BetterHash Mining Protocol BIP

[buwaytress]

Just read a little on Corallo's draft BIP: https://github.com/TheBlueMatt/bips/blob/betterhash/bip-XXXX.mediawiki that seeks to separate pool payout info channel from work info channel.

Summary of main motivation:
- more diverse block templates with miners allowed to construct their own - less risk of pool restricting protocol upgrades

Stratum is "difficult to implement and poorly documented" - you guys agree?

[-ck]

All great ideas... but it's 6 years too late and the reality is... antpool and S9 miners.

[aliashraf]

Just checked the BIP, thanks for sharing.

Good motivation, actually, almost in the exact day and hour that this BIP is committed to github (few minutes sooner?), I started a topic here regarding the same issues! I was surprised even more when I found both my article and this BIP share the same idea of separating coinbase transaction and the rest of the transactions committed to the Merkle tree. I take it as a very encouraging sign  

Now bad news:
Unfortunately the BIP fails to achieve decentralisation improvement in a considerable fashion.
Imo, it is due to the fact that this BIP is a conservative, second layer tweak to Stratum to help miners building the work by choosing the transactions they wish to confirm from sources other than the pool.

Sounds good but it is useless, I afraid.

Besides keeping the pool  in charge of accounting, which is problematic anyway, by keeping its role as a hub that can accept or reject miner's submitted shares, it keeps the power to enforce its policy. Specially when it comes to signaling and governance issues, we would have the same situation with pools as always. Nothing is going to be changed by this BIP.

So, sorry my conclusion is: Good motivation, weak design because of the definite lack of courage to go deep down in the PoW itself instead of remaining in the shallow area of Stratum.

[buwaytress]

Just checked the BIP, thanks for sharing.

Good motivation, actually, almost in the exact day and hour that this BIP is committed to github (few minutes sooner?), I started a topic here regarding the same issues! I was surprised even more when I found both my article and this BIP share the same idea of separating coinbase transaction and the rest of the transactions committed to the Merkle tree. I take it as a very encouraging sign  

Now bad news:
Unfortunately the BIP fails to achieve decentralisation improvement in a considerable fashion.
Imo, it is due to the fact that this BIP is a conservative, second layer tweak to Stratum to help miners building the work by choosing the transactions they wish to confirm from sources other than the pool.

Sounds good but it is useless, I afraid.

Besides keeping the pool  in charge of accounting, which is problematic anyway, by keeping its role as a hub that can accept or reject miner's submitted shares, it keeps the power to enforce its policy. Specially when it comes to signaling and governance issues, we would have the same situation with pools as always. Nothing is going to be changed by this BIP.

So, sorry my conclusion is: Good motivation, weak design because of the definite lack of courage to go deep down in the PoW itself instead of remaining in the shallow area of Stratum.

I'd have responded earlier but the last time I checked this thread, I didn't see this comment, I guess it was deleted and then restored? The coincidence in timing was what actually caught my attention, though I suppose it's quite normal for interest in certain topics to converge in this space.

I do feel that almost every proposal (including yours) has generally been met by a lack of enthusiasm, mainly because these fixes appear to be too late. The BIP is, as you say conservative, but that's keeping in character with Bitcoin core developers. But if I understand correctly, this BIP is a rewrite/replacement of Stratum, and would help with signalling/governance, because miners can now construct their own block templates with new protocol upgrades reflected.

[Raize]

Stratum is "difficult to implement and poorly documented" - you guys agree?

I'm not sure I would say Stratum is bad, but I would like to see a new protocol at least considered. Now that we have open source firmware for the S9s to activate ASICBoost and mine freely on other pools I think there will be more interest in this sort of decentralization and possibly more interest in pools that give miners more control.

It's worth considering looking at a new protocol at least, despite whatever political pushback may be expected, there might still be receptiveness to it given enough innovation. I'm glad someone gave this one a look. I'm a bit ashamed to say that I didn't even notice this BIP back when it was originally released.

[minefarmbuy]

https://twitter.com/minefarmbuy/status/1160279470214828032?s=20

A short thread on "betterhash".