Measurement and the ballot
The comparison between the two branches ends in a community vote, and the result decides which engine the official network is built on. This page is about what comes before the ballot — because a ballot with no measurements in front of it is a vote on sentiment, and the louder side wins.
Why that table carries no operating figures
An operating figure frozen into a document makes both branches judged by a measurement nobody can reproduce — and that error does not announce itself, because the number still looks perfectly concrete. This is not hypothetical: when the project first measured both in parallel, the two branches produced two quantities carrying the same name without the same meaning. One branch produces blocks on a steady time interval; the other produces a block only when there is a transaction. Printing two seconds per block figures side by side would read as "this one is hundreds of times slower", when the truth is that one of them does not produce empty blocks.
The lesson is a rule, and it applies to every comparison from here on: using the same measurement is not enough to compare — both sides must also be doing the same work. Getting genuinely comparable numbers means putting the same load on both first, and that is a gate not yet passed.
So each branch's live figures belong to that branch's own explorer, not to this document.
The vote, and what makes it mean anything
The comparison ends in a community vote, and the result decides which engine the official network is built on.
| Criterion | How it is measured |
|---|---|
| Degree of decentralisation | raise the number of machines signing until finality degrades, and record where it degrades |
| Finality | time from submitting a transaction until it can no longer be reversed, measured on a public endpoint |
| EVM maturity | run a standard contract suite and record which pass without modification |
| Wallet compatibility | add the network to a standard wallet and complete a transaction |
| Creating a chain | time from the creation command until the chain answers on its own endpoint |
| Interoperability | move an asset to another chain and read it back at the far end |
| Cost of operation | resources per chain, and what deploying a chain actually costs |
The right-hand column is the one that matters, and it is a guardrail rather than decoration: a published score with no reproducible measurement behind it is not a score — it is an opinion with a number attached. The community sets the weights between criteria, and no branch scores itself.
⚠️ The vote is a gate that is NOT OPEN. How the ballot works, who is eligible, and the window will be published before the first vote is cast. Until then there is nothing to sign up for, and no page of the project collects anything from you in exchange for a future share. Anyone inviting you to register to "hold your place in the vote" is offering something that does not exist.
Why this sequence is harder than it looks
Three things have to be finished in order, and each is a gate: produce the same load on both branches → publish the measurements with their method → open the ballot with a mechanism published in advance. Skip the first and the two numbers cannot be compared; skip the second and the ballot has nothing to stand on; skip the third and nobody can recognise the result.
Figure 13 — Three unpassed gates stand between the testing stretch and the ballot. An arrow looping back on itself is how this document draws a gate that is not open.
In short: what decides the worth of this ballot is not how many people vote, but whether a reproducible measurement stands behind each score. Until those three gates are passed, every statement about which branch is better — including from inside the project — is an opinion, not a result.
→ Next: Platform architecture — The layers, and where responsibility sits