Status as of Sunday, October 22, 2017

[Steve Sokolowski]

Good evening!

• We didn't get much done this weekend. I fell ill, and Chris is still recovering from what appears to be a different virus, so I'm hoping he doesn't start over again with whatever I have. We hope to be back to normal in a few days.
• There were a few things we accomplished. First, we discovered a problem with electricity calculations. A year ago, we added a factor that only stores a random percentage of shares in the database. We correctly multiply the value of the shares by the factor to get a rough estimate of earnings, but did not do that by electricity calculations. Because the factor is 10, electricity values were 10% of what they should have been. Chris will be releasing an update this evening to resolve this problem.
• We resolved an issue that prevented some users from logging in with two-factor authentication.
• Over the next week while we continue to prepare our future plans, I'll be focusing on making permanent reductions to support ticket numbers and improving security. For support tickets, we already added and deployed a countdown timer that displays when the latest possible payout time can be if a balance is mature. We hope that this timer will cut down on the number of canned replies provided to customers who ask when their payouts will be issued.
• In security, I'm going to make a few changes to try to reduce the number of tickets from customers who have money stolen. First up will be blocking Tor access to accounts. When available, this feature will be enabled for all new accounts, and people connecting through Tor exit nodes will see a message stating that access is disabled. Customers can uncheck the feature upon signup if they wish it not to be enforced for their accounts. Many hackers use Tor to try to hack accounts at various websites, and this feature will force these people to either choose a target where they can remain anonymous, or reveal themselves and suffer the consequences. Existing customers will be able to opt-in, with the feature disabled by default.
• Chris was unable to determine why the mining server shut down yesterday. We'll continue to monitor the problem.
• I expect that the maximum number of users we can handle without parallelization of coin assignment is 11,000. Since we can't take the risk of such a big release without 100% of our attention, expect that registrations will be disabled for an extended period if the limit is reached before we hire additional employees.

[VanessaEzekowitz]

Because the factor is 10, electricity values were 10% of what they should have been. Chris will be releasing an update this evening to resolve this problem

You better double-check that. I have had "w=350 p=0.103" in my password arguments for about a week. That works out to about 87 cents a day (since the miners run 24 hours a day), but it's been showing as 4 cents on the site, so less than 1/20 of the right value.

[CryptoDamon]

You guys get well, your health is first priority.

[Aura89]

Because the factor is 10, electricity values were 10% of what they should have been. Chris will be releasing an update this evening to resolve this problem

You better double-check that. I have had "w=350 p=0.103" in my password arguments for about a week. That works out to about 87 cents a day (since the miners run 24 hours a day), but it's been showing as 4 cents on the site, so less than 1/20 of the right value.

I'm going to have to agree with this. If it were a factor of 10, that would mean my power usage, which was showing around .11 cents a day before with 800 watts and .1163 as the cost, would be around $1.10, even though it should be more around $2, according to every single website out there that calculates wattage by cost

For instance cryptocompare states that 800 watts and should be about $2.23 a day.

There were a few things we accomplished. First, we discovered a problem with electricity calculations. A year ago, we added a factor that only stores a random percentage of shares in the database. We correctly multiply the value of the shares by the factor to get a rough estimate of earnings, but did not do that by electricity calculations. Because the factor is 10, electricity values were 10% of what they should have been. Chris will be releasing an update this evening to resolve this problem.

I by no means mean for this to be taken insulting as i am genuine just curious. But considering how much wattage something is and how much it costs per kWh, my question on this is why is it complicated?

The math on this is simple, there's really no variables other then what we would put in as the cost and how much wattage is being used. The math should be something similar to:

(Wattage x Hours Used)/1,000 x Electric Rate ($/kWh) = Total Cost

Since there's only 3 variables, wattage, how long the website knows that it's connected, and the electricity cost rate, what is the issue?

Again, this is a genuine question because i'm just not understanding why that math can't be implemented

If i do that math for my device for 24 hours time, that comes to: (800x24)/1000x0.1163=2.23296

[Steve Sokolowski]

Because the factor is 10, electricity values were 10% of what they should have been. Chris will be releasing an update this evening to resolve this problem

You better double-check that. I have had "w=350 p=0.103" in my password arguments for about a week. That works out to about 87 cents a day (since the miners run 24 hours a day), but it's been showing as 4 cents on the site, so less than 1/20 of the right value.

