Unicast etc meta-observation

Sun Nov 21 04:14:11 UTC 2021

Reading over many of these notes my observation is that many here are
good at understanding the technical points of the proposals and
throwing around 224/4 this and 127/8 that.

Then the discussion mostly disintegrates into anecdotes with hands
waving furiously, "my anecdote is VERY important!"

What would be useful would be some attempt to quantify what those
anecdotes are trying to express such as the costs and disruptions they
imply.

Something more like engineering or economics or engineering economics.

Let's use a measure, "Cost And Disruption to the Internet's
Infrastructure", call it the CDII.

For example:

1.  Proposing these address re-uses: CDII zero, other than the cost of
endless discussion.

2. Writing them down into revised standards: CDII zero, other than the
effort to agree and write them down.

3. Deploying an experiment: CDII about zero one hopes, again other
than keeping "lab notebooks" and reporting back results.

4. Deploying wider, voluntary usage: CDII larger as this begins to
require changes in, for example, actual routing, firewalls, etc in the
wild and may begin to affect those not intentionally involved.

5. Full implementation: Still voluntary of course because everything
is. But this might be the point where someone requesting IPv4 address
space is assigned a block in these re-purposed spaces. CDII grows.

There are sub-threads implied for the various repurposed blocks with
different CDIIs.

Many seem to feel, for example, that re-use of 224/4 would have
relatively low CDII, 127/8 perhaps a somewhat higher CDII, etc. What's
the detailed list?  What are the relative CDIIs?

And there are devils in the details of quantifying CDII per se. For
example CDII for infrastructure operators vs end users for a given
deployment.

But it's not totally open-ended or unknowable or, perhaps put better,
not impossible to enumerate and make intelligent estimates.

It's not a very deep or bushy tree. Just a few choices at each
decision point with CDIIs assigned.

Plus some attempt to aggregate those CDIIs and decide what is
acceptable and what is not, thresholds. Plus time horizons and
quantifiable measures of success and failure along the way.

Something like this could be more productive than describing one's
connected lightbulb or networking in Armenia one more time even if
such anecdotes would form the basis for that quantification.

Since this and similar discussions have been going on literally for
decades maybe the time is past due to try to quantify something like
this proposed CDII model.

-- 
        -Barry Shein

Software Tool & Die    | bzs at TheWorld.com             | http://www.TheWorld.com
Purveyors to the Trade | Voice: +1 617-STD-WRLD       | 800-THE-WRLD
The World: Since 1989  | A Public Information Utility | *oo*