The highlights from the speech were rather interesting, particularly the comment about timeliness of standards. ANSI is still working on standardizing the 5.25-inch format as of today. How long before they will have a released standard?

Standards must be timely, as in within one year to 18 months of initial request for standardization. The proposed standard must be promoted as soon as possible to disseminate information to the widest possible audience. Implementation of the standard must not unduly impact the cost of my product. i.e., if the cost impact is 1-2 percent it will be adopted, 10 percent and it won't. The number of competing standards that answer the same problem with a different solution must be at a minimum or non-existent. In other words either the solution has to totally supersede what I am doing today or I need justification as to why this solution is better than another one.

Most of the standards we use are updated regularly. The standards that have not been updated fall into two categories: (1) Old standards where only companies using legacy hardware care about the standard. These customers typically are more sensitive to price than quality. (2) Active standards where the using industry has effectively updated the standard by requesting additional capabilities that have not (and may never be) written into the standard.

A Good Standard

• A standard that answers a real problem as succinctly as possible.
• A standard, that if written to solve a particular problem, expires at a set date.
• A standard that is well promoted and supported by its industry.

Documents require a "900-pound gorilla" to become standards. Specifications that are written and either not implemented or ignored are not standards, no matter what the title. If the industry "900-pound gorilla" adopts a specification, it becomes the "standard."

As an example: prior to the breakup of AT&T, they controlled the telephone market in the US. When AT&T selected the four-pin modular connector it became the standard. Period. Ditto, when IBM selected the 15-pin high-density D-sub for computer monitors in 1986.

But when there is no "900-pound gorilla" to select a way forward, the resulting confusion can delay or sidetrack the adoption of standards and truly inhibit progress. Companies will "wait for the market to decide" before committing resources. If enough companies wait, nothing happens.

The computer industry is one that is notorious for this problem. No company wants to be first, but no company can afford to be last. A real sword of Damocles decision. Many companies have guessed wrong and they are now in the dustbin of history, which makes the surviving companies that much more reticent to make a decision to back one standard versus another.

But this is only one lone wolf's opinion and does not represent my company.

Don Chambers

JOIN IN!   We'd like to know how you feel about standards. Please e-mail me at the address below. Tell us who and where you are, and what position you hold. Then, kindly answer the following questions:

1. What kinds of standards are most important to you in your work?
2. How do you use them?
3. Are they updated on a timely basis?
4. What is YOUR definition of a "good" standard?

We will print a representative sampling of the answers we receive, and, unless notified otherwise, we will provide attribution. Please let us know if you do not wish to be identified and/or to have your employer's name listed if you're among the people whom we quote.

Thanks,
Jaren Green, Editor
jaren.green@ihs.com

Copyright © 2001. All rights reserved. Global Engineering Documents is a trademark and IHS and its logo are registered trademarks of Information Handling Services Inc. All other logos are registered trademarks of their respective owners.   Privacy Statement.