I'm going to have to agree with this. If it were a factor of 10, that would mean my power usage, which was showing around .11 cents a day before with 800 watts and .1163 as the cost, would be around $1.10, even though it should be more around $2, according to every single website out there that calculates wattage by cost

For instance cryptocompare states that 800 watts and should be about $2.23 a day.

There were a few things we accomplished. First, we discovered a problem with electricity calculations. A year ago, we added a factor that only stores a random percentage of shares in the database. We correctly multiply the value of the shares by the factor to get a rough estimate of earnings, but did not do that by electricity calculations. Because the factor is 10, electricity values were 10% of what they should have been. Chris will be releasing an update this evening to resolve this problem.

I by no means mean for this to be taken insulting as i am genuine just curious. But considering how much wattage something is and how much it costs per kWh, my question on this is why is it complicated?

The math on this is simple, there's really no variables other then what we would put in as the cost and how much wattage is being used. The math should be something similar to:

(Wattage x Hours Used)/1,000 x Electric Rate ($/kWh) = Total Cost

Since there's only 3 variables, wattage, how long the website knows that it's connected, and the electricity cost rate, what is the issue?

Again, this is a genuine question because i'm just not understanding why that math can't be implemented

If i do that math for my device for 24 hours time, that comes to: (800x24)/1000x0.1163=2.23296

It actually took me about 4 hours to figure this problem out yesterday, because I looked at the same math you did and couldn't figure out what was wrong with the conversion.

But as I said in the other post, the problem wasn't the formula, it was the performance improvements. The performance improvements cause some shares to be dropped a random, and the remaining shares are multiplied by the number of shares that would be dropped on average to come up with approximate earnings.

I think the sampling rate is set to either 10 or 20 right now. When the random number between 0 and 10 or 0 and 20 equals zero, then that share is recorded and all its values (except electricity cost) were multiplied by 10 or 20 to get the value of the share. Electricity cost was not multiplied, resulting in incorrect data.

Now, the electricity cost is multiplied along with the value of the share, so the electricity values should also be correct.

[Aura89]

You better double-check that. I have had "w=350 p=0.103" in my password arguments for about a week. That works out to about 87 cents a day (since the miners run 24 hours a day), but it's been showing as 4 cents on the site, so less than 1/20 of the right value.

I'm going to have to agree with this. If it were a factor of 10, that would mean my power usage, which was showing around .11 cents a day before with 800 watts and .1163 as the cost, would be around $1.10, even though it should be more around $2, according to every single website out there that calculates wattage by cost

For instance cryptocompare states that 800 watts and should be about $2.23 a day.

There were a few things we accomplished. First, we discovered a problem with electricity calculations. A year ago, we added a factor that only stores a random percentage of shares in the database. We correctly multiply the value of the shares by the factor to get a rough estimate of earnings, but did not do that by electricity calculations. Because the factor is 10, electricity values were 10% of what they should have been. Chris will be releasing an update this evening to resolve this problem.

I by no means mean for this to be taken insulting as i am genuine just curious. But considering how much wattage something is and how much it costs per kWh, my question on this is why is it complicated?

The math on this is simple, there's really no variables other then what we would put in as the cost and how much wattage is being used. The math should be something similar to:

(Wattage x Hours Used)/1,000 x Electric Rate ($/kWh) = Total Cost

Since there's only 3 variables, wattage, how long the website knows that it's connected, and the electricity cost rate, what is the issue?

Again, this is a genuine question because i'm just not understanding why that math can't be implemented

If i do that math for my device for 24 hours time, that comes to: (800x24)/1000x0.1163=2.23296

It actually took me about 4 hours to figure this problem out yesterday, because I looked at the same math you did and couldn't figure out what was wrong with the conversion.

But as I said in the other post, the problem wasn't the formula, it was the performance improvements. The performance improvements cause some shares to be dropped a random, and the remaining shares are multiplied by the number of shares that would be dropped on average to come up with approximate earnings.

I think the sampling rate is set to either 10 or 20 right now. When the random number between 0 and 10 or 0 and 20 equals zero, then that share is recorded and all its values (except electricity cost) were multiplied by 10 or 20 to get the value of the share. Electricity cost was not multiplied, resulting in incorrect data.

Now, the electricity cost is multiplied along with the value of the share, so the electricity values should also be correct.

Awesome. I figured it had to be more complicated then what i was thinking lol