Four chains shipped governance changes today, and none shipped a feature
Serguey Asael Shinder2 min readRead today's desks side by side and a pattern falls out that is easy to miss when each item is read alone. Almost nothing shipped a user-visible feature. What moved was the rules for who decides, and the formats that outlive the people who wrote them.
Robonomics proposed retiring Sudo — a developer multisig with privileged power over the chain — and replacing it with two OpenGov tracks, with the removal of Sudo itself as the first referendum. Solana's changelog carries a discussion about removing the fixed 100M compute-unit block limit, which would move the decision about how much a block may contain from the protocol to individual validators. XRPL took in a draft for coordinating who has signed a partially-signed transaction. Bitcoin merged a BIP for encrypting the wallet data that is not the seed.
Different chains, different maturity, one subject: authority and its bookkeeping.
Three things follow that are worth holding onto.
A privileged key is a governance decision that has not been made yet. Every chain starts with one because it must, and each keeps it longer than intended because removing it requires a working replacement and a moment where nobody can fix anything by hand. Robonomics is proposing to spend that moment deliberately. The honest detail in its post is the question to other teams: what did you wish you had kept as a recovery path?
A limit in the protocol is a limit on what you can be required to run. The Solana discussion is presented as flexibility, and it is — but a fixed block size is also the thing that lets a small validator know what hardware suffices. Removing a ceiling does not distribute the decision evenly; it hands it to whoever can sustain the most, and everyone else discovers the new requirement after the fact.
A format decides what survives. The BIP and the XRPL draft are both about state that is not money and not consensus: backup descriptors, signature progress. Neither is exciting and both are where real losses happen — funds that are unrecoverable because the descriptor was lost, custody that fails because nobody could establish who had signed.
The practical reading, for anyone assessing a chain rather than trading one: look at what happens when the founding team stops being available. Who can upgrade the runtime. What document a wallet written five years from now would need. Whether the block limit is a number in the protocol or an emergent property of whoever is richest. Those questions were all being worked on today, in four repositories, by people not talking to each other.
⚠️ Nothing here is investment analysis or advice. These are engineering and governance observations drawn from published proposals, and a proposal is not a decision.