[governance] Uni.X to Uni.X .NETworking - Addressing Outside the 64-bit Space

[Jim Fleming]

Uni.X to Uni.X .NETworking - Addressing Outside the 64-bit Space

The 64-bit Uni.X to Uni.X .NETworking of grid-agent nodes breaks down into
the
following natural aggregates [based largely on backward inter-working with
legacy systems].

18+32+14

2+2+4+3+7+ 32 +1+2+6+1+1+3

Any time an addressing scheme becomes bounded by, for example, 32 bits or 64
bits, there
is a natural reaction to consider addressing outside of the in-use
addressing. That is often
needed for transitions and for maintenance work that is "back-stage" -
hidden from the
view of the average user.

With 160-bits and two 64 bit fields resulting in 128 bits, there is not a
lot of space left for
more addressing. One place that TWO more bits could be found is in the 19
code-points
hidden or un-used in the Length values. Sixteen (16) of those 19 code-points
can be used
resulting in a small 4-bit field. Those 4-bits can be divided SSDD or SDSD
to provide 2
additional address bits, outside of the 64-bit addressing. SSDD is likely
preferred because
the bits would likely be added Prefix bits to encapsulate the entire 64-bit
space into 4
layers, which allows for cases where two redundant layers are in service and
two other
layers are being migrated in or out and tagging is desired. If SDSD is used
then one would
likely end up with one Prefix bit and one Suffix bit resulting in a two
64-bit spaces and for
each address in each space a spare or alternate address.

The added addressing would only be available to the three protocols, ICMP,
TCP and UDP
because the NOP protocol uses all 10 of the Length bits in conjunction with
the 16 Check-Sum
bits to create a 26 bit field which is divided into 2 bits for Length and up
to 3 Bytes of Data.
When 1 Byte is used, the Check-Sum reverts to that use and the Byte is
stored with the length
code in the 10 bit Length field. When 2 bytes are used, the spare Check-Sum
is used as a
check-byte only on the 18 in-use bits of the 26-bit field. If the rest of
the 160-bits are corrupted
the packet will not likely make it to the right place. If it does, the small
8-bit check byte is
focused on the small 2-byte pay-load of data and 2-bit length. You have
8-bits checking 18-bits
making the check more reliable than if 8-bits were checking 152-bits. With a
large bit field, if
multiple errors occur, a small field of check bits can be useless or a false
sense of security.

>From a governance point of view, the 64-bits are the focal point. Addressing
around the edges
is handy for maintenance of the universe or planet but may not be part of
the day-to-day
governance. People of course have to be concerned about covert governance
activities around
the edges, back-stage, etc. and should make sure they pay attention to how
all of the bits are
used or not used and routed and of course recorded.

The routing does not necessarily follow the natural aggregate governance of
the address space.

18+32+14
2+2+4+3+7+ 32 +1+2+6+1+1+3

Just because someone has address space from region A and another person has
address space
from Region B, that does not imply that their traffic has to flow via some
super-Region node
that connects A and B. Wireless (WIFI and WIMAX) are hopefully going to help
people see
that packets can be routed directly from A to B across un-natural national
boundaries.

If one were forced to break down the aggregates based on some un-natural
national boundaries
or geo-centric boundaries, then the following might work.

2+2+4+3+7+ 32 +1+2+6+1+1+3

2 - supports up to 4 planets
2 - supports 4 regions on each planet
4 - supports 16 land-based super states in each region
3 - supports 8 governance regimes in each super state
7 - supports 128 major metro areas in each of the above
32 - is a .NET the size of the research legacy network for each of the above
1 - is handy for redundant even-odd mated-pair applications
2 - provides 4 major sub-nets at a location
6 - is handy for addressing up to 32 phones, PCs, agents, etc.
1 - is handy for redundant PCs, phones, agents working side-by-side, etc.
1 - is handy for duplex processing in each PC
3 - can be used for 8 processes or daemons to interwork or intrawork or for
some traditional sub-net arrangement with 2 special addresses and 6
end-to-end devices.

With the above, imagine you are one of 4+ billion people who have a unique
32-bit prefix
and you drop it into the above. You pick one of 4 planets, then a region,
then a super-state,
then one of 8 political parties and then settle into one of 128 metro areas.
>From there, you
deploy your redundant network with 4 major sub-nets each supporting your 32
children
[How many people have more than 32 children?] and then each child has a
phone and
video game, and some have dual processors, and then in each one there are 8
real end-to-end
locations that can communicate end-to-end with all of the other planets,
regions, etc. etc. etc.

Is 64-bits enough ?
Is it routable ?
Can the 64-bits be placed inside of a 128-bit field and the other 64-bits
used for data at
the moment and later be used for routing ? IF the above does not work out as
enough
addressing...

Are there other ways to govern 64 bits ?
18+32+14
2+2+4+3+7+ 32 +1+2+6+1+1+3

18+6+32+7
2+2+4+3+7+6+ 32 +1+2+1+1+3

Can the market decide ?
Should programmers pick something ?
Do people think a coin-toss or random arrangement would really be better
than looking at the reality of the past, today, and the likely future ?

Do people understand they can obtain a unique address from several regimes
and their
wireless devices could still route to each other across the street or in the
same apartment building.

Do Computer Professionals for Social Responsibility know anything about
"Computers" ?


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