Kevin Peter Abate / Perfil

There are days when I feel like I have my dirty hands too deep into the candy dish. As it turns out the price of deploying a contract to Ethereum's mainnet is sky high and there doesn't seem to be anything anyone can do about it. I can understand not wanting people to upload crap code, but in reality if you dig around in etherscan.io for the source code of the most used contracts, they all appear to do the same thing with the same code. Most of the time it's code that handles issuance of tokens and payment channels, and they seem to mostly use OpenZeppelin. Don't think that I'm criticizing tried and true protocols, but for what it's worth, smart contracts are starting to look a little bit overrated. I appreciate the simplicity of the design, but what else is there? There's balloting and auctions, but if you aren't pretty much a billionaire already, it's going to become prohibitively expensive to use. The impact of ETH 2 on transaction fees and deployment costs could be limited and if it's limited then the upside may be capped for now. Meanwhile, a new library for Python that encapsulates most of the difficulty of using Bitcoin is just starting to emerge. I've been tinkering with that one on one of my cloud servers and it seems like smooth sailing for now. The setup took a couple of days and there was a little bit of digging for direction, but overall it appears that I have something workable that can at least push out some wallet information, even if Python isn't ideal for handling web requests. I still like Perl 5 and I don't feel compelled to obfuscate.

But I've been tinkering around with fuzzy logic and it looks like I may have found a winner. I'm testing on a smaller demo account and if I can go from $20,000 to $100,000, as a goal, then I'm certainly shrinking the minimum account size for feasibility of use. Velvet and Velvet Sea are tools for hands-on traders. The only thing that I really had to work on was more tightly controlling the window of opportunity and the probability of whether the EA opens a trade or not.

So like I said, there are days where I feel like I have my dirty hands too deep into the candy dish.

Have a nice weekend. I should probably take a little break here and draw things out a little bit better.

Hee haw... new update to velvet! Added fuzzy logics to handle take profit and stop loss via the commands "fuzzy.limit(float equity);" and "fuzzy.stop(float equity);"

Looks like the project cracked the top 1000. We'll have to see if it cracks the top hundred.

https://www.mql5.com/en/market/product/71985

I hesitated to release it, because I didn't like what I saw in the back testing. So it can work, and it can work well, but if it doesn't then it doesn't. I think there's account.stop([number]) and account.limit([number]) in the old ordering layer, too, so if you use both then you can run the velvet commands alongside the advisor-along, say a trendline-and if there was order.close({above|below}), then I had "order.closeall();" that you could enter as a description of a label or a textbox, you had to try both, I'm pretty sure it only works with label, but I don't exactly remember. There was an option using that command where you can specify the symbol.


I had to think about how this was supposed to work for a minute, because I thought that if I ERC20'd it would have parentheses and parameters. That, it seems, may not be supported, unless I relay everything through a single contract and ended up having to honor each channels tokens interchangeably. Work was going to be getting the front end done for the web3 interface, which for the time being will continue to be Ropsten. It is my plan to offer a more progressive price appreciation rate for accounts built with time, to offset the relay fees to get the token. Then you have a clear estimate on how long it will take to form a window where you turn a profit by offering a discount on the channel. You don't like the price point don't buy or sell tokens on that channel. Because there was slow drip and faucet sized channels and they were going to have different interest rates, because sometimes it's expensive to be poor. Those rates were more in proportion to the fees associated with the deposit procedure as well, so I didn't want to flood the market on slow drip, because I thought slow drip might produce a viable stable coin, because then people had to enter it in themselves or they wouldn't have money 0.025 Eth per day is less. The gas fees might be ten percent of that. If they can't offer their tokens for sale in a year at a profit by discounting the market then why hold it?



I just don't want to have to panic when I wake up in the morning.

Kevin

I had a few addresses on ropsten that I was looking at and I had a couple of pieces of code that were maybe going to work. First, I wrote an auditor contract which really doesn't do anything fancy except look at the code hash and the length of the code and compares both of those for two contracts, then I deployed three of them, so that they could be self-auditing. And then I deployed a much more complicated second contract that acts as a bank. I set it up with a fee schedule for transfers between bank contracts, but for transactions between accounts under the same bank contract the transfer fee should be less. So maybe this is okay if you want to have accounts tied together through a bank charter with lower fees for not going outside the charter, but it is hefty to get it mined in and that means that most of the relay fee goes to the miner under the current model, because you can't just have one contract here, with a million accounts. It's not okay to have to traverse all that data to find the millionth account and then charge that person gas. That's not okay---they'll run out of gas every time. So in order to incentivize miners to look for this contract, I've got it on Ropsten.

