Update examples to reference Mesa instead of Berkeley by Shigoto-dev19 · Pull Request #2751 · o1-labs/o1js

Shigoto-dev19 · 2026-02-02T15:19:29Z

Closes #2673.

Changes

Update Mina node, archive, and explorer endpoints to Mesa
- GraphQL endpoint: https://plain-1-graphql.mina-mesa-network.gcp.o1test.net/graphql
- Archive endpoint: http://mesa-archive-node-api.gcp.o1test.net
- Explorer URL: https://mesa.minaexplorer.com (replacing both minascan.io/berkeley and berkeley.minaexplorer.com)
Default faucet network parameter changed from devnet to mesa
- Note: we might need to change this again to devnet assuming that mesa would be the default devnet once the hardfork is applied.
Rename example files referencing Berkeley to Mesa
- simple-zkapp-berkeley.ts -> simple-zkapp-mesa.ts
- voting/run-berkeley.ts -> voting/run-mesa.ts
- Note: the replacement here is straightforward but renaming the examples with live or devnet suffix might be more meaningful.
Replace berkeley hash option with network-agnostic legacy in mina-signer
- hashPayment and hashStakeDelegation option changed from { berkeley?: boolean } to { legacy?: boolean }(inverted logic)
- Migration: { berkeley: true } becomes { legacy: false }
- This is a breaking change, but it will be handled alongside other breaking changes when the hard fork is applied
  Chose legacy over mesa because the flag controls V1 vs current hashing behavior, not anything network-specific; this is consistent with the existing sign-legacy.ts naming and avoids needing another rename for the next network.

TODO

Verify Mesa node and archive endpoints are still up-to-date at merge time (they may change before the hard fork)

Notes

Some examples require pre-funded keys to run. I've manually funded the hardcoded keys used in the examples have via the Mina faucet.
- Note: each key was funded by 10 Mina which might be drained in the future and block CI.
Mina.faucet() is currently non-functional: the faucet endpoint (faucet.minaprotocol.com) is deployed behind Vercel, which intercepts programmatic fetch requests from Node.js with a 429 status and returns an HTML security checkpoint page instead of JSON. The o1js client code is correct; this is a server-side issue where Vercel's bot protection blocks automated/non-browser traffic before the request ever reaches the faucet API.
Some examples like the voting/run-mesa.ts example generate random keys and rely on Mina.faucet() at runtime, so it will not work until the faucet issue above is resolved.

…-signer

Shigoto-dev19 · 2026-02-02T15:37:32Z

Funded the account used here: B62qjTDdVdxkkRNRbkpdq6D9RUvX2C7RUaW3voHMAjtmk6WLxuXHfS9.

See https://mesa.minaexplorer.com/transaction/5Jtpjt2hpHionPm3JCZ4Z7PS9qxE1jj2ZMLQEWN7TLNMQbj8iLBq

Shigoto-dev19 · 2026-02-02T15:39:13Z

            pendingTx.hash
          }`
-        : `https://minascan.io/berkeley/tx/${pendingTx.hash}?type=zk-tx`)
+        : `https://mesa.minaexplorer.com/transaction/${pendingTx.hash}`)


mesa.minaexplorer.com is what works for Mesa Testnet for now; we might need to update this with another explorer in the future.

Shigoto-dev19 · 2026-02-02T15:42:01Z

Funded both B62qqzhd5U54JafhR4CB8NLWQM8PRfiCZ4TuoTT5UQHzGwdR2f5RLnK & B62qnScMYfgSUWwtzB1r6fB8i23YFXgA25rzcSXVCtYVfUxLHkMLr3G.

See https://mesa.minaexplorer.com/transaction/5JtsqFPnq8sWdLtym1v4jCc2fPxnEu59aGLrQseFtykgvhCJoWsW & https://mesa.minaexplorer.com/transaction/5Jv8TG2DYsKbTMxVGVjT4x7QCXNavziSuZH1suRz2gwezRWCMxFw respectively.

Shigoto-dev19 · 2026-02-02T15:43:25Z

Funded B62qpfgnUm7zVqi8MJHNB2m37rtgMNDbFNhC2DpMmmVpQt8x6gKv9Ww.

See https://mesa.minaexplorer.com/transaction/5Jtq7XUPKR1uHfntmBd3cbaeNNT3J9D9pvyCHNsFzDFqQHK7LNSm

Shigoto-dev19 · 2026-02-02T15:44:34Z

Funded B62qn9f2eQdHyzX3SaqzZ2bPex5xD2vRTZtEb7D4yRVNo7Mm9tDdrrA.

