The IBM PC isn't a standard for the industry at all - it's a standard for IBM, and a shifty target at that. IBM never set out to create a standard. They birthed a well-conceived market-grabber that bridged the gap between the adaptable but non-business Apple II and the workaday, dull world of CP/M computers.
Well then, fine. . . . What's the big deal about standards anyway?
Imagine a nationwide electrical power system that supplied different voltages to each region of the country. Or required differently shaped plugs in New York than the ones in California. Electrical appliance makers would spend a lot of their research and development resources designing alternative motors and plugs to meet the requirements of each nonstandard power supply. That's what goes on in many parts of the world and why you need a transformer to make your hair dryer work in Europe.
Standards are developed (or as often, just evolve) in almost every industry for solid reasons: to make components interchangeable in completed products; to focus design efforts on innovation rather than adaptation and ultimately, to keep the cost of products affordable. These reasons all tend to benefit the consumer.
Success in the personal computer business is a curious chicken and egg situation. Buyers are sophisticated enough to realize they shouldn't consider a computer that doesn't have a broad base of available software; software companies won't create new products for computers that aren't being bought in large quantities.
Some young genius dreams up a new way to organize information. Then, because programs have to match up precisely with a computer's operating system and hardware features, a decision has to be made about which machine or generic group of machines is to be the vehicle for the embryonic product.
Imagine the software developers who were putting the finishing touches on programs for the Osborne I, just when the announcement was made that Osborne Computer had filed for Chapter 11 protection under the bankruptcy act; they didn't lose all the work put into the software designed for this fairly non-standard machine but much of the potential return on their time investment has been lost.
Enter the "IBM PC standard" dilemma. The standard everyone talks about is technical in nature but simple to understand - it comes down to whether a given program can run on a variety of computer brands without modifications. If the IBM PC was a standard in that sense, there wouldn't be all the discussion there is about whether the Compaq or TI Professional or Corona computers are "truly compatible" with IBM's offering.
The reason the compatibility controversy rages is because IBM set out to design a machine that was unique and specifically their own, just as Apple did with the ubiquitous Apple II or Radio Shack with their line of TRS-80 computers.
Few people have ever accused IBM of a charitable outlook toward their competitors. When they announced the PC in 1981, they did take the unprecedented step of releasing Voluminous technical information that permitted others in the industry a chance to design add-on components, software and even competitive computers - with one important exception.
In the heart of the IBM PC lies a section of Read Only Memory (ROM) that is a specialized chunk of circuitry to handle input and output routines and parts of the BASIC language interpreter, among other things. This 40,000 byte ROM area has proven to be a stumbling block to hardware and software compatibility as it is patented and undocumented. This was not an accident; IBM could have made the contents of this critical area as publicly available as they did other PC internals, but they chose not to.
Other computer manufacturers have gotten around this unknown territory by analysis of how the PC and specific software function, accomplished by exhaustive testing. This costs much time and money, commodities that IBM's competitors have in limited amounts, especially compared to the immense resources IBM can bring to bear (see box).
Is IBM not being fair to the little guys? Probably not, but that isn't the issue. By slavishly following the IBM party line, the rest of the industry has put themselves in the position of being jerked around by any design whim (good or bad) that IBM decides is useful to their computer line or even their stockholders.
As noted on the opposite page, operating system changes have already caused trauma for the compatible vendors. But what if IBM decides to change the size and content of the ROM that creates so many compatibility problems, or perhaps load some totally new operating system into this obscured memory location? In laying out the internal architecture for the PC, IBM allocated 256 KB of the total available memory for ROM and currently uses only 40 KB. That should cause some nightmares for their competitors, if it isn't already doing so.
This isn't intended to be just an anti-IBM polemic. I think the personal computer industry needs more than one vendor, no matter how skilled. I think consumers need choices ~ in hardware, pricing and' innovation.
Unless the non-IBM computer industry gets its act together and develops its own standards, not just a de facto approach caused by immediate adoption of any change in PC specifications, most hardware makers will find themselves in an increasingly difficult competitive position; many will find themselves out of business. And as that happens, we will all be the poorer for it.