Mine to local Bitcoin Core GBT

[Nexus9090]

Not sure if this is an appropriate forum to post idea requests, since I'm new to this I'll give it a go here. At worst you can delete it or point me to the right place.

Would it be possible to add mining to a local bitcoin core via GBT instead of just Stratum protocol pools.
Yes, I've heard all of the arguments about pools having lower latency etc etc and to be honest I'm not convinced.

Anyway, It'd be nice to have if there was an option to mine locally to bitcoin core.

Many Thanks

Garry

[mutatrum]

I've been playing with this idea as well. The ESP32-S3 should be fast enough to do the block assembly, but this needs to be verified.

[Nexus9090]

I think it would be a good feature to add. I'm still not convinced pools offer and real advantage especially with low latency fibre to home these days. It may have been true 5-8 years ago but home networking has moved on since then.

I'd really like to see mining to bitcoin core GBT implemented.

[skot]

How much work is required after getblocktemplate? If it's a lot it seems like this would be more suited for the node..

[mutatrum]

• Config of bitcoin node
• Poll getblocktemplate
• Construct mining work, mainly the calculation of the merkle tree
• Submit a valid block

Rough estimate for work prep is a few 100ms.

There's some extra's that make it better:

• Subscribe to new block header over ZMQ to force new work and reduce polling rate
• Custom payout transaction (not sure how to handle this)
• Caching tx hashes, reduces work prep time
• Listen to new incoming tx over ZMQ, either to include high fee txs immediately or to trigger a fresh template
• One funky idea is to expose a stratum API
• A list of other bitcoin nodes/submit URLs for redundancy and faster propagation

So, it's doable, the S3 can handle it especially with the extra ram.

It's a silly idea, and on a lot of fronts it doesn't make sense. But somewhere deep down, it also makes total sense to do this.

“We choose to getblocktemplate, and do the other things, not because they are easy, but because they are hard.” 🚀

[Nexus9090]

Thinking about this again.

At the end of the day using a pool is about trust and can anyone really be trusted given the value of a block these days?

Like you say somewhere deep down there's always an element of doubt where pools are concerned so it makes total sense, at least if you've got the option to mine locally to GBT in a decentralised way how things used to be done, you know you've only got yourself to blame if it doesn't go well.

Anyway, I do hope you'll be able to make progress with it.

I wish my software development skills were better and I'd wade in to get it done but my skills are better with HW development since that's what I've done most of my life.

Thanks for your efforts.

[benjamin-wilson]

I think this is more appropriate for hash-command. We're loading up a lot of work for the s3 already, and core has a limited number of rpc threads so having a bunch of bitaxes could overwhelm.