BUIP015 - Decentralize mining with the FAIR PoW algorithm and an user-configurable PoW setting

[freetrader]

@hellobitc0inworld : I believe you are correct that any Nakamoto-style SHA256 hard fork would start out at a low valuation.

I don't see that as a bad thing for several reasons:

1. Having a low valuation makes the fork less interesting initially for opposing miners (they can earn more by regular mining on their fork than attacking). This is important because the difficulty on such a fork would have to start low, unless it has arranged for a significant % of existing hashpower to switch over to it immediately (this will not happen under current circumstances, given the centralization of mining)

2. In the beginning, a low price can attract a lot more buyers and broaden the economic base a little. This helps to strengthen the growth of mining on the spinoff chain.

3. It may be just my impression, but the incumbents have alienated some significant players in the commercial corner of the ecosystem. I could imagine some exchanges and pools stepping (and even some ASIC mining operations) stepping up to support a viable fork that retains the POW.
For them, it would offer business opportunities - fork arbitrage, winning additional miners who support that cause etc.

It becomes a whole lot more difficult for the rest of the ecosystem to adapt quickly when the POW is changed (not that I'm saying a POW fork is not viable - it will find support too, but adoption may be significantly slower, and there is no keel-over effect possible where significant parts of the existing mining power defects).

 

[SysMan]

using bit of time: PoW FAQ, my IMHO

Keep in mind, there are almost only tech&brain limits for any computations optimization... CPU/GPU is power inefficient multipurpose chips.
This inefficiency increase FEE/Transaction cost for your network (high cost electricity & cpu/gpu mining should be covered by economical interest - otherwise ppls will not mine/ read "secure-blockchain" by anything - simple real life rule, result of work should cover spent resources).

any FPGA,GPU/CUDA/ARM or even ASICs is just optimized way for such computations - just natural evolution of task solution.
This optimizations doesn't born by itself from scratch, its not cheating, its result then engineers&investors invest their time&money taking all related long time risks & developing more optimized ways, this also creates direct investment in base-cost-market-cap of p2p-any-coin network & provides coin economy security protection barrier by computational-hashing work
If any-coin as infrastructure are enough big as p2p, enough stable, enough economically stable and represent point of interest for engineers&investors - they can construct more optimized math task solutions method - this is just part of natural technical evolution.
You may live in illusions what your laptop/desktop/server cpu/gpu is most powerful but it isn't - because it multipurpose main stream device, its not optimized & efficient for specific task.
Almost any math/memory calc's can be optimized, its just questions of time and interest - to take the risks of research & construction.
But it you like to keep your small-coin, with small-economy you can make any "new-way hashing" - until you are small and _not interesting for anyone_ - CPU/GPU will be main PoW in your coin.
Also changing the rules of network, I mean changing any rule in big game - you will loose trust of involved parties.

For example changing one PoW to another PoW: - you change the base economy rule, 1st as impact this will without any doubts will change cost of coin - because hardware protection by hashing power will be different(even if you done it smooths), transactions & coin/assets protection will be changed. You will lose part of network/community users who as miners invest their money, as next impact lose of trust, no-one after this will never invest in mining - if rule changed once it can be changed again.

Example changing PoW to PoS: - as impact increase any-coin risks, removes additional protection barrier from in-depended hardware level trx sign, and will leaves it almost on level of coins control coins "who own the most-money&nodes - own more transactions".

BTW illusion - what CPU mining will provide same gold-rush-reward like it was on start in 2012/2013 of bitcoin complitely failing - now reward will be div between 3.5 millions users.
Since 2012/2013 lot of users invest their power/resources/computers/brains, build businesses on top of bitcoin - increasing bitcoin network market-cap.

 

[freetrader]

@SysMan : I can see you are following the debate closely. Good, at least someone is paying attention, and I like that you are not afraid to give your perspective, even if our views may be different.

This inefficiency increase FEE/Transaction cost for your network (high cost electricity & cpu/gpu mining should be covered by economical interest - otherwise ppls will not mine/ read "secure-blockchain" by anything - simple real life rule, result of work should cover spent resources).

Of course, and there are successful altcoins out there who are reaping rewards while Bitcoin is shackled. They are powered in no small amount by a lot of disillusioned Bitcoiners hedging their crypto investments, in some cases leaving Bitcoin entirely.

It's not hard for people to see where Blockstream / Core's roadmap is leading.

And it's not a friendly environment for small business on the blockchain. If anything is losing part of the network/community, this is it.

CPU/GPU is power inefficient multipurpose chips.

Bear in mind that there are ASICs out there in the hands of smaller mining operations, not just $100M warehouses and larger, some of whom today subsist only because of subsidized electricity (and other factors that have little place in technical debate).

BitFury no doubt deserves credit in making ASICs available to a wider public.

You may live in illusions what your laptop/desktop/server cpu/gpu is most powerful but it isn't - because it multipurpose main stream device, its not optimized & efficient for specific task.
Almost any math/memory calc's can be optimized, its just questions of time and interest - to take the risks of research & construction.

Yes. Time for Blockstream and Core to stop Bitcoin's growth is almost over.

You may know better than nearly anyone around what it takes to play catch-up with algorithms in the ASIC development game.

But it you like to keep your small-coin, with small-economy you can make any "new-way hashing" - until you are small and _not interesting for anyone_ - CPU/GPU will be main PoW in your coin.

Few here are under some illusions that a CPU/GPU mining phase will last forever (or in the case of keeping POW - for any substantial amount of time).

You need to understand that this is not about some "gold-rush-reward like it was on start in 2012/2013". We live in the real world, our businesses are based on Bitcoin in many cases, just like yours. We need Bitcoin to grow, and Blockstream is living up to its name.

The miners who are now not signalling growth for Bitcoin are losing trust, fast.
Don't be on the losing side by not taking developments into account.
We hope your company will succeed and continue to strengthen Bitcoin, not contribute to its ultimate defeat. Even if Bitcoin is just one member in the cryptocurrency world, we love it for what it brought (and still brings) to the world. We want to see Bitcoin succeed, and retain its position at the head of the table. Altcoins have their unique selling points, and Bitcoin does not need to compete on all of them, that is true. But we signed up for the electronic cash that Satoshi described, not some settlement network.

 

[bitcartel]

Updated to include Equihash as a possible algorithm, which has been selected as the PoW scheme for Z.Cash:

https://z.cash/blog/why-equihash.html

Also, worth mentioning is ASICBoost (patent-pending):

• http://www.asicboost.com/
• http://bitcoin.stackexchange.com/questions/43579/how-does-asicboost-work

Some core developers are looking to reduce it's impact:

As part of the hard-fork proposed in the HK agreement(1) we'd like to make the
patented AsicBoost optimisation useless, and hopefully make further similar
optimizations useless as well.​

http://lists.linuxfoundation.org/pipermail/bitcoin-dev/2016-May/012652.html

Is this central planning? Protecting one set of miners over another? What about miners who develop improvements in secret? Or have advantages such as free electricity? There's a good argument here that it's unfair that CPU/GPU miners were not protected from ASIC miners, which would have preserved decentralization

 

[johnyj]

ASICboost fiasco is another prove that the fact that someone can change the protocol is the biggest flaw of bitcoin, it makes it unstable and not trustworthy, the so called consensus can be manipulated. To succeed, protocol must be set in stone from the beginning, people can only adapt, not change it, otherwise it is just a matter of time before bitcoin totally changed to something else