Dotted Toiletpaper (Re: [ga] Cyberspace Security and the Root(s))

[Marc Schneiders <marc@fuchsia.bijt.net>]

On Thu, 19 Sep 2002, at 22:42 [=GMT-0000], Roberto Gaetano wrote:

> The ideal situation is indeed to have a system of servers that carry
> identical data. In fact, what about 13 copies of the same data,
> scattered in different continents,

Some continents missing, one very overrepresentated.

> whose information is guaranteed to be identical by a
> co-ordinated effort of the 13 managers?

Why is that more ideal than 3 times 13 managers (or 2 times 10)? 39 (or
20) servers are more difficult to kill than 13. The distance of many users
to them would be shorter. Co-ordination need not be more complicated.

> Maybe thatīs what we need to implement, I wonder why nobody has thought
> about this simple and ideal solution!

Because we are all entrenched in some form of root-religion and (ab)using
technical issues to argue politics/governance? And thus avoiding the
question that really matters: Who governs these important structures of
the internet? And to whose benefit? And since nearly all people govern for
their own benefit, it is best that nearly all people are represented in
the governing body. Unless we are willing to assume, that the 15 or 20
people on a Board are all selected from the small percentage of humans
that can be called saints.

There are people who argue from the principle 'more is better' that
multiple roots are good, and thus try to legitimize something completely
different and principally unrelated: Multiple roots carrying different
data are OK. Not a valid conclusion naturally.

There are other people who try to show that multiple roots are evil in
themselves, because they believe in the greatness of ICANN (or another
organization they hope will develop from/after ICANN): one authority over
the Names and Numbers. Any technical change that alters the root status
quo is feared, because it might call forth questions of who have authority
over the Holy List. And why. And whether or not they keep it in good
shape.

That is of course what we should be talking about here. Another sort of
zonetransfers than some major ccTLDs refuse to ICANN/IANA.

Another sort of discussion, or rather fight, than the one Stuart Lynn is
forcing upon all, the grab for power by mere assertion:

"We are King ICANN, so if it is our royal pleasure, we should be allowed
to determine what design the toiletpaper has that you at DENIC use.

And we have determined, it is in fact Old Internet Practice, going back to
the days of Postel (we call it IPC-2), that dots are not allowed on
toiletpaper.

So, we order each ccTLD to send us for inspection an (unused!) sheet of
the paper they use in the little room, labelled with their Country Code in
the top left corner (assuming portrait as the perceived shape of the
sheet).

Please, do not put it in any other corner, since we also have to ask you,
_not_ to put the customary dot before your Country Code. So we cannot make
out whether we might be reading your Country Code upside down, if it could
be in any corner. (If the possible writing of the two characters from top
to bottom should also be taken into account, we might also mistakenly read
ZX for XN. And IL could become LI, if writing from right to left would
have to be expected. In the example just quoted of great relevance, as
anyone only faintly familiar with the scripts used in IL and LI can
attest.)

So no dots needed, and no dots wanted, since we intend to use the samples
you send us after inspection ourselves. So they cannot have dots. Not
that we mind breaking our own rules.

Don't forget to add the accompanying fee: $20,000.00. We cannot start
before we have the money. And only after this matter is settled we can
have a look at your requests for changes in the Holy List. You have to
deliver first. Because we have the Holy List. Because we are King."

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