See https://mesa.minaexplorer.com/transaction/5Jus3kzqXkGxfyqRjg38TkKga12gY2zeFuaNTmr3xCtxeKt8w9vD

Shigoto-dev19 · 2026-02-02T15:45:27Z

+        // the block time on mesa is currently longer than the average 1.5-3min, so its better to target a higher block time
+        // fetching an update every 20s is more than enough with a current block time of 1.5 min


I didn't change the comments and the poll configuration taking into consideration the network instability.

Shigoto-dev19 · 2026-02-02T15:47:19Z

+ * **Breaking change:** The `berkeley` option has been removed and replaced with `legacy`.
+ * - Old: `hashPayment(signed, { berkeley: true })` (use current hash)
+ * - New: `hashPayment(signed, { legacy: false })` (use current hash)


I am not sure if the breaking change comments are necessary; I just added them as a notice especially regarding the inverted logic.

i dont think breaking change comments make sense since they are meta comments that only really matter if you look at transitions of versions rather than the code itself. the code itself is not breaking; its only the iteration of versions. if you use this function without having ever looked at the previous version you wouldn't know the difference, hence it's not breaking to you at all

Removed JSDoc comments in c09bcfb

Trivo25 · 2026-02-03T07:34:29Z

  hashPayment(
    { data, signature }: SignedLegacy<Json.Payment>,
-    options?: { berkeley?: boolean }
+    options?: { legacy?: boolean }


its a little bit confusing to swap the logic around and may be more disruptive than needed. why not just call it latest instead of legacy? that way we don't have any change in behaviour, it's not called berkeley anymore and it's set up properly for future hardforks

Makes sense 👍

Fixed in c09bcfb

Trivo25 · 2026-02-03T07:37:01Z

  const minaGraphQlEndpoint = useCustomLocalNetwork
    ? 'http://localhost:8080/graphql'
-    : 'https://api.minascan.io/node/devnet/v1/graphql';
+    : 'https://plain-1-graphql.mina-mesa-network.gcp.o1test.net/graphql';


this is our testnet mesa node endpoint - not a devnet endpoint - we will very likely shut down the testnet (or at least that endpoint) after mesa is complete. the old URL should still be valid since it's for the public devnet (which will fork as well)

That's what I noted in the PR description too

Trivo25 · 2026-02-03T07:37:06Z

    archive: useCustomLocalNetwork
      ? 'http://localhost:8282'
-      : 'https://api.minascan.io/archive/devnet/v1/graphql',
+      : 'http://mesa-archive-node-api.gcp.o1test.net',


Trivo25 · 2026-02-03T07:37:34Z

+import { PublicKey, Types, fetchAccount, fetchLastBlock, setGraphqlEndpoints } from 'o1js';

-setGraphqlEndpoints(['https://api.minascan.io/node/devnet/v1/graphql']);
+setGraphqlEndpoints(['https://plain-1-graphql.mina-mesa-network.gcp.o1test.net/graphql']);


same as above - we should take note to change (and deploy the example) right after devnet forked

Trivo25 · 2026-02-03T07:39:35Z

 * Requests the [testnet faucet](https://faucet.minaprotocol.com/api/v1/faucet) to fund a public key.
 */
-async function faucet(pub: PublicKey, network: string = 'devnet', headers?: HeadersInit) {
+async function faucet(pub: PublicKey, network: string = 'mesa', headers?: HeadersInit) {


devnet and mesa are not the same thing. mesa is the name of a hardfork and devnet is the name of a network; mesa will be deployed to devnet so devnet should still be valid later since the network will continue to exists only with a different protocol version (mesa)

This answer the question in mind from what I noted in the PR description.

Fixed in da16a24

Trivo25 · 2026-02-03T08:02:52Z

Mina.faucet() is currently non-functional: the faucet endpoint (faucet.minaprotocol.com) is deployed behind Vercel, which intercepts programmatic fetch requests from Node.js with a 429 status and returns an HTML security checkpoint page instead of JSON. The o1js client code is correct; this is a server-side issue where Vercel's bot protection blocks automated/non-browser traffic before the request ever reaches the faucet API.

Some examples like the voting/run-mesa.ts example generate random keys and rely on Mina.faucet() at runtime, so it will not work until the faucet issue above is resolved.

