Bitcoin Forum
January 31, 2023, 12:16:28 AM *
News: Latest Bitcoin Core release: 24.0.1 [Torrent]
 
   Home   Help Search Login Register More  
Pages: [1] 2 »  All
  Print  
Author Topic: How to design a CoinChoose-resistant coin?  (Read 1565 times)
mat5x (OP)
Full Member
***
Offline Offline

Activity: 181
Merit: 100


View Profile
June 09, 2013, 03:59:43 PM
 #1

The biggest threat to altcoins isnt ASIC mining BTC and just buying up lots of ALT for cheap, depressing the price anymore.

Coinchoose pretends that a 0.5BTC deep market represents the pricing value on 100s of BTC of several hours/days of mining by punters tracking most profitable coin. Punters think they can all mine 10000s of ALT then go to the market anytime and get full value by dumping 100x the coin volume of ALT ever traded. Ridiculous.

This creates an unnatural (and unattended!) hashing wave for new ALT, which pumps up the difficulty after a time, and then, like most coins, cause a 4x increase in diff, drops the price on the floor.

This isnt healthy for any coin, but the autominers tracking the bogus valuations on CC are a force to be watched, with 100-200MHs on scrypt, if not more.

It doesnt matter if someone DOS's CC again like yesterday, some dumbass site with some random valuation will appear, everyone will take its math as gospel, and point their miners at it (it DOES however provide some WONDERFUL price gaming opportunities, which are pretty hard to resist!) But that really fucks the market up.

So leave CC and clones as is, since there's nothing to be done about ignorance. The question is how to code a coin that resists such pump/dump waves.

Obviously a coin that retargs every block would be great, but then you have a static value vs radical change in difficulty issue, which results in TRC's yoyo difficulties.

ELC however values coins for the hashrate put in, which makes that yoyoing less effective. Unfortunately, ELC nInterval=2160 * 120s.  This mostly causes STALLCOIN (unrelated to STABLECOIN, actually, which is about horses I think), tho with ELC there's much less value loss to miners than most ALT because value ~ diff in ELC.

But ELC wont solve this, its subject to CCwaves just as much as any ALT.

I have a few ideas, but id like to get more opinions on CC's effect on ALT market first.



Elacoin/ELC - block reward proportional-to-difficulty coin! http://elacoin.com
1675124188
Hero Member
*
Offline Offline

Posts: 1675124188

View Profile Personal Message (Offline)

Ignore
1675124188
Reply with quote  #2

1675124188
Report to moderator
1675124188
Hero Member
*
Offline Offline

Posts: 1675124188

View Profile Personal Message (Offline)

Ignore
1675124188
Reply with quote  #2

1675124188
Report to moderator
1675124188
Hero Member
*
Offline Offline

Posts: 1675124188

View Profile Personal Message (Offline)

Ignore
1675124188
Reply with quote  #2

1675124188
Report to moderator
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
1675124188
Hero Member
*
Offline Offline

Posts: 1675124188

View Profile Personal Message (Offline)

Ignore
1675124188
Reply with quote  #2

1675124188
Report to moderator
ImI
Legendary
*
Offline Offline

Activity: 1946
Merit: 1019



View Profile
June 09, 2013, 04:02:06 PM
 #2


I thought novacoin and bitbar are retargeting every block?
mat5x (OP)
Full Member
***
Offline Offline

Activity: 181
Merit: 100


View Profile
June 09, 2013, 04:12:46 PM
 #3


I thought novacoin and bitbar are retargeting every block?

Ill have to take a closer look. NVC was launched before CC existed or became popular, so any coin that was lucky enough to get established before CCwaves is likely unaffected by it (until that one day they read 500% PROFITABLE...).

Bitbar ill have to check closely since its a lot newer. Ill have to see how they avoid the huge hashwave diff retarg issues without turning into a TRC yoyo.

Funny stuff, I ran some spreadsheet models on ELC with shorter retarg nInterval, and it makes pretty obvious the 30% transfer of value from future miners the CCwavers cashing in on easy diff, moving it to hard diff. (max retarg up and down on most coins is 4x, so its easy to model: throw 150MH/s ontop of 20 exactly at the first block of retarg and remove it last block of retarg and see what happens even with a sliding window avg over the next nInterval).

And thats with a coin that pays per diff, it must be much worse with a static diff.

