In the Works: Start at the standard
(IDG) -- Today, standards are at the fore of many technology discussions. But what really constitutes standards? It's a question that sometimes confuses even the people who write them. And this confusion can affect vendor plans and more important, your buying decisions.
As chairman of the Internet Engineering Task Force, my part of "In the Works" will focus on emerging standards. So I figured I should explain what makes a standard and what you should keep your eye on.
Simply put, a standard is an agreement on how networking products interact. You can liken it to everyday life. For instance, when we travel, we generally use the currency of the country we are in because it is the standard for measuring and exchanging value in that locale.
Also, consider light bulbs. No matter who makes them, they still screw in the sockets the same way.
Such conventions are useful because they simplify our lives, but they are not magical. If we started making light bulbs that screwed in counterclockwise, life wouldn't end, but it would be more difficult.
The same is true for networks. We use particular protocols, such as PPP, IP, TCP, SMTP or HTTP, because they are useful and solve a problem. We also use them because others agree to use them, and we can therefore communicate between different vendors' implementations of these various protocols and procedures.
The Internet Engineering Task Force has two kinds of documents that detail these Internet-related protocols and procedures. We call them Internet drafts and requests for comment (RFC).
When a proposal for a standard is submitted to the IETF, it becomes an Internet draft, which means that it is a temporary paper under discussion by our working groups. It may be revised several times, may ultimately be discarded, or may eventually be published as an RFC. An RFC is a permanent, numbered document that is entered into a library where engineers and vendors can use it as a technical reference.
However, these RFCs are only suggestions for how to do something. This contrasts the documents of other standards groups such as ISO or ITU-T, which consider their documents the right way to implement a technology. For instance, they specify the exact use of the pins on a connector. You would never see the IEEE come up with something like RFC 1149: "IP on Avian Carriers" or publish bad ideas and poetry created by its participants as official documents.
So, as you can see, RFCs aren't always standards and the submission of a document doesn't automatically make it a standard.
Yet, vendors frequently refer to these documents as approved standards when trying to sell you their products. They may boast "my equipment is good, because my protocol is described in some RFC," which has become obsolete or which most people don't implement. I'll stop short of saying vendors try to create confusion, but if it helps them boost sales, some are happy enough not to clear things up.
So, "caveat emptor." If interoperability is important for your network, always test statements by vendors that their wares are actually standards-compliant. How? Start by checking the RFC index here. See whether the IETF has marked the document as a proposed standard, draft standard or standard. Then, make sure that competing vendors and packages that must interoperate use the same protocol. After all, a standard is an agreement of what to make and use, and if the other vendor is making and using something else, it isn't a very strong agreement.
Internet pioneer urges transition to new Net protocol
RELATED IDG.net STORIES:
Information superhighway: Construction ahead
Internet Engineering Task Force
|Back to the top||
© 2001 Cable News Network. All Rights Reserved.|
Terms under which this service is provided to you.
Read our privacy guidelines.