The following text is copyright 1995 by Network World, permission is hearby given for reproduction, as long as attribution is given and this notice is included.

Making the Signs and Painting the Lines

By: Scott Bradner

Well, in a couple of months NSF will be out of the backbone business. Under the current schedule, NSF will tell ANS to turn off the NSFNET by the end of April and by then all of the old NSFNET-connected midlevel networks will be happily exchanging their bits over commercial network service providers. (Most of them are already, to one approximation of happy or another.)

So, NSF will be out of the, to use a rather over used and abused analogy, highway building line of work. For a while, they will help some of the midlevels pay part of their data connection fees, sort of like mom and dad helping out with the rent for the kid who has moved out of the homestead. Some of the kids are a bit shaky on their feet, others are doing rather well, but they don't refuse the parent's help, but this, too, shall pass away.

There is more to building this data highway than just laying down the roads. Someone has to set the rules, put up the signposts, hand out the license plates, (how far can I stretch this?) and disambiguate the town names (that's far enough!).

These functions on the Internet include allocating addresses, assigning domain names, assisting confused users, and providing routing information, along with developing and defining the standards required to ensure continued interoperability.

Up to now these administrative infrastructure functions have also been funded by various arms of various governments. The U.S. name and address registry (known as the InterNIC), other national network information centers (NICs), the Internet Assigned Numbers Authority (IANA) and the new Routing Arbiter have been, and are now, funded by specific grants from a number of U.S. federal government agencies or, for the non-U.S. organizations, local governmental funding. The Internet Engineering Task Force which is, the Internet standards developing organization, is supported in one way or another by the NSF, the Corporation for Research Initiatives, the Internet Society and fees paid by meeting attendees.

As the Internet matures, (if that term can be used about an environment where, as I write this, I received a invitation for some automagic self configuring electronic magazine on a mailing list devoted to electronic data interchange (EDI)), there is less and less justification for governments to perform these functions. Alternate reliable funding models must be developed to support these needed functions.

Service fees would seem the most likely, a fee to register a domain name for example, or even a fee to keep one registered. (It was erroneously reported in these pages last week that the Internet Society had decided to charge for IP Addresses. While the suggestion has been made by some people, the Society has not the authority to impose any such charge nor has it developed any plans to propose that others do so.)

This whole area of the funding of the administrative infrastructure is one that will take considerable discussion before any definitive plans can be made. This discussion will take place in many forums within and outside of the government agencies that are curently providing the funding. These will be long and hard discussions, and they are overdue.

Desclosure: The Commercial Internet Exchange (CIX), FARNET, Inc. and the Internet Society are jointly sponsoring a forum in April where discussions on these and other Internet operational issues will be held. I am a co-chair, along with Bob Collet and Jim Williams, of this meeting so I may appear to be focused a bit more on these issues than I otherwise might. For information of the meeting take a look at

Disclaimer: The opinions expressed above and the semi-plug for the meeting are mine alone and I have no idea if Harvard, the Internet Society (individually or collectively) or my co chairs share any of these opinions.