moving when all the scenarios are not yet complete [RE: DSTM]

[Pekka Savola <pekkas@netcore.fi>]

A new topic to better suit the subject..

On Wed, 23 Jun 2004, Christian Huitema wrote:
> > But when I saw the mail Teredo was going to move without doing
> finishing
> > all scenarios/analysis docs I thought we were opening the flood gates
> > for all mechanisms.  I could be wrong?  But all mechanisms should be
> > applied to that entire set of work efforts.
> 
> I have two issues with this statement.
> 
> First, I don't think we should wait until *all* scenario documents are
> completed before we standardize or publish *any* new transition
> technology. The bar ought to be lower: we should progress a transition
> technology if we agree that it is clearly needed by at least one
> scenario. Otherwise, we could keep inventing new scenarios, and we would
> always have to wait until yet another scenario analysis is completed.
> 
> Second, we have to define what "completed" means. What is the decision
> point? We have actually all but completed the "unmanaged networks"
> evaluation: we went through the working group last call, and the
> document is now on the IESG plate. Based on this scenario, doing work on
> Teredo is not spurious.

What Christian said. The analysis of 3GPP, Unmanaged and ISP have all
already left this WG (with rough consensus) and are at IESG
evaluation, or beyond the IESG evaluation (and none of documents got
IESG pushback on Teredo/BGPtunnel/etc.).  I fail to see what more one
could expect here.  Does it matter what we decide e.g.  in the
Enterprise document, if Unmanaged, ISP, or 3GPP already requires a
certain kind of mechanism?

There is, however, an argument to be made if a mechanism would be
required in multiple scenarios, where one or more of the scenarios was
not finished yet -- then the question would be what would be the basis
on where to evolve the specification? (for example, let's consider
ISATAP in 3GPP: one could remove direct tunneling support, but then it
might become useless (at least to some) in Enterprise -- for
mechanisms which dependencies it might be worth a bit more of
wait-and-see, but e.g. Teredo has only been proposed in the Unmanaged
scenario.

So IMHO this is an argument for moving forward (if there is consensus,
as there seems to be in Unmanaged and ISP) in all the scenarios which
are already at the IESG (read: when there is consensus in the WG, and
there is no objection at the IESG).  The mechanisms required by
scenarios which are not done yet IMHO cannot go forward at this point.

-- 
Pekka Savola                 "You each name yourselves king, yet the
Netcore Oy                    kingdom bleeds."
Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings