[ANN] Catcoin - 0.9.1.1 - Old thread. Locked. Please use 0.9.2 thread.

[zerodrama]

Also 10.7% down did I miss something? and this is even worse, it means that the diff takes longer to go down than to go up.

FAIL.

Let's use an EXTREME example.

Say we want to allow a maximum 100% jump upwards?

DIFF SERIES UP: 1, 2, 4, 8, 16... FOR A 100% JUMP UP we MULTIPLY (not add) by 2.00x (otherwise there's no SYMMETRY and favors downward drop which rewards big dump miners at the expense of loyal small miners).

What if we want to balance the downward motion?

100% SERIES DOWN: 16, 0, 0, 0... HOUSTON WE HAVE A PROBLEM!

IT SHOULD BE 50%: 16, 8, 4, 2, 1.

In the case of 12% -> 112/100 max up, 100/112 max down. 12% up, 10.7% down.

[zerodrama]

I can't believe what I read here. You are still thinking a little tuening here and a little there and your averageing function with limits produces a satisfying solution!?

You already did 2 changes which aren't cheap. You have to fork and everyone has to update. And you really want to continue that path? Another change attempt and fork and another likely failure? And then... another one?

If this charts and data that you see aren't conlusive enough for you to see your base is off and you are going the wrong way than I guess I wasted my time with my gently hints.

You go a painful length in just to avoid an already existing SIMPLE and elegant solution at all cost.

For what?

I lost the hope. You don't know what you do.

Kimoto is not simple.
Kimoto antagonizes the coders against the users.
Kimoto is not explained.
Kimoto has competition.

[loader140]

Anyone want's to buy CatCoins?

  PRICE? 

Do they have a loadable address and hologram? or are they just for fun?
What currency will you accept? Will you ship to UK? Price?

Either way, can I reserve one

[gadado]

Kimoto is not simple.
Kimoto antagonizes the coders against the users.
Kimoto is not explained.
Kimoto has competition.

It's actually close to a Joke. I first thought he makes a joke reading the formula and the code.
It's clever no doubt. That man is good. And he is funny.

[zerodrama]

The average block time over 24 to 48 hours is 10 minutes. The range is a bit uncomfortable right now.

UPDATED FORK ANALYSIS:

FORK START
BLOCK    TIME                                DIFF       TOTAL COINS
21346   2014-01-31 05:57:46   108.562   1067350

1ST BOTTOM AND TOP
21385   2014-01-31 19:06:53   1.306   1069300
21427   2014-02-01 01:10:27   140.753   1071400

42 BLOCKS
363 MINUTES 30 SECONDS
AVERAGE: JUST UNDER 9 MINUTES

2ND BOTTOM AND TOP
21471   2014-02-01 15:47:06   1.193   1073600
21508   2014-02-01 22:33:34   79.186   1075450

37 BLOCKS
404 MINUTES 28 SECONDS
AVERAGE: JUST UNDER 11 MINUTES

3RD BOTTOM AND TOP
21543   2014-02-02 02:16:25   1.499   1077200
21588   2014-02-02 08:51:11   225.885   1079450

45 BLOCKS
394 MINUTES 46 SECONDS
AVERAGE: JUST UNDER 9 MINUTES

****NEW DATA****
4TH BOTTOM AND TOP
21637   2014-02-03 10:00:12   0.899   1081900
21680   2014-02-03 16:48:56   117.85   1084050

43 BLOCKS
408 MINUTES 44 SECONDS
AVERAGE: JUST UNDER 10 MINUTES

STILL WORKS!

****PUBLIC SERVICE ANNOUNCEMENT****
THE FORK WORKS!
----------------------------------------------------------------------------------------------------------------------------
THERE ARE PROBLEMS THAT NEED TO BE SOLVED AT THE POOL LAYER.
GRAVITY WELL IS A HACK AGAINST A PROBLEM THAT WILL KEEP GETTING WORSE.
WE NEED COIN SOFTWARE THAT CAN BE UPDATED BY CREATIVE PEOPLE AS WELL AS DEEP CODERS.
GRAVITY WELL MAKES THAT MUCH HARDER.

The community in the long run is the life of the coin. Every coin has proved this principle. The more black magic there is in the code, the less the community is able to adapt and participate. This effectively kills the coin or makes it the province of established owners of the old system. SEE New York City HEARINGS. They don't understand it. They are afraid of it. They can't even compromise because it is all technowizardry to them. And yeah they're also trying to own it.

[zerodrama]

Gentrification of bitcoin and litecoin is turning them into speculation traps.
Kimoto Gravity Well allows steady take over of the currency by people WHO DO NOT CARE ABOUT THE GOALS OF THE COMMUNITY.
We already have that problem on Wall Street.

[SlimePuppy]

I can't believe what I read here. You are still thinking a little tuening here and a little there and your averageing function with limits produces a satisfying solution!?

No - you are incorrect.  Nobody is talking about 'a little tuning' - though we aren't ruling out that possibility, either.  What's really important is that for the first time since the original dev dropped and ran, we have a functioning coin that CAN run in a 21st century world.  Yes, it cycles.  Yes, we know we can do better.  But now - for the first time - customers can reliably use the coin in commerce, there's enough liquidity in the market and transactions clear quickly enough that more traders are coming in, and we have won a MAJOR endorsement!

NONE of that would have been possible without this fork.  End of story.

I lost the hope. You don't know what you do.

That is your choice and while I don't agree with your assessment, I'm not in a position to tell you you've made the wrong choice.  Good luck with your trading career.

[skillface]

I think instead of looking for yet another solution in code, we should be looking to temporarily borrow some major hashing power during the next high difficulty period to help smooth out the curve, I think. Pretty sure I remember hearing that Bitcoin and even Litecoin had to do something similar during their early days as well.

The code works pretty much as well as can be expected, the mistake was letting it walk down from 108 difficulty at the beginning, I believe. We should have probably set difficulty to 60 or so and went from there, the algorithm wouldn't have tried to compensate as much as it currently is.

Average block time is still ~10 minutes overall, so that proves that something's working, at least. The coin is now mostly usable, minus Cryptsy still being.. well, Cryptsy. Once that's sorted I think we'll start to see some miners return, helping to smooth out the difficulty spikes.

[envy2010]

Gentrification of bitcoin and litecoin is turning them into speculation traps.
Kimoto Gravity Well allows steady take over of the currency by people WHO DO NOT CARE ABOUT THE GOALS OF THE COMMUNITY.
We already have that problem on Wall Street.

Changing these three lines:

Code:

  
if(pindexLast->nHeight >= fork2Block){
        numerator = 112;
        denominator = 100;

To this:

Code:

  if(pindexLast->nHeight >= fork2Block){
        numerator = 1018;
        denominator = 1000;

Still shows good results, but it's not as sensitive to GH/s spikes:

[Nullu]

Gravity well is a hack? What?

It's the only algorithm that seems to actually work against whales and pools. I don't even understand the current difficulty algorithm now. Probably because I lost track of it in the last thread.

You shouldn't have to borrow hashing power, or add in more layers of complexity. This has been over engineered. You're trying to fix problems that are not easy to overcome. I'm still a huge fan of Catcoin, but I couldn't possibly invest any further into it. I would mine it, if it wasn't so unpredictable right now.

[SlimePuppy]

Correct me if I'm wrong, but from your comment I get the feeling that you didn't make the effort to look it up and understand how it work, and I think it's fair to make a judgement about it after you do those things. and btw it is a simpler solution than SMA36LIM12

This may be obvious to you, kuroman, but how many people on this forum do you think are at least marginal C programmers?

Let's all use words, examples, and conversations that are available to the entire community.  Limiting the conversation to a few, either through expediency or hubris, doesn't help the greater good.

No solutions are off the table right now - not all possible solutions are ON the table yet. LOL  We're not yet at the stage where we need to fight over options.  Thankfully, we now have time to propose and more importantly TEST the solutions on the testnet, not just on an Excel spreadsheet or in a Matlab pageprint.

Everyone - put your 'thinking caps' on and get those ideas on the table!

Thank you!

[zerodrama]

Code:

  if(pindexLast->nHeight >= fork2Block){
        numerator = 1018;
        denominator = 1000;

1.8% ARE YOU SERIOUS? No offense, bro but might as well just give away (on drop) / freeze (on jump) the coin at that point. That's equivalent to no change at all. Welcome to FLATCOIN.

I predict every KimotoCoin will be owned by botnets and massive farms. Everyone will be at the mercy of people who DO NOT CARE ABOUT THE FUTURE.

I think if the coin goes between diff 20 and diff 40 after it settles down, that will be perfect. Stable enough to send coins and wobbly enough to chase the strip miners.

[vondi1122]

how many people on this forum do you think are at least marginal C programmers?

Let's all use words, examples, and conversations that are available to the entire community.  Limiting the conversation to a few, either through expediency or hubris, doesn't help the greater good.


Everyone - put your 'thinking caps' on and get those ideas on the table!

+1

[SlimePuppy]

Gravity well is a hack? What?

It's the only algorithm that seems to actually work against whales and pools. I don't even understand the current difficulty algorithm now. Probably because I lost track of it in the last thread.

You shouldn't have to borrow hashing power, or add in more layers of complexity. This has been over engineered. You're trying to fix problems that are not easy to overcome. I'm still a huge fan of Catcoin, but I couldn't possibly invest any further into it. I would mine it, if it wasn't so unpredictable right now.

I've been mining CAT since Dec 26th.  Speaking as someone that mined every one of the 530 CAT in my wallet, it's my position that this coin has been improving since it's launch.

There are other algos that work.  Gravity well is the current fad, certainly, but it's not the only way to go.  Maybe we'll use it, maybe we'll modify it, maybe we'll select another option.  But we're not in the final section process yet - we're just beginning!

This is a good time for everyone to RELAX!  We have brought CAT back from death - this time it can run without manipulation. 

You suggest the current algo is 'over engineered' even as you incessantly promote the gravity well? 


Ideas are good! Even the ones that make the cat laugh.

[envy2010]

Code:

  if(pindexLast->nHeight >= fork2Block){
        numerator = 1018;
        denominator = 1000;

1.8% ARE YOU SERIOUS? No offense, bro but might as well just give away (on drop) / freeze (on jump) the coin at that point. That's equivalent to no change at all. Welcome to FLATCOIN.

I predict every KimotoCoin will be owned by botnets and massive farms. Everyone will be at the mercy of people who DO NOT CARE ABOUT THE FUTURE.

I think if the coin goes between diff 20 and diff 40 after it settles down, that will be perfect. Stable enough to send coins and wobbly enough to chase the strip miners.

It's not at all flat. That allows a 0.982^36 = ~48% change in 36 blocks, and a 28% change in only 18 blocks. Why would the coin ever need to double or halve difficulty in less than 36 blocks? A 28% diff increase is more than enough to make the coin go from very profitable to quite unprofitable to big miners.

[SlimePuppy]

Gentrification of bitcoin and litecoin is turning them into speculation traps.
Kimoto Gravity Well allows steady take over of the currency by people WHO DO NOT CARE ABOUT THE GOALS OF THE COMMUNITY.
We already have that problem on Wall Street.

Changing these three lines:

Code:

  
if(pindexLast->nHeight >= fork2Block){
        numerator = 112;
        denominator = 100;

To this:

Code:

  if(pindexLast->nHeight >= fork2Block){
        numerator = 1018;
        denominator = 1000;

Still shows good results, but it's not as sensitive to GH/s spikes:

<snip>

Envy - will you please tell us the assumptions you're using and/or give an example of what you're 'feeding' your model?  I'd like to understand what environment you're working in.  Thanks in advance!

[Nullu]

Gravity well is a hack? What?

It's the only algorithm that seems to actually work against whales and pools. I don't even understand the current difficulty algorithm now. Probably because I lost track of it in the last thread.

You shouldn't have to borrow hashing power, or add in more layers of complexity. This has been over engineered. You're trying to fix problems that are not easy to overcome. I'm still a huge fan of Catcoin, but I couldn't possibly invest any further into it. I would mine it, if it wasn't so unpredictable right now.

I've been mining CAT since Dec 26th.  Speaking as someone that mined every one of the 530 CAT in my wallet, it's my position that this coin has been improving since it's launch.

There are other algos that work.  Gravity well is the current fad, certainly, but it's not the only way to go.  Maybe we'll use it, maybe we'll modify it, maybe we'll select another option.  But we're not in the final section process yet - we're just beginning!

This is a good time for everyone to RELAX!  We have brought CAT back from death - this time it can run without manipulation.  

You suggest the current algo is 'over engineered' even as you incessantly promote the gravity well?  


Ideas are good! Even the ones that make the cat laugh.

Gravity Well is simple to understand, and fixes a lot of problems that others have complained about. I don't understand why people are so dismissive of it. I've heard complaints that it only addresses specific problems. Well maybe the problem is you're trying to do too much with the difficulty algorithm.

If you want to reward loyal Cat miners/investors, then introduce a small PoS element. I only promote Kimoto Gravity Well because it works. Nobody can counter-argue that it benefits botnets and farms. Every single coin is vulnerable to botnets and farms regardless of what algorithm it employs.

Satoshi said one-cpu-one-vote. It sounds like some people want to undermine this principle. You can reward loyality, but you can't punish greed.

[kuroman]

112/100 = 12% change up, 100/112 = 10.7% change down. This has been noted on previous pages.

The difficulty dropped below 1 because even if 35 blocks were solved INSTANTLY after the fork, the 36 block average time would still have been over 10 minutes -- and so even after 35 drops of 12% the algo "thought" the diff was still too high. Though it wasn't instant, once a couple blocks were found the next 30 or so came very fast. Then, very suddenly, those long blocks were far enough back to be left out of the average, and the average time dropped to below a minute.

The 36SMA12LMT system is intended to react fast to lots of hash entering and leaving the network (which it does well) but a side effect of that is that it overreacted to what it perceived as a lot of hash having "just left" a the time of the fork. If the difficulty had started at 50, block times would have been more reasonable for the 36 before the fork, and this oscillation would never have started.

Also, having the hashrate rise faster than it drops is good in that it prevents hyperinflation. As long as the drop back down  is fast enough to keep block times reasonable (at least under an hour), that's more suitable for reaching an equilibrium state.

Check out the rest of my comment, You've answered me before I wrotte the second part, also thank you for 10.7 information not alot of us had this piece of information as we all got the 36SMA12LMT as 12% goes both ways.

We are a community coin. We don't add features that most people can't relate to. If we add black magic, fewer people will contribute to the vitality of the coin. Then we die.

We are going to restructure the coin so that this stuff is easier to deal with.

Correct me if I'm wrong, but from your comment I get the feeling that you didn't make the effort to look it up and understand how it work, and I think it's fair to make a judgement about it after you do those things. and btw it is a simpler solution than SMA36LIM12

You are even wronger. SMA36LIM12 is simpler than the ArbitrarySciFiVariableName spaghetti you posted.

It's not whether it works. Maybe it's perfect (except for WTF broken FPU!), but it matters if the average person understands it. If they don't, then they have no voice.

Kuroman, I like the analysis you've provided in the past.  I understand what you're saying about our cycles here and think I can help build a bridge between you and Mav.  It's important, first, to remember that this fork was not intended to be a 100% solution.  It could have been with a bit more time (or less wasted time dealing with trolls) but we didn't have the time available.  This is designed to get us moving and to make sure we never got stuck too high or too low again.  As Mav has said multiple times in various venues, this is at best an 80% solution.  We CAN stay here for months and ride the waves - we'll be alive, we'll be attracting enough miners, and end-users can actually start using the coin.  This gives us a working platform the PR team can work from - and they're kicking ass!

Yes - we can and will do better!  And now, because of this fork, we have plenty of time to create a 100% solution without having a gun to our heads.

A quick additional comment - we know from our testnet work that this algo WILL converge and stabilize.  We didn't have that before.  Also - keep in mind that in conjunction with this fork we also got a significant endorsement from the Oprah Winfrey empire - I'll guarantee you that we didn't see that coming.

I hope you and envy will continue to work towards the long-term goal of getting CAT to Mars.  

