Re: BGP

Damien T. (damient@livewire.comsec.net)
Sun, 25 May 1997 14:31:32 -0700 (PDT)

At 10:56 PM 5/24/97 -0500, Jake (not the snake) wrote:
>On Sat, 24 May 1997, James D. Butt wrote:
>> I think that it is very important to understand what you are
>> implementing. The RFC's are not bad reading in fact I would recommend
>> reading all the RFC's about all of the protocols you are using....
>
>It is also important to understand that the RFC's are just that. RFC's.
>They are not necessarily the standard. Way back in the olden days, when I
>was learning about routing and subnetting, I placed too much faith in the
>older RFC's I read, assumeing that they were the "law". They have since
>been either superceded, or don't apply anymore.
>
>Id say that it is a good STARTING place but just becuase you read
>something in an RFC, does not make it come to pass.

Actually, I generally look for someone else's written work that is based on
the RFC as a starting point. As MegaZone is no doubt familiar, people often
grasp a concept easier if the salient points are distilled from the detailed
technical nuts and bolts and then discussed in a practical, real-world sense
rather than a more abstract fashion. The evolution of Livingston's
configuration manuals are a case in point, many of us still having the old
spiral-bound versions laying around.

For me, the RFCs are the "last word." The final arbitrator on how things
"should" work.

I agree with you in that some things seem to be drifting away from RFC
compliance at a much more rapid pace than ever before. Two factors that
have been bothering me in recent weeks in terms of feeling like the Internet
is drifting away from standards (other than Microsoft implementations) are
the assorted weirdness and turf protecting regarding domain names (now is
the time for Karl D. to chime in <g>!), and the "de-peering" of UUnet and
MCI in their quest to apparently become paid transit providers (anyone
experienced any fallout from this?)

Since I think I've raised more (paranthetical) questions than I should have,
I'll end this here <g>!

Damien