bug: waitForTransactionReceipt doesn't accept client with multiple possible chains about wagmi HOT 6 OPEN

shouldn't this issue be linked inside viem repo, as it was originally? The bug seems to be related to viem/actions which are part of viem repo. In the getClient.test-d.ts I need to write code like below to recreate the issue

import { chain, config } from '@wagmi/test'
import { expectTypeOf, test } from 'vitest'

import { http } from 'viem'
import { waitForTransactionReceipt } from 'viem/actions'
import { createConfig } from '../createConfig.js'
import { arbitrumNova, optimism } from '../exports/chains.js'
import { getClient } from './getClient.js'

test('behavior: viem actions', () => {
  const config = createConfig({
    chains: [arbitrumNova, optimism],
    transports: {
      [arbitrumNova.id]: http(),
      [optimism.id]: http(),
    },
  })
  const client = getClient(config)
  waitForTransactionReceipt(client, { hash: '0x…' })
})

from wagmi.

using client.waitForTransactionReceipt works fine (might be the intended usage?)

from wagmi.

If anyone wants to look into this, this is the test that you need to pass:

getClient.test-d.ts

test('behavior: viem actions', () => {
  const config = createConfig({
    chains: [arbitrumNova, optimism],
    transports: {
      [arbitrumNova.id]: http(),
      [optimism.id]: http(),
    },
  })
  const client = getClient(config)
  waitForTransactionReceipt(client, { hash: '0x…' })
})

from wagmi.

shouldn't this issue be linked inside viem repo, as it was originally?

It could, but it's less likely that Viem will have a client with a chain union for client.chain. In this case, getClient returns a client with client.chain: typeof arbitrumNova | typeof optimism. Seems like the move is to solve this downstream of Viem in getClient.test-d.ts first (since Wagmi should have type coverage for this scenario), then update Viem if necessary.

from wagmi.

@tmm it looks like in #3929 you're gonna take this on as part of a larger refactor

is there a rough timeline on when it might land? we need to bump wagmi to pull in a coinbase wallet fix on Zora, if there's a rough timeline it'll help me decide whether to wait or to refactor some @wagmi/core usage in favour of consuming the actions from a client instead

from wagmi.

@tmm it looks like in #3929 you're gonna take this on as part of a larger refactor

is there a rough timeline on when it might land? we need to bump wagmi to pull in a coinbase wallet fix on Zora, if there's a rough timeline it'll help me decide whether to wait or to refactor some @wagmi/core usage in favour of consuming the actions from a client instead

refactored out our @wagmi/core usage in favour of consuming actions via public/wallet clients instead, i'd been meaning to do that anyways

from wagmi.