Andy

Andy thank you for your comment, the fact is the 12% limite is supposed to make the diff converge indeed, but that is under some strict conditions such as, fix hashrate jumps, the jump of hash rate is made when the block time is 600s ....ect which is definitly not the case not to mention that the SMA36 makes everything goes beserk and due to other variables it is countering the convergence from the 12% limite.

It uses this function KGW = 1 + (0.7084 * pow((double(PastBlocksMass)/double(144)), -1.228))
So if fast rate is high the function above is applied if/when it's the opposite 1/KGW is applied
Here is the whole code behind it, there is nothing magical about it just a simple math formula applied dynamicly :

No explanation of the function. Not magical. Magical. I don't think it means what you think it means. So, at this point it is magical.

I have looked at it. I think I know what it does. I also think it can be done a lot better. Without magic.

I over-estimated your understanding capabilities it seems, if you can't figure out as much that you have KGM applied when hashrate increase drastically and 1/KGW applied when it deacrease dramatically while KGW is a clear formula and the code it self is clear than writing the SMA36LIM12.

If you say it's hard then explane to me why every single new scrypt altcoin use this solution? You are just being overprotective for a solution that doesn't work properly.

Kimoto is not simple.
Kimoto antagonizes the coders against the users.
Kimoto is not explained.
Kimoto has competition.

Let me fix that for you

-You don't get it =/= not simple
-You don't know how to code (neither do I since it's not my domaine of expertise but I can read C and can understand alghorithms) but since almost every new scrypt altcoin uses it, it proves how wrong your are and how it is easy to implement.
-It is explaned and I posted the code which is simple, so again if you don't get it and having trouble understanding does mean it is not explaned, and I'll give you an example, even if I give a primary school kid the book or just the lesson on a high school math subject, he will not understand it and think it's antagonizing, not explained...  ect ect he will have the same reaction as you right now, while the lesson on it self is nothing much just some high school math.
-And what the problem if it having competition (why don't give some example while we are at it) and competition is good, it means that the best solution will prevail, and will push developers to innovate more, Quantum Mechanics woudn't be where it is if it wasn't to it competition with classical physics and general relativity, if it wasn't for the competition between Nicolas Tesla and Thomas Edison, you won't be using alternative current at your home and electronics won't have involved as much... ... ...

This may be obvious to you, kuroman, but how many people on this forum do you think are at least marginal C programmers?

Let's all use words, examples, and conversations that are available to the entire community.  Limiting the conversation to a few, either through expediency or hubris, doesn't help the greater good.

No solutions are off the table right now - not all possible solutions are ON the table yet. LOL  We're not yet at the stage where we need to fight over options.  Thankfully, we now have time to propose and more importantly TEST the solutions on the testnet, not just on an Excel spreadsheet or in a Matlab pageprint.

Everyone - put your 'thinking caps' on and get those ideas on the table!

Thank you!

I don't understand your point here? I posted the KGW code to people that are concerned about it and were claiming it is overly complicated which is not the case as proven is explaned several times.

The solution works, not on a testnet which doesn't represent the real life conditions, but on real coins that've been and being hit by profitability pools.

NB: this is why I didn't want to join the catcoin-dev channel discussion despite being present and folowing what happening there, people are blankly dismissing other people ideas, I saw some heated discussions, where no one want to make compromises. Also calling people math geek/nerds and discrediting them on that same channel doesn't help really because and let me remind you however you break the discussion the basic level of all this will be math and code.
And this kind of agressivness and stance that pushs people off (especially from Zerodrama), and it is the reason why I stoped and many did at some point when the fork plan was initiated from commenting, and just to make things clear, I'm not criticizing the fact that we've forked, and I understand very much the principale on which the solution works on and how it was supposed to solve the problem heck I even defended it against people that made a rushed judgement and said it doesn't work at day one of fork and told them to let trough some cycles my comment is still there, but after 9-10 cycles I thinks it's fair enough to have an initial judgement and I've given my analysis on why it doesn't work, and a fix from my percpective while rooting for an available solution that is proven to works properly .

[Maverickthenoob]

ENOUGH. Everybody take a step back for a second. Dev team included.

A little (blatantly obvious) honesty: The current algorithm didn't work as well as we planned, but to those of us tearing us apart for it, I didn't exactly see any of you lining up at the door with any of the simulations I'm seeing now. I would have put a stop to this before we even released it if we had, or at least tweaked it some more to make it less (over)reactive.

