Breaking News from a Composer Operator
Potential Composer:
“I heard Neptune needs more composers.”
Vault-Tec:
“Absolutely! Decentralization is very important.”
Potential Composer:
“Great! How do I participate?”
Vault-Tec:
“First, buy a powerful machine.”
Potential Composer:
“Done.”
Vault-Tec:
“Excellent.”
Potential Composer:
“Did I earn anything?”
Vault-Tec:
“No. You have to give it all to the guessers. ”
Potential Composer:
“But why do I see somebody earning 64 NPT through composing?”
Vault-Tec:
“Pool. They have their own GPU composer.”
Composer Operator:
“So GPU composers are the future?”
Vault-Tec:
“No, no. Nobody said composers have a future.”
Potential Composer:
“Okay… so I can become a guesser instead?”
Vault-Tec:
“Oh no no.”
Potential Composer:
“Why not?”
Vault-Tec:
“You need to join a pool.”
Potential Composer:
“Why?”
Vault-Tec:
“Because solo guessing does not exist.”
Potential Composer:
“Oh… so then I earn rewards through the pool?”
Vault-Tec:
“No no.”
Potential Composer:
“No?”
Vault-Tec:
“You watch the pool earn rewards.” 
Potential Composer:
“There must be some way to make money from mining.”
Vault-Tec:
“Maybe in the future there will be GPU composers and GPU guessers.”
Potential Composer:
“Finally! Then I can earn something?”
Vault-Tec:
“Well…”
Potential Composer:
“Yes?”
Vault-Tec:
“We expect 128 NPT for guessers after 40 seconds instead of 2.5 minutes.”
Potential Composer:
“…”
Vault-Tec:
“Oki doki! ”
Breaking News
After yet another round of awarding 128 NPT to guessers following 2.5 minutes of composing, the number of active composers has declined.
Fortunately, experts confirm that this is a very positive signal for decentralization.
The fewer composers remain, the more decentralized the surviving ones become.
Oki doki!
Jokes aside
I think the real question is not:
“Who can give more NPT to guessers?”
but:
“How do we encourage competition around producing the best proposal for the network?”
One possible direction could be to define some objective network-value function and allow the market to compete around proposal quality rather than primarily around fee escalation.
The exact metric is probably a separate discussion.
The important point is that proposal quality should matter.
Guessers should still be rewarded well.
Higher guesser rewards should still be possible.
But ideally fee wars would not become the dominant strategy.
I believe runtime set-guesser-fraction and recompose could help here.
The purpose of recompose is not to protect composer profits.
The purpose of recompose is to allow the market to converge toward the best proposal without forcing participants to discard accumulated mempool knowledge every time the fee market moves.
Today, changing guesser incentives often means abandoning an in-progress composition and starting over.
This may unintentionally reward fee escalation more than proposal quality.
With recompose a composer could:
-
observe a stronger proposal
-
update guesser fraction
-
recompose
-
preserve mempool visibility
-
continue improving the proposal
The composer with the best view of the economically valuable mempool should have a natural advantage because that proposal likely creates the most value for the network.
One final thought.
I fully understand that TIP-10 is the correct priority today.
However, once TIP-10 lands, I believe mining deserves another look.
The ecosystem is evolving rapidly:
-
public GPU guessers will likely appear
-
public GPU provers/composers may appear
-
hardware requirements will change
-
market structure will change
At that point fairness and decentralization become even more important.
Today, a significant part of the network’s guessing power appears to be concentrated in DRPool.
I am not suggesting anything improper is happening.
My point is simply that the broader community currently has no practical way to reproduce or verify the performance characteristics that lead to this dominance.
As a result, many participants are left trying to infer what is happening from market behavior alone.
This creates uncertainty.
In decentralized systems, trust is good, but reproducibility is better.
If public GPU guessers become available, independent composers could once again guess on their own proposals.
If public GPU provers/composers become available, competition may become significantly more transparent and reproducible.
That could be a major step toward both fairness and decentralization.
At that point we may discover that the current economics are perfectly fine.
Or we may discover that they need adjustment.
Either way, I think the discussion is worth having after TIP-10.
These are simply observations from someone running composer infrastructure and watching the market every day since February 2026.
Vault-Tec Research Department
“How do we improve decentralization?”
Reward composers
Encourage independent operators
Sustainable mining economics
Allow solo guessers
Everyone joins a pool
128 NPT for guessers
Hope more composers appear
According to our latest simulations, decentralization reaches its theoretical maximum shortly after nobody wants to become a composer anymore.
Vault-Tec™ — Building a Better Future Underground Since 2077.