Andreas Antonopoulos: Decentralized consumer mining will re-appear

[arnoudk]

Some nice tweets from Andreas on the (re)decentralization of mining.

https://twitter.com/aantonop/status/709225294805078016
https://twitter.com/aantonop/status/709225457288282112
https://twitter.com/aantonop/status/709225655527821313
https://twitter.com/aantonop/status/709225930095366144
https://twitter.com/aantonop/status/709226296090341376
https://twitter.com/aantonop/status/709226539888431104

[LiteCoinGuy]

i guess he is right over the long run. still i would like to see some more software changes that help decentralisation.

[arnoudk]

i guess he is right over the long run. still i would like to see some more software changes that help decentralisation.

Me too. I'd like to see Extreme Thin Blocks (from the Bitcoin Unlimited software) implemented much more broadly - as that reduces the time needed to verify blocks AND it reduces block propagation time!

I'd also like to see more people use bitcoin - as a currency - on chain. More users would result in more people running a full node. Even if the percentage of people declines, there will still be more nodes in absolute numbers. Roger made a good argument on this. So I am for an increase to 2 MB to help decentralization further.

I would also like to see things like Lightning Network - and I would like that LN is a choice that people can make. Either do an on-chain transaction OR do a cheaper LN transaction with a little more hassle (ie, having to prefund LN, waiting to get the funds to be able to spend them, etc). LN can be nice, and it should compete in a free market with on chain transactions. If it is as advertised, it will reduce on chain transactions without the need to make the on chain transactions artificially expensive. If artificially expensive on chain transactions are needed, for LN to work. then it is a bad idea because it requires central planning and coercion to work. I, however, believe it can compete on it's own merits and SHOULD compete on it's own merits. Without crippling on chain activity.

I would also like to see SegWit implemented. Mainly because it fixes transaction malleability and THAT really is keeping bitcoin back. I've had numerous ideas that work in theory - but break in practice because there is no good way to reliably recover from transaction ids changing. However, here too, there is central planning and coercion at work - because the data that SegWit stores in its SegWitChain is discounted 75%. Some people call that compression, but that is wrong, misleading and insulting to intelligence. It is an accounting discount. THAT part of SegWit, I dislike. I believe that SegWit should compete on a level playing field, not from some artificially created superiority position. All SegWit data will have to be transmitted, evaluated and stored byte-for-byte. Every byte has to be stored. There is no compression - there are just TWO datasets. I really strongly believe that SegWit can and should compete on it's own merits, and should do that on a level playing field.

However, neither LN nor SegWit increase decentralization in my opinion. LN is a scaling network - but scaling in Number of transactions, not significantly in number of users. SegWit doesn't improve decentralization at all.

@LCG: What developments do you want to see to improve decentralization?

[LiteCoinGuy]

hey arnoudk,

you named them: all!

Decentralization is the most important thing for bitcoin, otherwise it is meaningless. so i would rather slow down adoption to build out decentralization. We should not listen to the masses who just want a quick buck.

i have a chart for you arnoudk. i was not planing to release it yet but it fits here very well



you can see that the estimated mining effenceny is dropping over the years. that is what andreas is talking about.

[arnoudk]

True, the chips are no longer developing quite as fast as these last 2 years.

This means that bitcoin mining will start to happen on "commodity" hardware once again. Everyone will be able to buy chips that are more-or-less equally efficient. Maybe not now, but this will happen.

Then, you have the best of both worlds:
1. Everyone can compete again in bitcoin mining, only cost of hardware + cost of electricity + cost of cooling + cost of maintenance are important. (and for the small blockists: cost of a semi decent internet link).

2. Botnets cannot compete, none of the hardware they could hack can mine bitcoins. Even supercomputers cannot compete. No AWS, no Google, etc can do a 51% attack.

This is a necessary phase that bitcoin is going through. All this talk about changing the PoW (which is a LukeJr "InsaneIdea TM"), just sets back the clock and destroys a large part of the ecosystem. We ARE reaching the point where state of the art ASICs will become commodity hardware now. I believe bitcoin mining will decentralize in the future (but may centralize a bit more before we get there).

[arnoudk]

still i would like to see some more software changes that help decentralisation.

(provides some arguments why no more software changes are critical)

you named them: all!

Eh, what? Thanks, but what further developments were you waiting for? They are here (in Unlimited). You could go set up a Bitcoin Unlimited node now, have the Xtreme Thin Blocks and Xpress Validation built-in.

Do you need Core to implement it?

If so - WHY? They are just one supplier. A supplier with great technical expertise and a valuable asset to the bitcoin ecosystem - but they are no Deity. They are no God. They are no Ruler. They are no Dictator. They have no power, except what we, everyone individually, give them.

You don't need Core's permission

So, I feel like I rebutted the arguments that more software changes are needed first. So time to change your position on needing more development? Or would you like to rebut my rebuttal (and challenge me to rebut the rebuttal of the rebuttal - this is gonna be fun!)

[Vaultoro]

Mining will start decentralising again for sure. USB Asics could be sold as lottery tickets to solo mine. Asics could also be heaters. Many things will come out because of the commodification of mining hardware.

[arnoudk]

Mining will start decentralising again for sure. USB Asics could be sold as lottery tickets to solo mine. Asics could also be heaters. Many things will come out because of the commodification of mining hardware.

Exactly!

ASIC space heaters... heck even ASIC water heaters (or pre-heaters) and swimming pool heaters. It does not need to be as optimized as data center operations to be efficient and electricity could be a bit more expensive. It just needs to find a spot where the heat (that is waste heat now) is actually useful. When the price of ASICs come down, there is no reason to not switch them off when they are not needed.

Maybe electric swimming pool heaters can out-compete gas heaters or heat pumps - not on cost pre-mining revenue, but maybe on cost post-mining revenue.