Actually I’m more interested in being able to pool ASAP with others here with a good CPU miner that works with Equihash more so than a GPU miner. But either way I would be happy with some progress.
the latency of doing a single iteration will prevent effective pooling until the diff goes WAY up.
I don’t follow. What does latency and / or difficulty have to do with the benefits of pooling?
1 Like
No one has mentioned RAM. The blog says 1 GB per thread has been forced. I believe the whole reason for using Equihash is too reduce the profit difference between big miners and little miners. I believe it shifts away from electricity expenses as being the biggest cost factor to the initial capital expense of buying more RAM. The blog also says it should not give an advantage to ASICs. This is a fantastic idea.
The need for pools only comes from higher and higher competition which should not be a killer problem at first, when $ per coin is low. If RAM if the largest expense and you spend $1000 on RAM and run it 24 hours/day, then at some point you should get at least $1000 in coin plus your electricity expenses, unless everyone has bought too much RAM and no one is getting a return on their investment. Since it is not likely to jump over $1000 per coin so quickly, you should not need to join a pool to make sure you do not overshoot or undershoot the return on your investment.
If RAM is the primary expense instead of electricity, big miners will not have a much higher competitive advantage (efficiency in the return on their mining investment). I think you only need to join a pool if the per coin value is getting high, close to your mining equipment expenditures, or your timeline for partial return on your investment is short.
It’s interesting to note that due to a RAM glut in 2013, prices only now are below the 2013 prices. I wonder if there is a hold up to the “moore’s law” for RAM.
DDR3 Ram price chart from 2012:
http://camelcamelcamel.com/Corsair-Vengeance-2x8GB-Desktop-Memory/product/B006EWUO22?context=browse
Wei Dai’s b-money idea included thoughts on connecting the computational system’s cost, both the hardware and energy, reflecting the age-idea that currency should reflect a basket of commodities. Hayak, Keynes, and Richard Graham are a pretty good spectrum and they all thought a commodity basket was the best currency. Hayak said it was better than Gold. Keynes tried to get the world off the dollar, as far away from fiat as possible after WW2 by pushing for the Bancor. Warren Buffett’s mentor was pushing it in the U.S., but got no political traction even as 100% of economists seem to have always thought that is what a currency should be. So, the ideal cryptocoin will require a basket of commodities to mine and maintain it, all fundamentally based on Gibbs free energy and the Landauer limit, but this goes beyond typical crypto and economic discussions, even as it is the core of the connection. So if RAM were truly in some magical way reaching a limit, it’s not (all) a bad thing for Zcash.
There is no reducing the profit difference between big and small miners without drastic market interference. Attempting to make an ASIC solution infeasible is about having a level playing field as the market for them is too small and therefore too easily dominated by the big mining operators who unavoidably drive network centralisation.
Until there’s a market for zec, attempting to do projections on ROI for mining hardware outlay is highly speculative.
2 Likes
Your 1st and 2nd sentences seem contradictory. My view is that making ASIC infeasible (to level the playing field) is a drastic market interference, for a good reason. A free market evolves towards concentration of wealth and monopolies. Democratic voting creates a more level playing field (1 person = 1 vote) by causing government to write the rules (algorithmic protocol) to bias the free market (capitalism) away from concentration of wealth, towards socialization.
Developers are the “governing employees” that make Zcash more democratic, more social. Equal access to coins based on investment expense is a fair market, made possible by a “government” (algorithm). It’s a democratic idea, 1 vote = an equal investment expense.
It’s true my ROI is entirely speculative. My point was to show small miners will lose only if big miners lose. Equihash is a good system for preventing wealth concentration at the outset: if ASICs were feasible, special interests could be a problem at the outset. Look at bitcoin’s miners.
The economic/democratic problems I’m about to describe for any constant-quantity coin like Zcash are long-term. Given no other option, not even in theory, I’m choosing Zcash to be in the 1% instead of the 99%. All economic woes are a consequence of the physics of evolution. There is no solution. Humans are not capable of subverting the physics of evolutionary progress towards higher efficiency.
Anonymity that prevents government from unjustly targeting individuals is a form of wealth distribution. It can take away power of special interests who try to subvert democracy. But it can also prevent government from performing the good aspects of its democratic role. Among other things, if a constant quantity coin becomes the default currency, compound interest always results in wealth concentration in the lenders. Gold historically works only in times of anarchy and war. The people needing loans also need an inflating currency (but not inflationary prices). Ideally all the interest charges should be used to finance all of government. Interest should be the only tax, and that tax should fund the expansion of the society (which is ~ equal to its need for the currency) so that there is no inflation in prices or wages, which keeps contracts in that coin valid (think ETH).
