Which metric are we talking about?
ARK Invest and Glassnode published a joint assessment on September 1 that measures what is called network capture risk. The key number: the smallest group of block-producing entities you would need to assemble in order to cross the threshold relevant for control over the protocol.
According to their framework, that number is:
| Network | Number of entities to cross the threshold |
|---|---|
| Bitcoin | 3 |
| Ethereum | 3 |
| Solana | 19 |
Source: CryptoSlate, citing the joint ARK Invest and Glassnode score.
Read in isolation, this looks like a point for Solana: assembling a coalition of 19 entities is harder than lining up 3. So on this one axis, Solana beats both Bitcoin and Ethereum.
So why did Bitcoin finish first on decentralization?
Because the coalition needed to disrupt consensus is just one of several risks. According to CryptoSlate, the same framework also factors in ownership, infrastructure, software, auditability, and how quickly you can exit the network. In the overall composite decentralization ranking, Bitcoin finished first.
In other words: the metric of "how many entities have to join forces" and the metric of "how easily the network fails for some other reason" are not the same thing. The tension between these numbers is at the heart of the whole story.
So where is the software risk?
Software is listed in the assessment as a separate path of exposure. The risk here is not about a coalition of miners or validators, but about the fact that a single bug in the dominant client software can hit the network as a whole. That is a different kind of vulnerability, one that cannot be captured by the number "19 entities."
A clarification of what we know and what we don't: the sources themselves list software as a risk axis, but we do not have a specific quantification of this risk for Solana in the available materials. We therefore do not cite any number that would put a figure on this particular scenario.
What is changing in Solana technically right now?
At the same time, a concrete change is moving through the network. According to BitHub.pl, the maximum size of a single transaction was raised from 1232 bytes to 4096 bytes, roughly triple.
CoinDesk sets this in the context of the Transaction v1 feature, which according to their article (dated September 7) activates on Wednesday. The goal is to allow more complex proofs and large multisig operations to fit into a single transaction. A side effect: services that read data from Solana have to update.
This is exactly the type of change where software risk shows up in practice. A protocol change that the entire ecosystem of surrounding tools has to account for is the moment when a bug or an un-updated service can cause problems.
What is happening in on-chain activity?
In parallel, application activity is rising. According to Cointelegraph, the Fomo app generated 1.76 million dollars in revenue on Friday, beating the Pump.fun launchpad with 1.1 million dollars on the same day. Over a 30-day window, however, the same source says Pump.fun still remains ahead.
A single-day beat is not the same as a long-term trend. We put both numbers side by side precisely so the reader can put it in context for themselves.
What to watch for with this type of news?
- One number is not the whole picture. "19 vs 3" measures one specific axis, not the overall resilience of the network.
- Methodology decides. Who defines the threshold and how determines the resulting ranking. Here it is a joint framework from ARK Invest and Glassnode.
- Protocol changes are sensitive points. The activation of a new feature that the entire tooling ecosystem must adopt is a moment of heightened operational risk.
- Daily vs. monthly metrics. A one-day record says nothing about the trend.
Charliedesk does not give buy or sell recommendations. We show what happened and what the data says about it. Once it becomes clear whether the announced change goes through without an incident, it can be verified retroactively against this text.