What is a PC?

It's a simple question -- "what is a PC?" -- yet the answer is complicated.

If we use Mr. Peabody's Wayback machine to travel to September 1981, the answer is simple. A "PC" (that is, a personal computer) is an IBM model 5150 with it's gray cover, detached keyboard (with 83 keys), and either an IBM Color Display (5153) or an IBM Monochrome Display (5151). It has an Intel 8088 processor, probably one or two floppy disk units, and a video adapter card.

At that time, that was a PC. Any other equipment was not. The PC name was strongly associated with IBM.

Over time, the concept of "PC" expanded. IBM introduced the IBM PC XT (model 5160), which meant that there were *two* models of IBM PC.

IBM introduced adapters for memory and ports. Other vendors did also. Compaq introduced their portable PC, fighting (and eventually winning) the battle for a compatible BIOS. Hercules made a video adapter that displayed graphics on monochrome displays (the IBM monochrome display adapter displayed only text).

In 1984 IBM introduced the IBM PC AT which used the Intel 80286 processor. Now there were three types of PCs from IBM, some with different processors, and bunches from other vendors. Some had more memory, some had different adapters. IBM introduced the Enhanced Graphics Adapter (EGA) with the IBM PC AT.

Through all of these changes, the two constants for PCs were this: they ran PC-DOS (or MS-DOS), and they ran Lotus 1-2-3. The operating system and that one application defined "PC". If the device ran PC-DOS and Lotus 1-2-3, it was a PC. If it did not, it was not. (And even this definition was not quite true, since several computers ran MS-DOS and special versions of Lotus 1-2-3, but were never considered to be "PC"s. The Zenith Z-100, for example.)

Moving forward to the early 1990s, our definition of PCs changed. It was no longer sufficient to run PC-DOS and Lotus 1-2-3. Instead, the criteria changed to Windows and Microsoft Office. Those were the defining characteristics of a PC. (Even in the late 1990s, when Compaq and Microsoft built the "Pocket PC", the device was considered a PC.)

Today, when we use the term "PC", we think of a set of devices. These include desktop computers, laptop computers, virtual computers running on servers, and now, with the Microsoft Surface, tablets. The operating system has expanded to include Linux (but not Mac OSX), and there is no definitive application. We use the phrases "Windows PC" and "Linux PC". Windows PCs must run Microsoft Windows and Microsoft Office, but a Linux PC needs only a version of Linux.

We have the puzzle of an Apple MacBook running Linux -- do we call this a PC? I am tending to think not. Apple's advertising and branding has been strong.

The one characteristic is that all of these devices require the user to be an administrator. The user must install new software, ensure updates are installed, and diagnose problems. This action separates a PC from a tablet. Tablets do not require the user to "install" software -- beyond selecting the software from a menu. Tablets do not require the user to be an administrator. Updates are applied automatically, or perhaps after a prompt. Network adapters do not need to be configured.

Let's take the dividing line between PCs and tablets as administration. Some might call it "ease of use".

Yet even this definition is less than clear. Apple's OSX is better at installing applications: just drag the install package to the "Applications" folder. Linux has made improvements too, with Ubuntu's "Software Center" that lets one pick an application and install it. Microsoft's Windows RT is quite close to Apple's iOS for iPhones and iPads, which are clearly not PCs.

Despite the lack of a bright line in devices and implementations, I believe that we will look back and consider PCs to require administration, and non-PCs (tablets, smartphones, etc.) to allow use without the administrator role.

So that's my answer: If you need an administrator, it's a PC. If you don't, then it isn't.

Maybe the answer isn't so complicated.