0x699E76537ACBA6653A44B87Bf4547D4FCCD9BC25 is the auditor with function "audit(address,address)" and that just compares the bytecode between two contracts.
0xA14097B82F564F7873F02D3F0F3257C487C9315E is the prototype for the bank contract and I can send the ABI if you want me to message it out. I thought a good base fee was 2048 gas. Charge 1/8 of that for local transfers, that's 256 gas. Charge the base fee * 5 for relays. 80/20 between the miner listed as block.coinbase and the owner given as msg.sender in the constructor call; the constructor takes only one argument and that is the base fee, or gas fee. I wanted it to be priced in gas so that it makes sense to somebody who uses Ethereum - that local transfers are supposed to be less and that relays are supposed to be slightly uncomfortable, but not too richly priced. No charges are applied to deposits or withdrawals, only relays and local transfers, along with the usual gas fees. Everything had to work first, I suppose. To relay between two bank contracts there was a low level call that caused etherscan.io to fail to recognize the source code. Chips, because there were going to be contracts that I wrote that would not accept deposits, only relays, because if it was relayed, I could audit the sender and be sure that it either had the right bytecode or that it had no bytecode and it was a user if the modifier "strict" was enabled for the contract.

Looks like the deployment aspect of the project (over a thousand dollars per contract) is over my head as far as costs are concerned. However, I will continue working to put together a proof of concept on one of the test nets. The financial constraints may be steep, but that is not going to be my business, which is a disappointment that I won't be able to profit from it at all. Imagine putting how much effort into something and then it comes to fruition, and you get nothing for it because you couldn't afford to deploy the contract yourself. Wow, life. My lucky day. Is Ethereum dead in that aspect? The counter-balance here against the price of ETH is going to be the cost to deploy a reasonably sophisticated smart contract. 1,500,000 gas at 125-150 Gwei means that at $4000-4500 you're looking at at least $800-900 to get the contract mined in. Is it possible to negotiate directly with miners about including a fee that goes to the miner of the block where the contact is deployed? Or to at least share the fees on a split? That thousand dollar price tag assumes nothing about potential miner fees that could be collected indefinitely into the future. I had a lot of things to think about in terms of work-arounds as far as getting these privacy node contracts deployed to the mainnet, because I was not willing to give up. It was roughly a hundred lines of code and that's it. At this point, I had to face the possibilities that either the cost of deployment puts it out of reach, or that if I do get it deployed I won't get anything out of the project at all, which is par for the course so far.

I've been working on a contract that looks like the makings for a candidate privacy layer for an Ethereum token and I have a supply mechanism that I think is *very* workable. The other day I was working on a firewall as well and I think that firewall is gonna end up being a double edged sword, because the firewall is very dependent on checking the value of both tx.origin and msg.sender against a list of addresses, but the privacy layer for depositing never uses tx.origin. The firewall only interacts with the privacy layer, and then the token channel is guaranteed funding by the firewall. Clients will "buy" the token outright and that token is insolvent unless they are able to resell them on a provided market supported by a baked in mechanism for price appreciation, that includes a fixed price for token generation at a linear rate governed by a maximum purchase and a lockout period.

Like I said, I've been coding for most of my life and this comes as second nature. I thought about how this would be possible back during the recession and I wondered to myself what money really meant and what it was really worth. The gas fees are going to be obscene for significant purchases on the insured market; it could be that more than 5% of the amount you spend goes to gas fees, but if you buy at that insured market with the maximum purchase, you're guaranteed a refund less the gas fees. The mechanism for constraining supply is that tokens are only minted when they are purchased from the insured market and the price of the uninsured market depends initially on what price hodlers want for their insured prizes; the price is going to appreciate on the insured market at a fixed rate. I also thought that it made more sense to allow these parameters to be not singularly fixed, but set separately for various token channels. Like say one channel has a maximum insured purchase of 0.025 Eth/hr (~$100/hr), with price appreciation of 0.1% per day. That's 42% APY baked in, but on another channel you might have a token that allows up to 5 Eth / hour and it might have only 0.01% price appreciation per day, which is more like 3.7% APY.

These are my plans, and then I was going to retire the script I had written to the codebase, and explain how C3PO and R2D2 work and how easy it should be to finish the process of importing the indicators and creating custom commands for velvet objects described by C3PO in Metaquotes. I just... I don't know if I wanna cross that bridge just yet.

I have another confession: I was going to release an expert advisor that's free and based on that indicator only a few people tried out. All I had to do was type in the description.

