code-substitute: Switch from block_hash to block_number

[bkchr]

This will make it easier for light clients to work with the code-substitute.

For more information on this see: #10589

Closes: #10589

polkadot companion: paritytech/polkadot#4667

[tomaka]

This is still not ideal, because if there's a code substitute at block 10, and a light client warp syncs to block 15, it will have to download the runtime at block 10, the runtime at block 15, compare them, and if they're equal then do the substitution.

Basically the problem is that the client has no way to know whether the substitute is still relevant.
If there's a code substitute at block 100 and we warp sync to block 9000000, even though it's likely that there's been a runtime upgrade in between, the light client has no way to know and will still have to download the runtime of block 100 to check whether it is the same.

[bkchr]

This is still not ideal, because if there's a code substitute at block 10, and a light client warp syncs to block 15, it will have to download the runtime at block 10, the runtime at block 15, compare them, and if they're equal then do the substitution.

We already had discussed this once :D You don't need to download the runtime at 10, because you should use the wasm binary provided in the chain spec. You could, as we are doing it, check the spec_version on startup and then you will just need to compare them against the spec_version of 15.

Basically the problem is that the client has no way to know whether the substitute is still relevant.
If there's a code substitute at block 100 and we warp sync to block 9000000, even though it's likely that there's been a runtime upgrade in between, the light client has no way to know and will still have to download the runtime of block 100 to check whether it is the same.

Same as above, you have the binary in the chain spec. You don't need to download it, because if you download it you will also not get the correct one. (You will get the one from on chain, while you want the one that is present in the chain spec)

[tomaka]

Haven't review the code, but 👍

[bkchr]

bot merge

[paritytech-processbot]

Error: It's not possible to meet the minimal approval count of 1 in paritytech/polkadot#4667