RE: Point of Order

["Douglas Otis" <dotis@sanlight.net>]

Jay,

These conversations will continue for as long as it took to develop
tunneling protocols, name servers, storage switching fabrics, http servers,
authentication servers, dynamic user based mapping, end-to-end to buffer-to-
buffer control, fcp conversion, multi-connection asymmetric/symmetric
channels, perhaps even boot protocols.  All of this already exists, but not
in one protocol!

It tunnels, it maps, it authenticates, it bundles, it dynamically routes, it
translates names for luns, targets and IP addresses all within a single
protocol real-time with only a few wasteful discards, stateful mid-points,
and confounding sequences sending response ahead of data.  At this point in
time however, all of this is seen as needed and to ask for a vote, each item
would receive an affirmative.  So until Yankee thrift kicks-in and while not
invented here prevails, the sausage making continues.  You are advised not
to watch.

Doug

> I have been monitoring the iSCSI group for several months. It is my
> perception that there are way too many opinions and positions and way too
> little consensus.
>
> What is the protocol for establishing consensus on any given issue so that
> the WG can move forward to NEW issues? Tabling controversial
> subjects until
> a later date also seems to be a reasonable option so that progress can be
> made on a spec.
>
> It seems logical to me that each area in the spec should be discussed
> (brainstormed) and then voted on or shelved until later. Discussion should
> center around the spec and issues therein instead of on random flaming,
> flying, dueling email threads. IMHO, 200 emails in two days is too much!
>
> I think it would help greatly if the WG chair(s) would provide a periodic
> "point-of-order" to focus or re-focus discussion by the teaming masses.
>
> Regards,
> Jay Brinkmeyer
> Compaq Computer Corporation
>
>