Unfortunately I was unable to attend the AMA, but have read through the log. Some excellent questions were asked and thank you synth for giving detailed answers. I have a few questions and comments of my own.
The skynode hardware is planned to use ARM, which I agree is better than x86 for our purposes. Have you considered the Risc-V architecture? Obviously it is still quite new, so there are reasons not to use risc-v such as limited availability and inmature toolchains/ software ecosystem. However it is the most open spec CPU which is ideal for skycoins goal of eliminating potential hardware backdoors. Risc-v chips may be available in limited quantities and higher prices currently, but to drive production up and prices down people need a reason to buy risc-v over arm. Skycoin could be that reason.
Another question, what layer of the OSI model will skywire protocol occupy? From what I've read it sounds like Layer 3 network protocol, however that would require some way for nodes to bridge over tcp/ip (the regular internet) unless they are directly connected. Or is skywire layer 7 application protocol, riding on top tcp/ip like most everything else?
Finally, to the people complaining about distribution, maybe it's not ideal but I've seen far worse schemes used and result in successful project. I am still eager to see the specifics regarding future distribution, but I think it is actually pretty wise of them to retain control over distribution (for now at least) because it's still an alpha phase project and protocol implementation details are still being finalized. Also I think many people may not realize that if you want to eliminate miners (and all the associated issues) from the equation, there are few ways (maybe none? idk) to secure the network without having all coins (or atleast a large portion like in PoS) generated in the genesis block. So it is a trade off, one that I think is worth it since PoW is highly inefficient and consuming massive amounts of electricity globally when there are more earth-friendly ways to secure crypto coins.
Yes. We are planning on doing RISC-V.
We are trying to get a $30 million government R&D grant for public private partnership in order to fund dual use RISC-V open source CPU ASICs. That would be competitive in price and performance with existing ARM processors. The ideal thing is to get someone else to pay for it completely, but doing it this way is slower and is based upon relationships.
RISC-V already runs on FPGA, so the engineering overhead on our side would be rather low. Cutting out the wifi, video card and all of the junk on a modern ARM processor, also cuts down the chip area and improves the yield.
China currently has domestic MIPs processors for military, government and domestic applications. Russia has also developed their own domestic ARM processors for military and government.
I think the next generation is definitely looking at moving from MIPS to RISC-V.
We have to build our own demand and consumer market, for the volumes to make sense. When we need to ship several hundred million a year, we are definitely commited to dedicated ASICs. However, we believe a competitive market for these ASICs will develop even if we do no do them themselves. It may be better, to subsidize the engineering and development costs for promising projects rather than doing it ourselves (unless we can get a government or large sponsor willing to fund the whole toolchain).
The end goal is always open source hardware. Some of the early ICO money went towards modest funding of an end-to-end completely open source FPGA tool chain. However the open source compatible FPGAs do not have the gate count to run a full CPU yet.
The cryto and routing protocol for Skywire, was designed to be FPGA native. So we may end up skipping RISC-V and going to to FPGA for the networking layer, leaving secure RISC-V processors for applications and
Skycoin's CX programming language (github.com/skycoin/cx) is being designed for RISC-V type processors with explicit load/store at each cache level and with no CPU flags (add and add and jmp if carry are separate instructions).
>is skywire layer 7 application protocol, riding on top tcp/ip like most everything else?Skywire will run over anything that can transmit bitstreams. Including HAM radio, Wifi, Ethernet, UDP, TCP/IP and even embedding chained messages in a DHT. It is even designed to work with asymmetric connection topologies that can only transmit or can only receive.
> I am still eager to see the specifics regarding future distribution, but I think it is actually pretty wise of them to retain control over distribution (for now at least)When Skycoin's market cap is ten billion dollars, then we can start doing serious infrastructure investment.
Skycoin's market cap and specifically its usefulness is tied to physical infrastructure. The velocity of money and real spending (not speculative spending, but real economic activity), is going to be a linear function of specific infrastructure that the project needs to bootstrap.
