More than half of $XRP Ledger nodes are still running outdated software, with just six days to go before a scheduled amendment activation.
$XRP Ledger co-creator David Schwartz spent the weekend talking about a hard fork and downplaying, in his view, the probability of any contentious split in $XRP’s consensus due to their slow upgrades.
The ledger’s so-called “fixCleanup3_1_3” amendment cleans up a series of issues affecting NFTs, permissioned domains, vaults, and a new lending protocol. It’s supposed to activate on May 27 but ideally should hold more than 80% support from trusted validating nodes for two weeks.
The release notes, published by the $XRP Ledger Foundation, tell every server operator to upgrade to 3.1.3 immediately. Any node still running older code at activation risks becoming amendment-blocked.
Amendment-blocked nodes may not validate ledgers (ledgers are the $XRP Ledger’s version of blocks), process transactions, or participate in consensus.
Schwartz on the difference between a fork and a fix
Concern about a chain split in the $XRP community spilled into a long thread on May 18. A critic asked Schwartz how a contested 50/50 validator split would even achieve resolution.
First, Schwartz downplayed changes to voting power, saying, “anyone can create dozens of nodes just to game the voting scheme,” and rejecting any rudimentary governance proposals akin to one vote per node.
Indeed, annual costs for running a hosted $XRP validating node can be as low as a few thousand dollars — a trivial expense for a motivated attack on the $83 billion blockchain.
I don’t think anyone would want a “one node one vote” scheme. For one thing, anyone can create dozens of nodes just to game the voting scheme.
At a minimum, someone has to decide they want ledger production to continue under the old rules and has to be willing to run a validator…
— David ‘JoelKatz’ Schwartz (@JoelKatz) May 18, 2026
Schwartz then walked through his view of the consensus math.
“The validator split doesn’t matter,” he said. “All that’s necessary is that each side have enough validators to create a functional unique node list (UNL) containing validators who agree to produce a ledger stream according to their preferred rules.”
UNL strips away the convention of one vote per node. Instead, $XRP Ledger node operators modify their UNL on their node’s software to prefer certain validating nodes, although almost every user adopts two particular UNLs that the Ripple-backed $XRP Foundation maintains, dUNL and XRPLF.
Anyway, pressed on what a plausible fork would actually require, Schwartz estimated, “Realistically, you need to find at least a half dozen people willing to run validators to have a plausible fork.”
The reassurance lands oddly. He downplayed fears of a chain split because forking XRPL only takes half a dozen people willing to spend a few thousands dollars.
The clock is ticking
On May 17, XRPL Foundation contributor Hussein Zangana, known in the $XRP community as “Vet,” posted that only “40% of the network is updated.”
His dashboard showed less than 40% of 846 $XRP Ledger nodes on the new version.
By May 18, RippleX head of engineering J. Ayo Akinyele put the figure at roughly 44% and by May 19, CoinGape reported about 46%.
With less than a week to upgrade, almost half of the network is in the wrong place.
Importantly, however, all 35 seats on the $XRP foundation’s super-powerful dUNL have voted in favor of fixCleanup3_1_3.
CoinGape reported 100% support among these 35 validating nodes that actually count toward practical consensus, because again, almost all $XRP users adopt the $XRP Foundation’s dUNL as their UNL by default.
The slow movers seem to be nodes operated by exchanges, market makers, $NFT marketplaces, DEX front-ends, and other operators running their own servers. They risk becoming amendment-blocked on May 27, until they upgrade.
Ripple pushes urgent XRPL patch — but nodes must trust its new key
Another upgrade problem for the $XRP ecosystem
This isn’t the $XRP Ledger’s first amendment scare over the last 12 months.
In February, the foundation disclosed a critical signature-validation flaw in its Batch amendment. That bug let an attacker execute inner transactions on behalf of arbitrary victim accounts without their private keys.
An emergency 3.1.1 release marked Batch and a companion amendment as unsupported.
Another signature flaw forced the same patch-and-resubmit cycle on the Permission Delegation amendment in September 2025. Fortunately, that earlier bug never reached mainnet or caused losses.
Unfortunately, the fixCleanup3_1_3 amendment exists in part because earlier amendments shipped with problems. Vet has framed the current activation as routine network hygiene rather than a contentious fork.
Schwartz conceded in the same thread that XRPL goes through technical hard forks more often than most public ledgers. Indeed, there have been 25 adopted amendments to the $XRP Ledger since last year.
