[Cryptography] Clarification on the Web4 model (state, Δ-updates, and crypto inputs)

[Patrick Chkoreff]

On 11/28/25 11:17 PM, nouun via cryptography wrote:
> Hi all,

> 1. What the “state vector” actually is
> 
> Web4 maintains a single instantaneous state, not a ledger and not a history.

Web4 is defined as some loosely amalgamated collection of poorly defined 
magic things.  How does that maintain the array of numbers you describe? 
  Is it on a single server?  Split across multiple servers?  Can 
anything iterate over all the locations and verify that they sum to zero?

What does it mean if someone in your system "pays" 10^80 units of your 
stuff to someone else?  There's a local imbalance of A-10^80 and 
B+10^80.  What gave A the right to issue 10^80 units of your stuff like 
that?

About 20 years ago I did the Loom system (https://fexl.com/loom), which 
has a giant sparse array of 2^128 rows of 2^138 columns each.  Each row 
is an asset type.  The columns in a row are the locations where assets 
are stored.  Each column has a number, by far mostly zero, and the 
values of all the columns in a row sum to zero.  For each row there is 
only one location with a negative value, and that is known only to the 
single privileged issuer of that asset.

It was kinda cute and fun, but it is (1) centralized, and (2) ya gotta 
trust that the system is doing the math right in the "move" operations.


-- Patrick