Wilfried: I'm thrilled to see your response. The reason I proposed this change was to align it with the IP CIDR object. My hope was we could then leverage the IP CIDR work to quickly implement the CLNS. So, I am looking at implementation at RIPE NCC and MERIT. I would like to expand prtraceroute and whois servers to use the CLNS. The host.txt option is certainly open to question. I suggest we talk about the benefits of the proposal without this section. To discuss the "host.txt" section, I suggest we query the noop@merit.edu tuba@merit.edu lists to determine the distribution of the DNS NSAP object. If it is far enough along, then we should can this. Otherwise, it may be a very useful "bootstrapping" function to support. If you agree, I will send notes to this and the rare-clns-wg4 mail list. In the implementation these objects could be kept in a different sequence of files that could be easily cut when the DNS NSAP object is widely deployed. Again, many thanks for such a thought response! Sue Hares