I need you guys to all stop and take a breath and open your minds to solutions, and by all of you I mean ALL OF YOU, my team included.

This is the situation we can't control now, we released the fork, it's done, the coin arguably functions better than it did before, because at least now the confirmations don't full stop for 4+ days at a time. This is critical to the survival of this coin, and buys us time. I'm not happy this didn't solve the problem, and releasing a band-aid wasn't really the goal, and guys honestly, that is somewhat on me. I'm sorry I let you as a community down. I really am. I have a lot going on in my life, I'm getting married next week and I've been ill for the majority of this week. It's not an excuse, but merely an explanation. Please remember that I'm human and my team is human, and we had to pick this coin up quickly in less than ideal circumstances. We've established roles, and everybody is aware of what they contribute. Lessons were learned during all of this, so please, it was all part of the process. I'd like to think that we have one of the most active dev teams of any of the coins out there, that's got to be worth something. We also have one of the most active and vocal communities out there, and that is definitely worth something.

Now is not the time to tear each other apart over mistakes that were made though, I can't do any more but apologize for what happened. I can promise you I will do better and my team will do better in the future. But, we need your support, and we need some of the expertise that has shown up in the past 24 hours. It's just a fact of life, I said when we built this team we were always looking for more talent. We still are, and always will be. As an example of this I recently added Johnny_non to our team, who I don't think anybody can argue with has done an UNBELIEVABLE job turning around our PR front and websites. I'd like to take a moment to recognize that dedication, because it's been extremely important. He's got some cool shit brewing in the wings too that will be discussed here soon. The Oprah thing is absolutely huge for us. I can't stress that enough.

As for our coins future, enough bickering, bickering and name calling is not productive, one thing is for certain: this coin's community certainly does not lack in color and contrast. We have several options in front of us, and we're looking for more.

Yes, the Gravity Well is one of them, hozer's concern revolves around the floating point numbers that the algorithm uses to calculate difficulty. He can get into his concerns if he wishes. That algorithm is by no means off the table, however.

envy2010 and kuroman, and strelok have provided very detailed feedback which I've been looking through at length. I think the solution is in front of us, we just need to AS A COMMUNITY decide what direction to take this coin and what we feel the best approach will be to resolve this once and for all. We have bought ourselves significantly more time with this last fork, obviously it's not ideal, but it still functions as a coin.

I hope that perhaps everybody can take a step back and breathe and stop fighting like cats, and that a calm level head will provide us with a permanent solution that takes this coin to the next level.

Thanks for your time.

Kevin McCurdy

[SlimePuppy]

Gravity Well is simple to understand, and fixes a lot of problems that others have complained about. I don't understand why people are so dismissive of it. I've heard complaints that it only addresses specific problems. Well maybe the problem is you're trying to do too much with the difficulty algorithm.

It might fix at least some problems - how many others does it cause?  We don't know yet.  Some have already looked into the algo and have found ways for it to be exploited.  Consider a possible future where a large number of coins adopt the same fad algo - they become targets for cracking development.  I'm not saying it's a guaranteed problem, but it is something we have to consider as we evaluate ALL the options.

If you want to reward loyal Cat miners/investors, then introduce a small PoS element.

I cannot understand why you would suggest that we should not be using hash manipulation (even though we're not and I don't recall anyone in the dev team suggesting it) yet you would suggest revamping the coin to manipulate via a 'loyalty' system.  I mine at Coinium because I like their loyalty system.  I and others believe that modifying the payout system belongs at the pool level, not in the guts of the coin.  Changing the way coins pay out - regardless of what one calls it - changes the coin.  So far, this community has made clear they want this coin with these parameters.  Our work so far has been to stabilize the coin and continue to work towards the 10 minute block target.  So far so good.

Satoshi said one-cpu-one-vote. It sounds like some people want to undermine this principle. You can reward loyality, but you can't punish greed.

Actually, one can punish greed.  The coin already rewards loyalty.  Nobody in the dev process wants to undermine the coin, the community or St Satoshi, er Saint Catoshi.