Quote from: MalReynolds on October 21, 2014, 07:29:51 PM

Quote from: Litoshi on October 20, 2014, 08:39:05 PM

With PoT, we reward forgers for keeping the network up. Otherwise, we would either need a continuous central system for network continuity, or use a Proof of Stake system like NXT, where a few whales get all the forging earnings. The PoT value resets after a successful bloc is forged, giving every up-node a turn at forging. Any coin without active miners has a network continuity problem.

PoI, Proof of Identity, is used to assure a user that a merchant has been verified by Cryptsi, and to reward merchants for using XCR:

An application, 1000XCR, and verification by10 forgers are necessary to become a verified merchant. The 10 forgers split the 1000XCR equally. After this, the merchant will be identified by a unique address and will receive half of the transaction fees generated from his sales.

The whole point of PoT is that (1) everybody is motivated to keep a node going as long as possible and (2) everybody who HAS kept a node going has an EQUAL CHANCE to forge a block whether you are a whale or a minnow. Can PoT forgers be rewarded enough to keep them motivated to forge Crypti? Let's explore that question.

First, how much does an always-on personal Crypti node cost? Let's go with a $75 2W PCduino amortized over three years using $3 per year of electricity (17.5 KWh per yr at $0.15 per Kwh) and $20 per month for a continuous DSL line running 0.75 to 3 Mbps for bandwidth. The computer and power is thus only around $3 per month; the key cost is bandwidth at $20 per month. Let's say a personal always-on Crypti node is thus $25 per month or $300 per year.

There's only 1440 Crypti blocks to be forged in a day. Roughly speaking, given the EQUAL CHANCE aspect of PoT:

Your single node in a 100 full-time node network captures 14 blocks per day, 5110 blocks per year; breakeven requires $0.06 per block
Your single node in a 200 full-time node network captures 7 blocks per day, 2550 blocks per year; breakeven requires $0.12 per block
Your single node in a 500 full-time node network captures 3 blocks per day, 1095 blocks per year; breakeven requires $0.27 per block
Your single node in a 750 full-time node network captures 2 blocks per day, 730 blocks per year; breakeven requires $0.41 per block
Your single node in a 1500 full-time node network captures 1 block per day, 365 blocks per year; breakeven requires $0.82 per block
Your single node in a 3000 full-time node network captures 1 block EVERY OTHER day, 182 blocks per year; breakeven requires $1.64 per block

So forging fees for a 100-3000 Crypti node network needs to be somewhere between $0.06 and $1.64 per block JUST TO BREAK EVEN with PoT forging.

Let's say that the forging fees are adjusted so that they are around 2 % of the total amount of transactions going across the Crypti network. ATM and credit cards fees are around 3%, so this makes using Crypti a bargain compared to using a plastic card.

A 2% fee means to get $0.06 in forging fees per block, the AVERAGE block must have 50 times that much in sales transactions, or $3 per block. That's Crypti sales of $4320 per day or $1.57 million per year.

A 2% fee means to get $1.64 in forging fees per block, the AVERAGE block must have 50 times that much in sales transactions, or $82 per block. That's Crypti sales of $118K per day or $43 million per year.

So just to break even with a competitive 2% forging fee, a 100 node network needs to have annual sales of $1.57 million or $15.7K annually PER NODE.

And just to break even with a competitive 2% forging fee, a 3000 node network needs to have annual sales of $43 million or $14.3K annually PER NODE.

This $15.7K - $14.3K range is really a single figure per node regardless of network size, the difference is due to roundoff error in each scenario.

Bottom line, if you assume $30 per month to run a node and specify a 2% forging fee per block, each and every node needs to support an additional $15K in sales per year just to break even in the cost of running the node.

Now, here's the gotcha - the person running a node that generates $15K per year in sales is not a "miner" or a "forger" - they must be a VENDOR by definition. A vendor that must add their own (always-on) Crypti node onto the network to allow processing of Crypti as a currency in a transaction.

Note how in order to keep a reward forging network adequately funded, the size of the network must scale in lockstep with the sales revenue it is supporting. What mechanism is Crypti going to use to ensure THAT? Because if people become forgers at a whim "just because I want to forge some XCR" or "just because it makes the Crypti network stronger", then the economics of the network are DISRUPTED.

FOR A COMMERCE BASED CRYPTOCOIN THAT USES FORGING AS A REWARD, THE NUMBER OF NODES MUST BE KEPT IN LOCKSTEP WITH THE INCREASE IN SALES REQUIRED TO SUPPORT IT.

Just how does Crypti propose to do THAT?

Of course, if you were to DROP PoT....and DROP forging...and JUST GO WITH PoI / PoP... then a VENDOR could add a node anytime they wanted, and write off $30 per month as a cost of doing business in running their own node, and they wouldn't have to hit a target of $15K in sales per year to guarantee some forger who isn't them stayed motivated...

Eventually you guys will acknowledge my point and realize it's VENDORS, not FORGERS, that are the key to running a successful Crypti network. And on that day, you will drop this crazy insistence on perfecting PoT to motivate forgers get on with making Crypti the key cryptocurrency for commerce...

Interesting essay Mal. I have a few comments:

the key cost is bandwidth at $20 per month

This assumes that a forger will have a DSL line dedicated to just the node. I doubt that...... most if not all forgers will use the same DSL or work T1 lines that they use for all other internet communications. Merchants wil already have DSL lines set up for their credit card transactions. Homeowners already have their DSL set up, so a node connection presents no additional bandwidth fees. That lowers your estimate of $300 a year to $60 a year.

There's only 1440 Crypti blocks to be forged in a day

]FOR A COMMERCE BASED CRYPTOCOIN THAT USES FORGING AS A REWARD, THE NUMBER OF NODES MUST BE KEPT IN LOCKSTEP WITH THE INCREASE IN SALES REQUIRED TO SUPPORT IT

But you assume that the only transactions going thru the node for the next year are the Cryptsi Node Reward Program transactions. There will not only be other transactions, but each block can contain 254 transactions. So we ar elimited to 365.000 or so transactions a day. A reward for forging 1 million XCR is 5000XCR.

Today there were ten blocks that went thru with 100XCR fees, and one with a 1000XCR fee. That was the devs making a merchant account for the Node Reward Program. Lucky Forgers.

Additionally, you are also assuming that XCR will stay the same price relative to the USD as it is today. I doubt that very much, and so do our frequent bloggers.

For a merchant to become a verified merchant, they must apply and pay a 1000XCR fee. Then 10 forgers must verify the merchant. Those 10 forgers earn 100XCR each.

Of course, if you were to DROP PoT..

PoT is the backbone of XCR. The team here has been working on the specs for PoT the past few weeks. The current plan calls for it to be the main algo to determine what node forges a block. The other algos are just frosting on the cake, adding a little to the value, but not a lot.

PoT will not allow a long-time up node to dominate the forging. Each node is given one point for every block it is up, and once it forges a block, the PoT value is reset to 0. That allows the other nodes to in the queue to forge, and prevents a type of attack.

The other value for merchants is the sharing of the transaction fee. Even your wildly estimated cost of $300 a year ($1 a day maybe) can be made up with just one daily earned fee from a $200 sale.