Internet access and telco usage patterns

Eric Woodward ejw at
Sun Jul 7 02:54:16 UTC 1996

At 10:51 AM 7/5/96 -0700, Hong Chen wrote:
>As a matter of fact, it is quite doable. Aimnet developed a roaming
>server (check that allows international ISPs to 
>use each other's network to provide dialup services. A group of 
>ISPs have joined a consortium GRIC (Global Reach Internet 
>Consortium)  lead by Aimnet. The roaming server is based 
>on Radius protocol. 
>A telco company can install modems and route the authentication to 
>the specific ISP for authentication. 


I looked at this doing this about a year ago but the major stumbling block 
was that if ISPs share the authentication responsibility using distributed
RADIUS, they have the capability of keeping each other's passwords for the
user's that used the global access service.

Also, a service you likely know about, started up around the same time in 
Vancouver, where I was living at the time, called GeoAccess
(, was going to target this idea much more
aggressively than I (and plus I did not feel like competing with him in
particular), and decided on the model on centralized authentication, 
effectivele becoming a worldwide access ISP without purchasing a single modem
or terminal server.  But even ISPs participating in his "network" can log
the entered passwords.

Telephone companies might have a problem with the legal ramifications of this
"roaming" service.

>I just came back from Montreal INet 96 last week and a new roaming 
>IETF group will be started. We are working on the IETF draft 
>for the roaming and stay tuned. 

Please let me know what the name of working group is, and perhaps take this
to private email.  I would be very interested to know how the password
access problem is worked around, or at very least, rationaly pushed aside, 
and even contribute.

Eric Woodward.
ejw at

More information about the NANOG mailing list