Add baseFeePerGas to headers sent via EthSubscribe

[karlb]

Description

Adding the baseFeePerGas to blocks returned by EthSubscribe makes the results more consistent with the data returned on Ethereum and via other RPC-Endpoints on Celo. And it is useful!

Once we add the baseFeePerGas to the header, this implementation is not needed anymore. So this change is only temporary.

I did not check for RPCEthCompatibility since this is not done for baseFeePerGas in the normal JSON-RPC responses, either.

Other changes

• Enabled websockets for mycelo

Tested

Tested against mycelo with the following code

{
  "name": "subs",
  "version": "1.0.0",
  "main": "index.js",
  "license": "MIT",
  "dependencies": {
    "web3": "^1.9.0",
    "web3-providers-http": "^1.9.0"
  }
}

const Web3WsProvider = require('web3-providers-ws');
var provider = new Web3WsProvider('ws://localhost:9545');

const Web3 = require('web3');
const web3 = new Web3(provider)

async function wrapper(){
    const subs = web3.eth.subscribe('newBlockHeaders', function(error, result) {
      if (!error)
          console.log(result);
      else
          console.log(error);
    });
}

wrapper()

Related issues

• Fixes https://github.com/celo-org/celo-blockchain-planning/issues/59

Backwards compatibility

Adding a field to the response should not break sanely implemented consumers. IMO it is acceptable to consider this backwards compatible, although it is not true in a strict sense.

[github-actions]

Test failures:
TestTransactionIndices: core blockchain_test.go:1944: Oldest indexded block mismatch, want 100, have 67
This test report was produced by the test-summary action.  Made with ❤️ in Cambridge.

[github-actions]

Coverage from tests in ./e2e_test/... for ./consensus/istanbul/... at commit 4969b78

coverage: 49.3% of statements across all listed packages

coverage:  61.6% of statements in consensus/istanbul
coverage:  37.5% of statements in consensus/istanbul/announce
coverage:  54.5% of statements in consensus/istanbul/backend
coverage:   0.0% of statements in consensus/istanbul/backend/backendtest
coverage:  24.3% of statements in consensus/istanbul/backend/internal/replica
coverage:  64.9% of statements in consensus/istanbul/core
coverage:  45.0% of statements in consensus/istanbul/db
coverage:   0.0% of statements in consensus/istanbul/proxy
coverage:  64.4% of statements in consensus/istanbul/uptime
coverage:  51.8% of statements in consensus/istanbul/validator
coverage:  79.2% of statements in consensus/istanbul/validator/random

[palango]

Looks good to me, as a temporary measure until we add it to the header. But let's have @gastonponti have a look as well.

[gastonponti]

LGTM 👍