It worked in the past! please create an issue for that - this should be fixed

Trivo25 · 2026-02-03T08:06:30Z

Mina.faucet() is currently non-functional: the faucet endpoint (faucet.minaprotocol.com) is deployed behind Vercel, which intercepts programmatic fetch requests from Node.js with a 429 status and returns an HTML security checkpoint page instead of JSON. The o1js client code is correct; this is a server-side issue where Vercel's bot protection blocks automated/non-browser traffic before the request ever reaches the faucet API.

Some examples like the voting/run-mesa.ts example generate random keys and rely on Mina.faucet() at runtime, so it will not work until the faucet issue above is resolved.

It worked in the past! please create an issue for that - this should be fixed

Oh no, its not Vercel blocking it - we now require a zk captcha as part of the facuet because bots used to spam the faucet

Shigoto-dev19 · 2026-02-03T08:24:41Z

Mina.faucet() is currently non-functional: the faucet endpoint (faucet.minaprotocol.com) is deployed behind Vercel, which intercepts programmatic fetch requests from Node.js with a 429 status and returns an HTML security checkpoint page instead of JSON. The o1js client code is correct; this is a server-side issue where Vercel's bot protection blocks automated/non-browser traffic before the request ever reaches the faucet API.

Some examples like the voting/run-mesa.ts example generate random keys and rely on Mina.faucet() at runtime, so it will not work until the faucet issue above is resolved.

It worked in the past! please create an issue for that - this should be fixed

Oh no, its not Vercel blocking it - we now require a zk captcha as part of the facuet because bots used to spam the faucet

yes yes, it's about the captcha challenge.

Shigoto-dev19 · 2026-02-03T08:56:12Z

It worked in the past! please create an issue for that - this should be fixed

#2756

querolita · 2026-03-24T11:26:56Z

What's the state of this PR? Should we spend some time on it this week? @Shigoto-dev19 @Trivo25

Shigoto-dev19 · 2026-03-24T12:29:55Z

What's the state of this PR? Should we spend some time on it this week? @Shigoto-dev19 @Trivo25

@querolita Mainly update Mesa node and archive endpoints to use public URLs instead of our Testnet ones.

Trivo25 · 2026-03-25T04:52:03Z

We should be able to just merge this into develop-3.0, as that branch will become our new main

Shigoto-dev19 added 3 commits February 2, 2026 17:22

Update Mina node and archive endpoints and explorer URL to Mesa

8e1ca13

Rename examples file names to reference Mesa instead of Berkeley

e60cf25

Replace "berkeley" hash option with network-agnostic "legacy" in mina…

248a6f2

…-signer

Shigoto-dev19 requested review from a team as code owners February 2, 2026 15:19

Shigoto-dev19 requested a review from 0x471 February 2, 2026 15:19

Shigoto-dev19 added the no changelog label Feb 2, 2026

Shigoto-dev19 requested a review from querolita February 2, 2026 15:19

Shigoto-dev19 removed the no changelog label Feb 2, 2026

Update changelog

efb8bc5

Shigoto-dev19 force-pushed the shigoto/refer-mesa-in-examples branch from 82041b5 to efb8bc5 Compare February 2, 2026 15:27

Shigoto-dev19 requested a review from Trivo25 February 2, 2026 15:27

Shigoto-dev19 commented Feb 2, 2026

View reviewed changes

Trivo25 reviewed Feb 3, 2026

View reviewed changes

Shigoto-dev19 added 3 commits February 3, 2026 12:25

Replace "legacy" hash option with "latest" in mina-signer

c09bcfb

Update changelog again

db9749c

Revert faucet default network to devnet

da16a24

		// the block time on mesa is currently longer than the average 1.5-3min, so its better to target a higher block time
		// fetching an update every 20s is more than enough with a current block time of 1.5 min

Uh oh!

Conversation

Shigoto-dev19 commented Feb 2, 2026

Changes

TODO

Notes

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Shigoto-dev19 Feb 2, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Trivo25 commented Feb 3, 2026

Uh oh!

Trivo25 commented Feb 3, 2026

Uh oh!

Shigoto-dev19 commented Feb 3, 2026

Uh oh!

Shigoto-dev19 commented Feb 3, 2026

Uh oh!

querolita commented Mar 24, 2026

Uh oh!

Shigoto-dev19 commented Mar 24, 2026

Uh oh!

Trivo25 commented Mar 25, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Shigoto-dev19 Feb 2, 2026 •

edited

Loading