Some of these ASICs for SDR can cost $5 million to $20 million each. Some of the things we want, require expensive, highly specialized analog/digital ASICs. The RISC-V ASIC may be $30 million dollars and have no other existing consumer market, but be absolutely necessary to deliver particular security guarantees to network stakeholders.
For custom PCB boards and more control over the hardware toolchain and to bootstrap and open source ecology, among several companies is going to be at least $120,000 to $600,000 but offset by consumer orders.
We may also need to get full time staff, to compile our own linux distribution and for preparing deterministic build setups for packages and completely removing binary blobs from the kernel.
Security architecture and hardening is going to be a large issue, because I do not think the MtGox style exchange heists are over yet. The Intel and AMD remote management engine back doors are in the public domain now and it is going to be almost impossible for anyone running servers on AMD or Intel hardware to keep their coins from being looted and pillaged.
Open source hardware wallets and RONJA hardware is another area we can invest in.
Distribution:Skycoin's development model is basic upon ASICminers, because their approach was very successful. They were the first cryto-stock to exceed a hundred million dollar market cap and they did not even have a blockchain to track equity ownership because Ethereum did not even exist yet and I think Bitcoin was still at $8.
The more Skycoin are "distributed", the less each of them is worth. For instance
- ASIC miners takes money and invests in hashing power and pays a reward to their investors
- A. ASIC miners sells 1 million shares at $5 a share, and buys 50% of hashing power
- B. ASIC miners sells 2 million shares at $5 a share, and buys 50% of hashing power
- In both cases the value of the shares went up hundreds of times and the reward pays off the cost of a share hundreds of times.
- However, the pie is the same size for both A and B.
- In B, the pie is divided among twice as many people, so each person gets half as much
Skycoin really should be priced as an equity, with a cap on the issuance. Issuing more Skycoin, means the pie (the value of the network and the velocity of money from real economic activity) is divided among more coins, but the value itself of all coins equals the value of the network.
If there are 100 million Skycoin issued, then the value of the network is divided more per coin than if only 10 million Skycoin are issued.
Skycoin blurs the boundaries between different types of assets, because of its design
- a crypto-coin (fixed cap)
- an equity (coins representing shares of value of network/economic activity, stock buybacks from economic operations through skywire)
- a fixed income instrument (coin-hours)
- a crop-share
- a store of value
- value derived from use
- an open ended investment fund (infrastructure development/investment)
It is going to get very confusing for several reasons
- coins are issuing coin-hours (that can be used to rent network capacity flows)
- coins are also used as a currency
- a person can "rent" (buy coin hours) or buy a share of network capacity flows (buy coinhour producing assets, equivalent to a crop share or stake of network ownership). (stocks and flows are different units and have to be treated differently or the equations do not balance or even have the right units).
- there are buybacks (a tithe from skywire operation used for skycoin buybacks)
- there is directed infrastructure investment for demand creation (both in software, the ecosystem and later in physical hardware and physical infrastructure)
This is an experiment in a new economic model.
One example, besides ASICminer is STEEMIT
- STEEMIT has miners who receive coins for generating blocks
- A percentage of the coins received by miners are allocated to a content platform
- people who create content on the platform are given given (based upon user votes and content popularity)
- The content attracts users, some of whom then on to purchase STEEMIT tokens to tip content creators
These are all struggling attempts, experiments at creating a closed economic system of consumption and production (a microcosm of society), within the context of a crypto-currency. Bitcon, STEEMIT, gnosis and Ethereum are really embryonic. No one has succeeded in creating an actually, closed, self sustaining system of consumption and production. None of these coins have engineered, designed or implemented a closed loop economic system yet.
Right now, the crypto markets are completely driven by speculation and new capital inflows (unsustainable). People are investing hundreds of millions of dollars in ICOs and buying at a dollar, to sell at $5 while no real value is being created, no real economic activity is occurring and the development teams have done nothing by market and hype.
