Nodes

Height read at

25522

Seats on the roll

1,313

Quorum a block must reach

876 of 1,313 seated seats — two thirds plus one, read from the chain’s own seating rather than recomputed here

Head block

25522 under the threshold by this node’s count of the signatures it holds. This chain has never reached its threshold, so this is its normal state rather than an incident — and finality does not depend on it: the line is a fixed depth below the head and reads no signature at all

Distinct seats that signed it, as this node holds them

123 9.4% of the roll, counted from the signatures THIS node holds for the block. Not a chain fact: the set is outside the block hash and each node collects its own

Window

100 block(s), 25423 to 25522 the newest blocks only — this is a public route and a fold over a whole chain is work a refresh should not be able to buy

Seats signing, typically

120 mean over the window; fewest 98, most 137, about 9.2% of the roll

Blocks that reached the quorum

0 of 100 0% of the window

Distinct seats SEEN in the window

305 of 1,313 on the roll — 23.2% — each one identified by a signature inside a sealed block, over the 100 block(s) of 100 in this window that this node holds a signature set for, 25423 to 25522. 12,044 signature(s) in total before de-duplication. This is not a count of seats online. A seat missing from it has very likely not had a turn, and which signatures reached THIS node is a local matter

Distinct seats that sealed

55 of 1,313 on the roll, counted from the block headers over the same window. Far smaller than the figure above, and that is arithmetic rather than a finding: the rotation gives each seat one slot, so a full turn of this roll takes about 1 h and a window of 100 block(s) cannot contain most of it

Two different kinds of figure are above, and the difference matters more than any of the numbers. The seats on the roll and the quorum threshold are CHAIN STATE, read at a height: every honest node answers them identically. The signature counts are THIS NODE’S OWN EVIDENCE: a block’s signatures are carried outside its hash, are agreed by nothing, and are collected by each node for itself after the block is sealed, so two honest nodes legitimately hold different sets for the same height — seven seats asked for the same heights at one moment returned six different counts for one of them. A second explorer on this chain may therefore show different signature counts and neither is lying.

Finality does not depend on any of it. The line is a fixed depth below the head, ratcheted, and reads no signature at all — so a block carrying none is finalised exactly as one carrying the whole roll is, and a count beside a finality line must not be read as its cause.

What is deliberately absent is a count of how many seats are ONLINE. No single node can measure that — a seat that is up and silent looks exactly like a seat that is down — and a number for it would be invented.

Node id

STR.39DD37EB40E2

Address

zk13str_39dd37eb40e2cde705a4d9c6a6ab14d8224b965a_b89a

Class

quasar

Reachable

no — this node can dial out and cannot be dialled

Accepts inbound peers

yes

Role

follower — it reads and verifies the chain and seals nothing

The address and the port are deliberately not here. This is a public page; where this node listens is an operator’s fact and is read from the node’s own operations surface, which binds to loopback and refuses anything else.

A local count, and the one figure on this page that is not about the network. A seat that dials out from behind NAT and never dials this follower is a seat running perfectly and absent from this table.

9 peers in session, and this is all of them — not a page of a longer list.

82 more node(s) known through discovery and not in session.

These two are THIS process’s own view and are not a measurement of the network. XPlorer is one node; a seat behind NAT dials out and is never in session with it, so a small number here says nothing about how many seats are running. A peer directory also cannot REACH the size of the roll: it is capped per node class — 16 records on a pulsar, 256 on a quasar, 1,024 on a supernova — and this surface runs as a quasar, so against a 1,313-seat roll this count climbs towards 256 and stops there for ever. Measured across the live fleet, no sampled seat publishes a way to be dialled at all. The seats-seen figure above is read from signatures inside sealed blocks instead, which is both a larger number and a better-evidenced one.

Every node this one knows of through discovery, in session or not. A node says whether it produces blocks; which nodes hold a seat it cannot say, because a node is known on the network by a different key from the one it seals with. Identities only: where a node can be reached is not published here.

Known to this node

91 not counting itself: unexpired directory records, plus sessions that sent none, so it need not match the directory size in the banner

Seats among them

unknown a node’s network identity is a different key from the seat key it seals with, and nothing a node sends binds one to the other, so this directory cannot say which of these nodes hold seats. 91 say they may produce blocks — a claim every quasar and supernova makes, seat or not — and 0 run under a network address that is itself on the roll. The roll and the signatures on each block, above, are the evidence for the seats

Declare they validate

91 a claim each node makes about itself. Every seat makes it, and so does any quasar or supernova, seat or not

Network address on the roll

0 proven seats, at height 25522

Followers

0 cannot produce blocks: light nodes, and any node announcing no validator role

In session now

9

Quarantined

0

With a signed record

89 the rest are in session and announced nothing

Directory holds at most

256 records

91 known nodes, and this is all of them — not a page of a longer list.

One node’s directory, not a census. It holds what this node has heard, capped by its class, and a seat that never announces to it is a seat running perfectly and absent from this list.

The same directory by type, newest arrivals first, kept live.

Protocol

praeco-v1

A record lives for

30 minutes

Quarantine after

5 failed dials

Quarantine holds

32 addresses

Seeds

7 named at start-up; their addresses are not published here

A node learns peers from the ones it already has. It needs no registry and no server: a seed is a starting point, not an authority, and a node with no peers still verifies everything it holds.

What a node on this network may be. A fixed table compiled into the build rather than a list of nodes: it says what is possible, and says nothing about how many of each are running.

3 classes, and this is all of them — not a page of a longer list.

There is no server in this design and no coordinator to go down. Every node keeps its own chain, verifies every block it accepts against its own genesis, and reaches the others directly. What a seed gives a new node is somebody to ask; what makes the answer trustworthy is that the node checks it rather than that the seed said it.