The following text is copyright 1996 by Network World, permission is hearby given for reproduction, as long as attribution is given and this notice is included.

When a part is greater than the whole

One would think that the advantages of standards would be well understood by now but this seems to be a lesson that must be learned anew every few years.

The American Heritage Dictionary has a 10 part definition of "standard" used as a noun and a 6 part definition when used as a verb but I'm using it in the context of a point of reference against which individuals are compared and evaluated." I'm also specifically referring to standards that result from a fair and open process and not vendor produced standards or those resulting from closed consortia. In the networking business, standards generally refer to the physical plugs used to interconnect devices or to the format of messages exchanged over a network and the actions that result when a specific message is received.

Clearly standards benefit consumers. The fact that there are more than a 100 different companies making equipment that conforms to the 10BaseT Ethernet over twisted pair standard means that prices are low and extra features are plentiful. Since the devices must meet the 10BaseT standard, these features do not inhibit interoperability, but instead provide for increased port density, better manageability, the addition of frame switching functions and other things that may be used by the vendors to differentiate their products from those of the competition.

Some vendors seem to have a hard time understanding that standards are a good thing for vendors also. These vendors create their own technology or add proprietary extensions to standards-based technology. They think that because their product is now "better" people will buy that one instead of some other vendor's products. Over the short run these vendors may even be right but over the long run 100% of a small market is not as big as a part of a large market.

There were a handful of vendors that offered different ways to run Ethernet over twisted pair wiring before the 10BaseT standard was adopted. Together these vendors sold less product in a year than the 10BaseT product that is now sold every few days. The advent of standards facilitated the rapid expansion of the market. As it turns out, some of the early players did not prove to be efficient or flexible enough to compete but those that were are far larger than they ever would have been selling proprietary solutions.

Unfortunately we are now starting to see some of this same thinking in the developers of applications to run over the Internet. Particularly in the area of web browsers, vendors are starting down their own paths, sure that a pot of gold lies that way.

I think this is short sighted. The potential of commerce over the Internet will be realized when every user's browser can talk to every merchant's server. If it requires different browsers to talk to the servers of different merchants, it means that the merchants will have to include browser distribution in their support functions. This would be like requiring users to get telephones with different touch tones in order to use different voicemail systems.

There are a few trillion dollars in annual business that could be done over the Internet once some of the supporting technologies that enable secure and reliable message exchange are in place. This potential will not be even vaguely realized unless these supporting technologies are standards.

disclaimer: Although some claim that Harvard is a standard unto itself (or at least to itself) the above opinions on the power of standards are my own.