Topic: 10,000 YACoin Bounty - page 2. (Read 2331 times)

x-post from https://bitcointalksearch.org/topic/annyac-yacoin-ongoing-development-206577


@St.Bit: sure i know, i've added link to coinchoose ltc comparison charts. just compare the % profitability values (but first i need confirmations that it's calculated right)

That's easy.

You only need to trust someone if he has benefits from acting against your interrests. I don't see this here and even if you weren't honest I would be very surpriced if you steal 1kYAC now although you could have easily stolen 50kYAC from me long time ago. Even after the recent raise I'd still trust you with 50kYAC if I had to.

@Sairon:
You do know that there are like 60 cryptos to mine out there, right? EDIT: xD

a lil' screenshot

Umm, so I got something working...
Nothing fancy so far, but it works.
The difficulty and exchange rate textboxes are pre-filled with cached data from exchanges/block explorers, so it should be quite accurate.

You need to input you YAC hashrate (at current Nfactor), LTC (or any other vanilla scrypt coin) hashrate and BTC (sha256) hashrate. If you don't have hardware for eg, BTC mining, just enter 0. The daily profits for each coin are calculated and shown, along with info which coin is the most profitable in your setup.

Tell me if you have any suggestions.

http://yacexplorer.tk/static/calc.htm


...also, BTC is over $1000, yay! ^^

Thank you St.Bitt.  You can trust I will keep them safe until someone fulfills the bounty.  I don't know how you can trust me, but you can.

LOL, tell me something about that. Been meaning to do this since I first saw the suggestion but then accidentally closed the tab and as nobody posted into this topic I just completely forgot to do it...

If you hadn't posted here I would probably have forgotten about this pledge.
To prevent that I send my pledge of 5kYAC:

97d140523519a8bfb2774fe97a6c44ac2b3ecb4e74288a8efaac1b4e062f21bf

This bounty should be pretty appealing right now...  I kinda wish I was the government and these coins, social security funds.

YLNhNhRVspJMZLBN4PcSjkLsVTd3TMaq7R

Glad you like the idea sairon   Anyone working on this project?

ybcminer --scrypt --benchmark --Nfmin x --Nfmax x

this would suffice for testing nfactor x

Switch intensities within ybcminer interface

However this will not work for testing CPUs

Would you really set up a full miner incl. blockchain just to test wheter an unknown coin with a micro MC is profitable at the current n-factor?!

Since we already need benchmarks for some of the most common GPU rigs wouldn't it be far better to modify a small miner to benchmark a system for all n-changes <20? Nobody would run such a software if it weren't from the original hub (or in the client) and a little faucet for submitting your results would be very motivating to some.

Genius. ^^
This would be the simplest and most accurate method to implement (although it will require users to test their mining speed for both scrypt variants).

It can be coinchoose.com or a setup quite similar to coinchoose, but it doesn't have to be the same as long as it serves the same purpose.  Is there a way to account for different GPUs, CPUs and combinations with memory size?  I'm envisioning the ability for one to input YACoin hashrate (the best and smartest way may mean people will have to start mining at least for a couple minutes to truly know this information), and then reverse calculate for LiteCoin or even input LiteCoin hashrate.  Am I completely confused how this would work? 

I also agree that accuracy a pretty relative term.  I believe a big complaint with the coinchoose concept is that you can have an extremely profitable coin, but the market depth and volatility is such that it is really pointless data.  It is almost like there needs to be a 'weighted factor' that incorporates those characteristics to determine the most RELIABLY profitable coin.

I think such a website would benefit from submission of people's experiences at different N. It is bound to be more rough than the usual scrypt/SHA256 comparison, but the aim should be to provide some guidance to those who are not familiar with it.

The lower power consumption aspect of mining at higher N should also be mentioned. E.g., right now, if I mine a normal scrypt coin, my GPU temperature would go up ~40C while mining YAC now only push temperature up by ~20C. I'm guessing that may be half the GPU's scrypt mining power consumption.