I have a confession: I've been toying around with hardhat to write a smart contract that could be used to govern a proof-of-effort stablecoin. But right now I haven't finished the code and I haven't yet gotten my feet wet in estimating the gas fees for it. I thought that I had a pretty nifty concept that, if nothing else, makes a valid and relevant statement about the adoptability of different digital or virtual currencies for transacting daily commerce. I might be able to launch a token sub currency if I can configure one more aspect of its operations, which could happen fairly quickly, but then as far as making more estimations on gas fees for execution on the EVM, adjustments given those measurements, and developing an interface to obtain this token once it is, in fact, deployed, I'm still a bit lost on that process, but I have a solid piece of it coded that hardhat has compiled into an artifact. Once I do get all of this done I still had to come out with a decent whitepaper to explain my intentions and the unique way the contract will govern issuance of tokens and a description of how the secondary market will operate.

Cheers.

I think I may well begin testing this setup on half-hour and hour ordering intervals with a significantly smaller demo account, because look, 16.2 is a triple from 5.4 and we all know that the demo starts at 5 BTC as the largest enumerated size. If I go from starting at something valued at $150,000 ordering from 7 different currency pairs and I size that down to order only on a 15 minute interval rather than a one minute interval, I'm already reaching for an argument that I'd be able to find a small measure of success in a $10,000 demo account, and if that's ordering from 7 different pairs, then I can constrain myself to one, and switch thee ordering interval all the way out to no more than one order per currency pair per hour. And if I look and see, at least through the lens of my my own understanding, that I can make a significant profit rather than trashing the account, then I'm putting it on the market for sale. Maybe a $35 one time payment is good enough. You can get a Liberty Silver Round for the same price, instead, if you're worried about losing the money. I find that the hardest thing is to ignore this program, because it's always giving me new insight into what positioning is, I suppose, "optimal". The biggest advantage is that I've been able to eliminate my emotions from trading and that's made the returns realized by my demo account much more stable. Remember that 3% of $100.00 is $3.00; it's easy to think $3.00 isn't a lot of money and end up $30.00 in the hole, but if you lost 30% of your account??

You know you've got a problem when people start looking at you like you shaved off one of your eyebrows.

It looks like from my demo account that the move lower in the dollar is on hold for now. The trend until recently has been favoring outflows, but it looks like the move lower is on hold for now, with AUDUSD and NZDUSD testing support and looking lower and USDCAD poking a bit higher, but within a range. Programmed advice took a small position in NZDUSD, but nothing else is biting on the 1 hour chart. I stopped out the positions that I took in crypto over the weekend, because they were all wrong. Fortunately, I'd already made it to 17.5 when I stopped down at 17. All of them were going in the wrong direction and that was maybe only 3% and that stung when I lost it and stung more when BTCUSD went my way later after I stopped the positions. It was short everything but Monero, and every one of those positions went against me. The lesson is that it shouldn't sting; if you're in a position, you don't know why, and you look when you're down, and you're only down a little bit, when it could become a lot worse and you don't know why you're in that position, you've just gotta consider what you'd be losing if you ride that number for your equity down. I hate to get all dodgy right now, but in this position the market makers are in the position of power, because the price is indecisive and testing lower in a trend higher might sound really easy, but you can't ride the margin trades forever if there's going to be a correction, because a correction could be turbulent, and if there was a correction in the trend against the dollar, then going against the dollar will go against you as the market corrects within the trend and there would be ample leeway to apply a later trend resumption or continuation against bearish trade, and then you could get caught out on a position.

Things look rather precarious, right now.

So lemme just start by saying that I know 50 bucks a month is steep for an indicator, but this isn't one that's included with Metatrader and it's not just that I've been seeing returns it's that this one rides the line between being used by the markets and being available to the masses, so the signal doesn't have the aspect of instrument reactivity: whereas you might see slight price spikes with the more commonly used indicators when you see a signal cross, for the time being, you won't observe that phenomenon with this indicator, and thus you'll see a significantly lower probability of price slippage. I haven't released the expert advisor that goes along with this, yet, because I'm still testing it and I'm still trying to figure out how to mitigate pullbacks in equity within that demo account. It's hard to come to the realization that when you're trading that if you want to turn a profit, you have to do something more sophisticated than trade the random walk to a margin call. Still, seeing returns on a larger demo account doesn't mean that the algorithm scales down for use with smaller accounts and that's another challenge that I have to face before I release an "advisor" that will, in fact, place trades without being directed by the user.

The demo did make the climb to 16.5 from 14, though and that is reassuring. I think a lot of the losses were being incurred by investments in commodities and indices where either there's too much volatility or too big a spread price.

