Jim, Most end applications are concerned with receiving the data in the payload of a packet, and that the headers of the packet remain intact to the extent that the packets are delivered to the proper destination. Did you write an application that uses TOS bits on an end-to-end basis, and now you've found that changing TOS bits within the network is going to break your application? What compelling event brought you to your computer and prompted you to start this rant on NANOG? The S/N ratio _was_ actually increasing... -rb

From: "JIM FLEMING" <jfleming@anet.com> To: "Jesper Skriver" <jesper@skriver.dk>, "Daniel Golding" <dan@netrail.net> CC: <edelman@law.harvard.edu>, "!Dr. Joe Baptista" <baptista@pccf.net>, <klensin@jck.com>, <mcade@att.com>, "Bellovin,Steven M (Steve) - ALRES" <smb@att.com>, <nanog@nanog.org>, <karl@cavebear.com> Subject: Re: Does AT&T provide IPv4HT ? Date: Tue, 21 Nov 2000 15:18:36 -0600

I think we are in agreement. I am assuming that people will want to "touch" the TOS byte when they "create" a packet. I am also assuming that NANOG, as a group, would want to support "end-to-end" Internet services. That seems to imply that the NANOG people would endorse the notion of "what goes in comes out". That arrangement encourages "stability". It seems odd that we have all of these ICANN people running around telling everyone that the "stability" of the Internet needs to be preserved, and then on the other hand, people are flippantly claiming to have total disregard for the contents of packets as they transit the IPv4 "cloud". To top it off, some companies apparently want to be seen as "proud" or "cool" for changing the contents of customer's packets. ...and then want to be called...."real operators"...go figure...

Jim Fleming http://www.unir.com Mars 128n 128e

_____________________________________________________________________________________ Get more from the Web. FREE MSN Explorer download : http://explorer.msn.com