ICANN/DNSO
DNSO Mailling lists archives

[ga-full]


<<< Chronological Index >>>    <<< Thread Index >>>

[ga] Re: How do you "rebid" a cartel ?


----- Original Message ----- 
From: "jefsey" <jefsey@club-internet.fr>
> 
> 1. you never replace anything, you build in parallel and outdate the 
> previous solutuion. 

Jefsey,

I totally agree, we have to be practical. People that have been designing
telecommunication protocols for the past (too many) years, are fully aware
of the need and benefits of working "in parallel" with existing systems to
evolve and migrate. That requires careful planning for evolution. Revolutions
do not happen (often) unless there is a real paradigm shift or jump. Equipment
does not get upgraded overnight. Even when equipment is eventually
upgraded, the new has to often run in parallel with the old, for long periods
to "soak" and ensure that a cut-over has 0% chance of failure. This is not
rocket science, but it is also not child's play. There are people on planet
Earth that do this on a regular basis. Most of those people would not give
two cents for the opinion of some I* society thug or clown. In some cases,
you might be talking about a 1 bit change in a protocol. That could require
six months of testing in a full-scale lab. That could then require another six
months of testing in a full-scale lab in the field, handling test traffic. Then
one may be faced with a year of limited introduction in selected field sites.
In short, you may be looking at two years from the time a small change is
determined to be useful. That change could last 50 to 100 years. People
with five years of Internet experience, need to consider the big picture more.

Jim Fleming
http://www.iana.org/assignments/ipv4-address-space
http://www.ntia.doc.gov/ntiahome/domainname/130dftmail/unir.txt

--
This message was passed to you via the ga-full@dnso.org list.
Send mail to majordomo@dnso.org to unsubscribe
("unsubscribe ga-full" in the body of the message).
Archives at http://www.dnso.org/archives.html



<<< Chronological Index >>>    <<< Thread Index >>>