I'm releasing an indicator and I'm gonna charge for it. It's an algorithm that looks at the differences between the exponentially bifurcated averages and draws a histogram for the standard deviation and a signal line for the mean value of those differences. It does appear to have a negative correlation with bullishness, meaning that that having a negative value for the mean of the differences is a bullish indication and having a positive value is a bearish indication. The standard deviation then appears to become a measure of the instantaneous volatility of a trend. It works on the premise that for stable markets to exist prices have a tendency to seek an equilibrium condition if speculation suggests that the price is inappropriate, and while these are subjective terms, the algorithm will detect momentum in a trend as a measure of the "desire" to seek equilibrium in one direction or another and describe deviation from that trend quite accurately. The next couple days, I'll be doing the write up and putting together a sales package for the Metaquotes listing.

Thank you all for trying my software. I hope to continue work on this so-called Bifurcated Averages Indicator and the Velvet Project. I'm happy with intermittent progress, because it means that I'm not going backwards. Looks like a lot of things are falling apart out there in the world, but I want you to know that I'm not.

Be careful, alright.

I'm testing out the expert advisor that I wrote again. I had stopped testing, because there was a pullback at the end of last week, even though the demo account recovered to 14 from 9. But I really don't know what's going on with it at this point. If it works, I don't understand why. I might look for inspiration this weekend and try to make some progress then. It's all quite up in the air right now. I really concerned that even if I've written something that works, I'm especially concerned about the volatility, because if it's too volatile it won't scale down for smaller accounts. And also I've stopped trading in USDJPY and USDCHF with it, because the strategy has only been incurring losses in those pairs.

I really hate to feel like I'm playing a dice game, here, but that's how it is right now.

Backtesting is going really bad. Seriously idk wtf I'm talking about --- anything. So here I run into a roadblock for now. Maybe take a break a few more days and draw some stuff out on paper. I thought that it would be worthwhile to make a dll binary for my string handling functions, and when I get done with that I had to clean things up again. I know there's no dll binaries allowed for market products, but seriously, I know I said that this was not meant to be backtested, but I'm looking at actual strategies that I can test and I have no clue what I'm doing here. Maybe I get something that works in one case, but not another. Then there were other things like how much leverage am I using and what goes wrong when I get wiped out. Good for demo accounts---keep trashing them out. Bullion is the place to be, because there are going to be major mood swings if you're planning for the long term. I think it just depended on the mood of the market, really. Sometimes you get follow through and sometimes you get a reversal. They're reversing, and there's not a damn thing you can do about it. Sometimes they were not reversing and you needed to get out too. But yeah, I got nothing. That one algorithm was up last week, and this week is not so good. It didn't work with metals or indices. USDCHF and USDJPY didn't work for some reason. But it came through loud and clear with AUDUSD and NZDUSD. It looked like EURUSD, USDCAD and GBPUSD were indecisive. On the whole the account is up significantly, follow through, however, comes with a premium.

And don't trade crypto on spot markets. In general, that's probably a really bad idea. I singled out ETHBTC for a reason and you need to understand that reason before you invest in that market.

https://stackoverflow.com/questions/68586248/is-it-possible-for-an-exchange-operating-based-on-cryptocurrency-deposits-to-spo

If you thought that the 5 000 000 BIT account was a little steep as an entry, I thought I might launch a signal that was appropriate for use on a $30 account. My goal is to try to push that barrier down closer to the $10.00 level, because that is, in fact the minimum deposit for EagleFX. I'm sitting here looking at Velvet Sea running on my system in the background. There's nothing to it, just the configuration. The first position was triggered about 15 minutes ago. That's going to be my focus this week --- confirming the results from last week and testing Velvet Sea. Velvet Sea, however is chart locked, meaning that it has to be running on the chart with the smiley face shown. It may be slow going, but you can run it on multiple charts. It took about 4 hours to get it to trigger one penny lot position, but it keeps detecting clean bar close events and hasn't experienced any open/close/open/close cycles or really shown any negative equity at all.

I'm really happy that I'm showing clean execution. I'll put a signal up if I get a decent result.

Follow my demo account. The account number / read only password is 274684, n2mfriy.

This one uses Martingale Averaging to generate entries based on whether or not the ranking of moving averages with increasing periods is ordered or not. I more than doubled the account in a week. But it's a huge account: 5,000,000 BIT is a high number to begin with and you don't have to be very judicious about risk. However, it worked. And quite unexpectedly I might add. And that's an Eagle FX Demo .