Steve made his speech in January, Apple had a new source code license that was functionally open: anybody could take Darwin and do what they wanted with it. Thus Apple came into compliance with reality, which is infrastructure--the buzzword to bet on.
For context, let's go back in history to the PC Era, when computing was personal. Back then we came up with another structural buzzword: platforms, which were the combinations of operating systems and hardware on which applications ran. Now those platforms have to run on the Internet. And as more new applications run on the Net, platforms become extensions of Internet infrastructure. This is what happened to Apple. It's also what happened last year to IBM when it declared its love for Linux. In both cases management agreed to comply with their own engineers and the larger guild of UNIX masons who are busy building the world.
Builders of all kinds, whether of houses or software, like to work with commodities. And the fact is, we always build infrastructure with commodities, whether we're talking about 2 x 4s and sheetrock or TCP/IP and Linux.
It is rapidly becoming clear that Linux is commodity infrastructure. That's why it's no longer much to brag about. It's like saying, ``this car is made with metal.'' Of course, Linux and other free and open UNIX variants (like Darwin) aren't the only commodities. So are 32-bit processors for every application/appliance. Linux is rapidly adapting to nearly all of them, thanks in large part to the embedded Linux toolchain vendors. Countless other components and connecting technologies are also commoditized. So the supply infrastructure is making it possible for more and more companies to get into the networked appliance business and for established companies to create and iterate prototypes.
One interesting effect of this is the emergence--like crystals out of chemicals in the right conditions--of appliances that are themselves infrastructure. This is what Cobalt pioneered by taking a highly commoditized business--Linux-based servers--and productizing them as ``server appliances''. They took commodity infrastructure and productized it, giving it appliance-like ease of configuration, installation, operation and, of course, sales. It was easy for Cobalt to sell a database-in-a-box to ISPs who could slide it into a rack and in turn sell it--again, as a product--to their customers.
Sun Microsystems understood this, which is why they paid $2 billion US for Cobalt. (For more, see our interview on page 8 with Cobalt CEO Stephen DeWitt, who is now vice president and general manager of Sun's Server Appliance Business Unit.) They saw how the world of network infrastructure--Sun's native habitat--was going to turn increasingly into an appliance business.
In Invisible Computer, Don Norman predicted that buggy, crashy, hard-to-use PCs would gradually be replaced by functionally purposeful and easy-to-use appliances. Now it's looking like he was right. Even about PCs.
Doc Searls is senior editor of Linux Journal and a coauthor of The Cluetrain Manifesto.
Doc Searls is Senior Editor of Linux Journal
Getting Started with DevOps - Including New Data on IT Performance from Puppet Labs 2015 State of DevOps Report
August 27, 2015
12:00 PM CDT
DevOps represents a profound change from the way most IT departments have traditionally worked: from siloed teams and high-anxiety releases to everyone collaborating on uneventful and more frequent releases of higher-quality code. It doesn't matter how large or small an organization is, or even whether it's historically slow moving or risk averse — there are ways to adopt DevOps sanely, and get measurable results in just weeks.
Free to Linux Journal readers.Register Now!
- Three More Lessons
- Django Models and Migrations
- August 2015 Issue of Linux Journal: Programming
- Hacking a Safe with Bash
- The Controversy Behind Canonical's Intellectual Property Policy
- Secure Server Deployments in Hostile Territory, Part II
- Shashlik - a Tasty New Android Simulator
- Huge Package Overhaul for Debian and Ubuntu
- Embed Linux in Monitoring and Control Systems
- KDE Reveals Plasma Mobile