Retargetting every block would be interesting, but there's 2 issues

1st is say there was no luck involved, and hashrate = blocktime directly. Then difficulty you are mining is actually from the previous block's time. So if 100Mh/s leaves the coin and 20MH/s remains, you're mining a coin at 5x the diff for the same reward, until retarg next block. If the blocks are long enough, creates an opportunity to jump in and out of the coin quickly to mine only the easy blocks with high hashpower, effectively transferring value from high diff miners to easy diff miners. Unfair.

2nd issue of course is luck. Cant have too small a window to measure diff or a wayward lucky or unlucky block could send the coin sprawling into high/low diffs. Most coins have /4 x4 max scaling, but 4x 2 or 3 times in a row is a big diff change. 4^3 = 64x.

Elacoin/ELC - block reward proportional-to-difficulty coin! http://elacoin.com
qiuness
Hero Member
*****
Offline Offline

Activity: 714
Merit: 500



View Profile
June 09, 2013, 04:30:04 PM
 #4

the problem is not coinchoose, it is all the shitcoins that are being released
XRcode
Full Member
***
Offline Offline

Activity: 134
Merit: 100


View Profile
June 09, 2013, 04:32:51 PM
 #5

the problem is not coinchoose, it is all the shitcoins that are being released


You are partly right, the problem is not coinchoose... It's the idiots who believe it and run CryptoSwitcher Smiley
anderl
Hero Member
*****
Offline Offline

Activity: 714
Merit: 500



View Profile
June 09, 2013, 04:34:12 PM
 #6

an unskilled worker blames his tools.  not sure what you mean by "coinchoose resistant"  seems like you don't understand economics or programming.  another nonsense thread on bitcointalk.
mebezac
Member
**
Offline Offline

Activity: 112
Merit: 10



View Profile
June 09, 2013, 04:35:34 PM
 #7

This is probably a stupid idea, but how about a coin that starts with a decent difficulty, say 5 or 10 and then doesn't have a retarget, just a static difficulty. You would have to keep interest in the coin or it could end up taking a long time to solve blocks and do transactions. Like I said, this is a stupid idea as a currency because there's no way to know how long blocks and transaction times will take. But, as far as not being subject to crazy difficulty swings, a static difficulty would take care of that part of the coinchoose raping.

BTC: 168d57nW72Y6DidPgE88iL7vYmpQD45dYK|LTC: LLPttXuFF2uTo2CWeEwXU5CwcUcgQ4NGd3
XRcode
Full Member
***
Offline Offline

Activity: 134
Merit: 100


View Profile
June 09, 2013, 04:38:24 PM
 #8

an unskilled worker blames his tools.  not sure what you mean by "coinchoose resistant"  seems like you don't understand economics or programming.  another nonsense thread on bitcointalk.

Do you understand cryptocurrency?
Since when is it normal to have the network hashpower increase by 500x in a few minutes?
vinne81
Full Member
***
Offline Offline

Activity: 182
Merit: 100



View Profile
June 09, 2013, 04:38:37 PM
 #9

This is probably a stupid idea, but how about a coin that starts with a decent difficulty, say 5 or 10 and then doesn't have a retarget, just a static difficulty. You would have to keep interest in the coin or it could end up taking a long time to solve blocks and do transactions. Like I said, this is a stupid idea as a currency because there's no way to know how long blocks and transaction times will take. But, as far as not being subject to crazy difficulty swings, a static difficulty would take care of that part of the coinchoose raping.

So then when the network hash rate multiplies by 1000x, you could have (depending on the intended hash rate) blocks each second? Hello orphans
mat5x (OP)
Full Member
***
Offline Offline

Activity: 181
Merit: 100


View Profile
June 09, 2013, 04:44:49 PM
 #10

the problem is not coinchoose, it is all the shitcoins that are being released

So we should only have BTC? HOw about not even having BTC and having nothing? (Just taking your argument further).

Which central authority would you like to pre-approve coins for release, the US Govt? (im sure they're already on it, and BTC isnt in their list.)

This is the free(est?) market here, people can invent whatever they want. This allows us to discuss and create lots of coins.

Some coins fail because they're stupid, some just because of CC, some because of bad timing or luck on the market. Some really good coins have been killed (CoiledCoin) by assholes, some almost killed by premine fiascos (Novacoin).

Itll all come out in the wash. