Topic: [ANN] Catcoin - Scrypt meow! - page 26. (Read 470759 times)

Worldcoin retargets every block. We are doing the same on the next fork.

Is there any way we could change Catcoin difficulty parameters to something similar to Worldcoin when the next fork comes? Like it's not affected by all those multipools and coin hoppers and is always profitable to mine.

Never optimize for the moment.

Always optimize for the long term.

Have you thought about running simulations to see how these ideas work? The technique is to simulate the various mining agents that behave like real-world agents and also simulate the properties of the coin and see how coin production is effected by changes to the coin. If this seems like something that could be helpful as input to the fix, let me know on this thread and I'll jump on the IRC when I have time. I have simulated many systems.

Agreed.


Yes.


I'm sure that this solution will make the coin far more stable if the average accounts for a long enough period. If the goal is long-term stability then the average should look at the last 2 weeks (2016 blocks). I think the current 36 block retarget time is good, although with a moving average the retarget time is less critical.

But, whatever the specs are, the moving average has to be done for CAT to survive longterm. Even a 36-block average with a retarget after every block would be much better than the current specs.

Bard of Directors here.

I want a coin that is both cute and strong.

Cats stay. Bitcoininess stays.

Too many core changes will harm us later.

I personally am against the modification of the block reward in any way shape or form. I feel that fundamentally alters the coin, no matter how trivial. In my books, block reward and block time are completely off limits. But hey, this coin belongs to the community, so we will do what the community asks. If that includes changing the block reward, so be it, though my gut feeling says that the large percentage of the community will disagree with that action.

Some of our team is less concerned than I am about such changes, and it has consumed considerable amounts of our resources time wise discussing this very issue. What does the community think? What do you guys want to see in this coin improvements wise? How far are we willing to go? Are there variables that are absolutely off limits?

Personally I'm for maintaining the scrypt bitcoin that this coin was originally marketed on. Sure, the diff re-target is different, but I think that's a fundamental necessity it today's market.

I've attached a paper on "Simple Moving Averages" that hozer has found. I'd like you folks to have a look at it. It can be found here: https://www.tradestation.com/education/labs/analysis-concepts/a-comparative-study-of-moving-averages

Personally speaking, an SMA or SWA provides exactly the behavior I think will solve our issues. Granted this paper is trades based, but the concepts are applicable here. If the diff is appropriate to hash, then we won't see the pools drop in nearly so often because the coin won't be ultra-profitable for short periods of time. Even if they do the coin will recover better, faster and more accurately ensuring our regulars are impacted as little as possible. The rest should fall into place on it's own. It is by far the simplest of our current proposals. I'm firmly of the belief that simple = less issues long term. We can address missing or new functionality later, these diff changes need to happen soon. The biggest issue is the fact that the diff can drop or jump by 400% currently, it needs to step, not jump. If it steps it won't go blowing past where it should be with no hope of stopping. We can add a maximum jump per block or dampener if needed to ensure the diff average doesn't become too steep and overshoot in extreme circumstances. Perhaps 100% per hour is a reasonable number. I'm open to suggestions.

Cat slowly rising again https://www.cryptsy.com/markets/view/136, yay. Hope it carries on over the next few weeks.

So this probably means that the block reward will not be constantly 50 cat per block ?
Isn't the "new technology" just an average of the last n block interval s / 600 multiplied by the current subsidy ? Or something like that ?

Without getting involved in the mechanics of the interactions between the difficulty, profitability incentives, and production rates, it is fair to say that when production rate spikes, a knee-jerk reaction causes production rates to fall. Since our coin-hopping friends are not interested in mining our Coin when production rates start falling, they leave. This change in the number of people contributing work to our coin results in the coin production rate dropping substantially below the baseline (show here with dotted lines).

One approach that is being considered is to put a constraint on the number of coins that can be produced per minute (represented by the downward pointing arrows), independent of the "difficulty." The fact that this can be done is a recent discovery, and represents an innovation which Catcoin may be the first to implement. This will tend to keep coin hoppers uninterested in our coin (which means fewer miners leave, when the rate production goes down sometimes). Since the number of miners leaving is what causes the production rate to fall significantly below the standard 5 coins per minute average, by suppressing the coin production rate (down arrows), we expect and have modeled that the coin production rate would go up during the rest of the time (upward pointing arrows).

Then we expect we will see something like this:




Now, one of the effects or shortening tall bars, is for them to get wider, until they settle at just about 6 hours (36 x 10 minutes each) per bar. This is because when you reduce the rate of coin production, it takes longer to complete the 36 blocks of production, and the next adjustment in production parameters can take effect. The effect of putting a constraint or cap on the coins produced per minute, is to make coin production rate more or less consistent over time. Behind the scenes, "difficulty adjustments" would continue, and hash rate can move quickly to match the market price of Catcoins on the exchanges (the more valuable the coin gets, the more hash power it attracts and justifies). But these adjustments happen automatically in the background and are not that important as long as they don't affect coin production rate, which by design should average and stay pretty consistently at 5 coins per minute. Because of the newly discovered method of disconnecting "difficulty" from coin production, the coin can adjust with maximum speed and agility to influx of temporary hashing power by limiting coin production rate, and in an independent manner, also adjust the overall "difficulty" in response to market prices of Catcoins.

Please feel free to post questions, raise any objections, add new observations, propose alternative ways to look at the problem, etc. to advance us closer to a consensus. Thank you for your time.

Etblvu1

Hold on cats, the vet is on the way!!!

What is this Graph For?

This graph was put together to help form a consensus among Catcoin supporters, about the nature of the problem we are facing, often described as difficulty level oscillations. The graph intentionally omits directly showing the difficulty levels, because it's actually not that important - and instead focuses on factors people actually care about - i.e., what is the actual rate at which coins are being produced, and who are receiving the coins that are being produced? To create this graph, I collected about seven days of data from chatchain.info -  covering roughly January 7 through January 14. Anyone can verify the graph as presented matches up with the information that's available there.

Notes to Help Understand Graph

1. Y-Axis as Coin Production Rate. Just like with Bitcoin, the Catcoin specifications call for an average of 10 minutes between production of blocks. This means there should be an average of 5 coins produced per minute. The Y-Axis represents coins produced per minute, and the lowest notch is labeled with emphasis on "5 Coins / Minute" to show this is the normal coin production rate. This is where we should be.

2. X-Axis as Time. The X-Axis is simply time. Big labels mark off midnight that starts off each day, and little notches mark off 3 hour increments (3am, 6am, 9am, 12pm, 3pm, 6pm, and 9pm). The data is accurate to the pixel level - the widths show how long each "difficulty level" era lasted. Note that the faster the rate of coin production, the less time the era occupies. Correspondingly, if the rate of coin production slows to a crawl, the "difficulty level" era stretches on for sometimes days.

3. Area as Proportional to Reward. The rectangular areas you see (green, blue, and red) are fair representations of the actual number of coins produced / distributed / not produced. This did not become clear to me until producing the graph, and it is counterintuitive, but the spikes in production rate (represented in green, which also shows how many coins coin hoppers took away), is actually a much smaller number of coins, than the amount of coin production subsequently suppressed (the red zones). In effect, we are seeing "insults" to the coin once in a while, which causes an hour or two of unrealistically fast coin production rates; and this is followed by "overcorrection" to produce coins at a snail's pace.

4. Coin Hoppers are Green. I gave them green, like grass hoppers. Basically, when the rate of coin production is very fast, everyone who mines the coin can expect to obtain more coins for a given amount of hashing effort. So the coin is very popular to mine, but only for a very brief time.

5. Loyal, Consistent Miners are Blue. When coin production is high, loyal, consistent miners see their yield rise, and they actually collectively see approximately the 5 coins per minute production rate they expect should always be happening. But the lion's share of the coins produced, goes to coin hoppers (green).

6. Missed Production. When the rate of coin production gets very slow, we start seeing numbers like only 1-2 coins being produced per minute, and even less. This is very bad for the coin retaining consistent, loyal miners, because they see very few coins for the large amount of effort put in. Ideally, we see little or no missed production (represented in Red).

7. Actual Coin Production Rate. The black line represents the actual coin production rate. If you have a hard time seeing it (it is pretty thin), just trace a line along where the green and blue blocks are, ignoring the red. This represents roughly how many coins are being produced per minute, at each point in time, during the 7 days covered.

Please feel free to post any questions about the information presented, and I will endeavor to do my best to clarify anything that is confusing in the graph. I hope this will spark some discussion and help bring fellow Catcoin supporters closer to a consensus about the nature of the difficulties we are seeing, and how we might go about solving it.

Etblvu1

Few maybe little more hours left us to offical anounce fork for CAT coin, wich will repair difficulty for good.
Dev working on best solution since couple of days. With amount of testnets, simulations and ideas.

So I propose to buy CAT coin from sellers when price is so low.

I think now its best time for invest in CAT coin.

thx for the reply!

and thanks @zerodrama too.

when i'll come home i'll try!

I updated progress yesterday.  I haven't posted progress yet today because we're still working.

We've got a test-net running, have four solid options sketched out, and code is being adjusted.  Some testing has already happened and the new code will be checked once it's finished and compiled.  We're also building environments to compile wallets so that we'll have binaries ready prior to the fork.  (Still don't have an Apple compiler yet, though, but have one good Cat that's offered Mac help.)

We're all miners, investors holding at least 400 CATs each, and cat people.  We're very aware that coinhoppers are scooping the 'easy' coins while leaving all of us to slog through the slow times.

I can assure all of you that we're interested in making this coin a success.

We'll pass more details once the coding's done and tested - and we'll put the options out to the community for full vetting.

In the meantime, if anyone's interested in the play-by-play, join us on IRC freenode #catcoin-dev

Andy
Keeper of Ada Grey (The Mighty Huntress)

Do you like that stair-step function?

You have two groups of miners here:

1. In the first group are catcoin faithfuls, who mine the coin through thick and thin, spending most of their time in the thick with high difficulty.
2. In the second group are coin jumpers. These people and pools mine the easiest coin on the block. The coin jumpers mine catcoin after the catcoin faithful have mined all the difficult blocks.

When the coin jumpers finish mining the easy catcoin blocks (before it re-targets to extreme difficulty), they go straight to cryptsy and dump all of their coin, eating every buy order until the coin jumpers have no more coin to sell.

The catcoin faithfuls are rapidly dwindling however because whoever has taken responsibility for developing catcoin refuses to fix the issue.

The only positive aspect of this tragedy is that it provides a grotesque sort of entertainment for the type of people who rubber-neck automobile accidents.


[EDIT]
I wrote the above while the previous post was being written. Hopefully a fix is coming!

LOL.

THE COIN IS BEING TESTED.

FIXES WILL COME WHEN THE COIN FEATURES ARE DEMONSTRATED TO WORK.

OR YOU CAN JOIN US IN CHAT.

Cat Colin's price dropped a lot today, what's going on??

Current Difficulty    63.60426279
Est. Next Difficulty    6.38379115 (Change in 33 Blocks)
Est. Avg. Time per Round (Network)    1 hour 39 minutes 38 seconds
Est. Avg. Time per Round (Pool)    36 hours 35 minutes 15 seconds
Est. Shares this Round    260523 (done: 153.33%)
Next Network Block    21100    (Current: 21099)
Last Block Found    21093
Time Since Last Block    17 hours 30 minutes 2 seconds

No way to run a coin.

It doesn't move because the block time is slow, because of the difficulty attacks. It assumes that it's behind.