Agreed, but we lack a benchmark for a sufficient number of GPUs. I said that the data was highly innacurate, but even if it's half the profitability IRL it's still in the TOP 3 profitable scrypt coins when compared to LTC.

EDIT: So far I've implemented getNormalizedDifficulty API call on my block explorer. Now I only need to write getBlockReward and we have all it takes to add YAC on coinchoose.com if they agree. (And possibly tweak some parameters or build a GPU benchmark chart.)

I belive your math is correct, but too inaccurate to use it for comparisons.

The real problem here is rather the normalizing of the YAC hashrate. Different GPUs yield different hashrates, that's nothing new, but everyone knows it's rate for LTC so that's no problem if you compare LTC like coins. For YAC it is different since it depends far more on the memory than on the GPU.

The 1/(0.5*NF-9) is a good approximation when it hasn't run out of memory, but most card already have begun to show first effects of running out of memory. The decrease factor for an n-changes is ~0.5 with enough memory and  <0.25 when memory isn't sufficient anymore.

If someone buys cards for mining they usually buy some with just enough RAM for mining LTC. These cards already ran out of memory and got hit by the last n-change.

It isn't even accurate for the current n-factor, so how could we consider that for the future ...

We would piss off a lot of people by promising 400% although most of them couldn't possibly reach that. We have to use benchmarked decrease factors and list the profitability per cardtype.

I'd say the best is to use profitability with decreasefactor 0.5(n-9) with asterix and link to a more detailed site with the benchmarks. I belive there are some cards that would be very profitable on YAC, but we must not upset people with cards not suitable.

I can do this. I need links for the block explorers for YAC, YBC, ZCC, GPL, ONC, CENT, CPR.

Well, it's a nice idea and I'm all for it, but there is a slight problem. As the profitability is based (among other things) on hashrate/difficulty comparison, how do we compare scrypt-chacha (or however you want to call it) with plain scrypt or even sha256? As you can see on my graphs page, simply normalizing the YAC hashrate (by "hashrate / (0.5 ** (Nfactor - 9))" based on Nfactor doesn't yield acceptable data:


To be clear, this graph thing *could* be caused by some miners leaving after the Nfactor change, but I doubt there are so many to make such a difference in net hashrate. However, the change has an amplified effect on GPUs as they're getting *less* than half their hashrate after each change. Dunno about CPUs, but even the L3 caches might be already used up on most of them.


If we consider this data to be correct for a while, I can compute the profitability compared to LTC. The values we need are: difficulty ratio, price ratio and block reward ratio.

Current normalized YAC hashrate is just slightly below 40 Mh/s. This translates to an LTC-compatible difficulty of 0.55, which is ~1802.27 times lower than LTC at the moment. So far so good.

Now for the price comparison: LTC is worth 1 LTC a piece (who would have thought...) and YAC 0.0025 (sell price from bter, buy is at 0.0036). So the YAC is worth around 1/400th of LTC.

The last thing we need to take into account is the number of coins generated per block. LTC has stable 50 and YAC is currently at 49.69. Dividing YAC by LTC gives us another value: 0.9392.

Now multiply these three numbers and we get ~4.23. Multiply by 100 to get a % value: YAC has 423% profitability of LTC.

Please, do check my math.

How to compute diff:


Hashrate is in hashes per second (not kh/s or Mh/s!), 2^32 is constant for scrypt and 60 the block interval target.

Great idea.

I'm also supporting this idea with 5kYAC for a comparison of mining profitability with YAC. Doesn't matter to me if based on YAC or just if puts it into comparison.

Giving away 10,000 YAC for a website like coinchoose.com, comparing mining profitability except based off of YACoin.  YACoin becomes very profitable to mine with both CPU and GPU with each NFactor increase, and it needs to be advertised.

Edit:
Here is the address with 10,000 YACoin loaded:  YLNhNhRVspJMZLBN4PcSjkLsVTd3TMaq7R