This can only continue for as long as you have an exponentially increasing wave of new capital flooding in (and this probably going to continue for at least a decade). While the volume of money coming in to a market, is increasing at a faster rate than the money leaving, then everything can go up all the time, regardless of merit, purpose or common sense.
You are going to see people, holding vast holdings of hundreds of millions of dollars in nominal coin wealth on paper, who are unable to convert their "wealth" to anything or spend it. They are going to have $3 billion dollars on paper, but it wont be liquid because it will take them 500 years at the rate of liquidity the market will allow, to convert the coin into anything else without tanking the market.
You have to differentiate between demand that is created by speculators and demand that is created by real economic activity and use value. Speculative demand can evaporate as a matter of opinion, while demand and transactions generated by real economic activity is immutable and eternal, and exists independent of opinion (speculation).
The reason we have so many ICOs, if that coins can raise 10 million dollars, then for every $1 in marketing, they can produce $10 in new capital inflows. So everyone makes money (in the short term), but these flows are merely speculative capital inflows and do no represent real wealth or real economic activity. However, this situation can continue for as long as new capital continues to flood into the crypto asset class and may continue for a decade or two at the current rate.
Skycoin was built with a very specific economic philosophy and outlook. The people who understand that philosophy are not telling us to "distribute" more coins carelessly, they are saying "Why raise money by selling coins at $5, when you can wait a while and sell them at $50". The whole point of raising money, is that we can invest $1 and create $50 in network value. If we are only getting $0.50 on the dollar for every dollars we put in, then its malinvestment and we should not be doing it.
I have met people who are going to regret everyday for the rest of their lives that they did not buy Bitcoin when it was at a penny. I think if the Skycoin project delivers on 1/3rd of its objective, there are going to be people who will commit suicide and kill themselves, because they did not buy Skycoin when they could. And Skycoin is only designed to be a niche coin, in a specific, strategic, defensive part of the ecosystem.
So there is a compromise between a well defined targeted, specific audience and something like Dash, Monero and Bitcoin. The larger positions are often not defensible and any innovation (in marketing or technical) is easily and rapidly copied (open source software). Meaning that there are only a few defensible or long term stable niches in the markets.
If you actually look at what Skycoin is doing and can understand it. Then no one would want to copy it, they would not want to commit the resources to it, they would not be able to find the resources and even if they could, there would be far easier ways to make quick money. Why spend six years building something long term, when you can do zero development and raise 250 million dollars with an Ethereum token tomorrow, without having to deliver anything?
The reason I think people will literally kill themselves for not investing in this or that crypto, is fairly simple.
If you look at how many hundreds of trillions of dollars there are offshore being hunted down and money in toxic assets, non-performing debt and speculative bubble that have not popped yet, the number is beyond comprehension. In some scenarios if governments do not freeze accounts and do "bail ins" and grab money out of circulation, restrict capital flight, or suspend asset sales we could see a bubble phase of fleeing capital, where cryptos end up priced at a multiple of total world assets (or atleast at levels comparable to the value of all gold and government fiat).
You have banks and scam artists looting whole countries like Greece with engineered debt scams and seizing pension funds, then sizing up other countries and asking "Can we do this to Germany?", "Could we get away with this in America?", "Can we do this to Russia again like we did in 1994?". These people are plundering and looting whole nations, to the degree that they have so much money they have already bought up all the liquid real assets available.
I have no idea how they are going to fit their trillions of dollars of loot and plunder into this 80 billion crypto asset class, but that is what they are going to do. When the whole asset class goes up 100x, anything you buy is going to go up. You are going to see people peddling the peak of human retardation, becoming billionaires overnight.
I think under some scenarios, we could be entering a bout of temporary insanity and the largest asset bubble in human history, for at least the last five hundred years. At the same time as the largest financial crisis and catastrophic implosion of empire in the past five hundred years.
