When scraping a non-mainnet chain, why do we require a mainnet RPC?

[tjayrush]

You're right. It does require an RPC endpoint for mainnet (but it can be a slow, public one if you're scraping a different chain). There's a couple of issues here:

1. It should fail WAY more gracefully.
2. The docs should report this requirement.

I wrote this issue: #9 (comment)

I guess there's two way to handle this:

1. Require the end user to have an RPC for Ethereum mainnet,
2. Assume that, if there is no RPC for mainnet, and the scraper is proceeding on a different chain, that there is no downloadable manifest for that chain, so proceed with a re-build from block zero.

[tjayrush]

https://gist.github.com/michaelbnewman/49717d8044dc2875f00b2adba8fa82d6