Equitable computation is a more intelligent network (solves problems) for deep reasons, despite being less efficient. In A.I., the most effective systems evenly distribute computation. The constant quantity of total available CPU time and memory space is the “currency” that needs to be distributed to grant access. There’s a conversion factor between the CPU time and memory space that is not unrelated to Einstein’s meters=icseconds based on Landauer principle. Genetic algorithms, agoric economic agent systems, bayesian techniques, & neural nets seek to redistribute computation among a wider variety of “genes/agents/nodes” and the “weighting factors” (“wiring” or “links” in the web) between them by distributing computational requirements more evenly, economizing the resources towards solutions. An unused node, gene, or web page (no links to it) and a very low price (in agoric agent-based computing) or very low probability (bayesian) are all computational elements that can be eliminated from the algorithm with minor error (a universal NAND gate with no wiring to it is the simplest example of an unused computational element).
Like everyone else interested in cryptocurrencies, I want to make as much profit with the least amount of work. Constant quantity currencies might be ideally suited for the 1% and a subversion of democracy. Wei Dai expressed a similar concern about bitcoin. That’s why he likes tracking commodities. An ideal coin would expand in lock-step with its M2-like usage to keep wages and prices constant, which keeps contracts in that coin valid, and prevents early adopters from profiting (gaining more access to society’s finite resources) without having to work for it (contributing to society).
A constant quantity currency is contrary to equitable (intelligent) economics for these reasons. It is only optimal when the resources it represents control of are a constant, as in A.I. systems constrained a specific hardware system. It will be beneficial in times of war and anarchy for the survivalists have planned ahead. Anonymity amplifies this benefit.
Just to follow up on this bounty thing:
If a dev wants to / is willing to write a pool server from scratch (rather than simply modify one of the ones that already exist), then it would be better for the community and decentralization if it used getblocktemplate rather than Stratum.
Getblocktemplate was designed by core devs specifically to help decentralize mining – but Stratum became super popular right around that same time. So getblocktemplate never really took off.
1 Like
The only way that an ASIC unfriendly proof of work is interference in a market is if that market was previously defined in terms of ASIC friendliness. That isn’t the case with Zcash.
btw Until reality matches your theories and other aspirations, they have no value.
POW was originally defined as the number of hashes performed. Selecting an algorithm that prefers RAM is a market interference. ASICs are capable of more hashes per dollar cost when the algorithm focuses on more hashes per dollar cost. So the market was indirectly defined in terms of ASIC friendliness.
It is blocking some processors from running at their maximum capability. This enables all types of processors to compete more fairly. It coincidentally saves on energy. This is the same as government busting up monopolies that were originally more efficient due to the efficiency gained from their size and dedicated investment. But once they become excessively large, they have the capability to manipulate the market to the detriment of all, such as getting over 51% hash power. This is a free market effect. Government and A.I. systems stop these monopolies by interfering in the free market. Free markets are fair (have an enforceable rule of law) only at the individual transaction level. This does not mean they are fair or efficient at the system-wide level. Maximizing hash rate is a goal needed for fast transactions, but interference in the hash rate marketplace is needed in order to assure it remains a distributed system, in exactly the same way government needs to intervene to prevent the 1% from dominating the 99%. Democracy stops this with varying levels of socialism.
Your second paragraph is equally wrong. My previous post is based on observation, not theories.
A free market is defined by there being no barrier to entry. The instant some uninvited third party (such as people doing business as ‘government’) comes between people engaging in a voluntary exchange, there is no longer a free market as it has been reduced to a regulated market.
Crypto currency networks do not care how their blocks are supplied, ie: there is a free market for hashing. Just because the Zcash developers have not provided solutions for the vast array of computational devices to be utilised in the market they are creating does not prevent other individuals from taking agency in their own lives and creating their own solutions - as hard as that may be for opponents of free markets to grasp…
To quote a particular clear headed individual re: your advocacy of government, market interference, democracy, socialism & etc, ‘Gack!’ 
1 Like
Okay, now someone please tell me at least how to establish a “conventional” mining pool for zcash?
How much the information (related to Bitcoin-based currencies mining-pool setup) is applicable to zcash?
And how much applicable it will be once zcash released?
Until there is a standalone miner, pooling of any kind isn’t possible.
I’m also interested in participating in a P2Pool setup. Any progress in the last month with getting a standalone miner set up?
1 Like
See this thread for (AFAIC) current progress:
1 Like
Interested in participating in a p2pool.
1 Like
I’d be interested in participating in a p2p mining pool as well.
1 Like
INterested in P2P Pool as well,
1 Like
Could someone share a link to Zcash mining pool deployment manual? 
1 Like
Hello just finished my pool https://zec.anopool.eu/getting-started its realy easy, now im just tweaking frontend.
So will there be any kindly posted to all cryptocommunity p2pool node deployment howto?