[governance] Uni.X to Uni.X .NETworking - 160 bit Messages and 320 bit Messages and DHT

[Jim Fleming]

Given the 160 bit Uni.X to Uni.X .NETworking message format:

0101.0101.SSSSDDDD.000000.LLLLLLLLLL
SSSSSSSDDDDDDD.SD.SD.SD.SSSSSSDDDDDD
SD11GTTT.PPSSSDDD.CCCCCCCCCCCCCCCC
SSSSSSSSSSSSSSSSSSSSSSSSSSSSSSSS
DDDDDDDDDDDDDDDDDDDDDDDDDDDDDDDD

The NOP Protocol (PP=00) supports up to 3 bytes in the messages.

With the 320-bit message format, the number of bits doubles. Most of those
bits are used for up to 16 bytes of data in the message. There is no
check-sum
in the 320-bit message format, but the hop-count is larger for those wanting
to
attempt to send longer distance messages. One penalty is of course that you
use
up more band-width with double the bits on the wire or wireless channel. You
also expose more bits to potential errors and have no check-sum to detect
that.
The 320-bit message format can be more useful when the storage mechanism or
transport mechanism provides a layer of reliabilty. In both case, 160 bit
and 320
bit the messages can be stored on disks and replayed later or streamed on a
remote LAN.

The 64-bit addressing can remain the same in both cases.

01.01.DDDD.000.DDDDDDD.0.1.<<<<32 bits>>>>.0.000000.0.1.DDD

When one-way address formats are used with existing transports the 64 bits
can
all fit in the old 32-bit fields. That allows for global, no-source-address,
sends to
specific nodes. It also allows for nodes to broadcast to the ether with no
specific
destination address. All 64 bits in the legacy header are used for one-way
addressing.

With DHT, Distributed Hash Table, the entire 160-bit message can be the key
or
another 160-bit message can follow and be used as the key. That makes keys
and
messages look very similar. Some call that security via obscurity. If you
take your
4-bit alphabet and use it to encode a key, you can have 40-symbol names. If
you
use your 5-bit alphabet (with 32 symbols) you can have 32-symbol names. Many
domain names, including the dots, including spacers, fit in the 32-symbol
size. The
so-called TLD (top-level-domain) is also included and just part of the name.
The
DHT supports ALL of the 160-bit keys and does not care if they come from
names.

There is some concern that 160-bits is not large enough for some
applications.
The 320-bit message format can also be used as a key.  The 64-bit addressing
stays
the same in both cases. When the 320-bit keys are used, they can be split in
two and
two 160-bit queries can be sent to the DHT and the returned data can be
combined.
That can put more burden on the 160-bit key-space. Another approach is of
course
to attempt to route the 320-bit messages, with 64-bit addressing and up to
16 bytes
of data and to somehow blend the 160-bit DHT and 320-bit routable node
exchange
points. That is an area where people could do more testing on how they want
the TTL
of the DHT to impact the results obtained. The 16 bytes of data become part
of the
key and addressing. The nodes do part of the hash computation when the
message is
received. This may result in a scenario where people send a query which
amounts to
saying, "Using your DHT, do this DHT query.". That allows for many DHTs and
for
people to use each other's DHTs, as well as the "global DHT" with a 160-bit
key.

Some of these decisions can end up applying to meat-space governance because
the
size of keys concern people. There are applications, especially with CASH,
where people
plan to store cash (or gold) at known exchange points. Imagine leaving a
dollar bill in a
certain book on an obscur shelf in the library. The key would indicate the
location. If you
know the key, you can pick up the dollar and move it to another location.
The money does
not leave the system, it just moves around with the key exchanges.

If you want to see this in action, some people use 128-bit keys to store and
move gold.
https://loom.cc/gold/

With Uni.X to Uni.X .NETworking, the addressing is 64-bits for the nodes,
and the keys
are 160-bits and 320-bits which match the basic message sizes. That allows
messages to
be used as keys and vice-versa. The 64-bit addressing fits in both of the
message formats.
The 64-bits are routable with existing technology, especially with careful
reservations of
bits for future use. The keys are not viewed as routable, they are used to
Put and Get the
data. The applications assume this as a base and treat the data as objects
and send messages
to and from the objects. The data becomes smart data, or active data. There
are languages
that allow you to develop programs that allow the objects to communicate.
The technology
has been available for a long long time, THE Big Lie Society censors it and
of course will
now attempt to re-invent it and claim they invented it. That is their lie
they have to live with.

_______________________________________________
governance mailing list
governance at lists.cpsr.org
https://ssl.cpsr.org/mailman/listinfo/governance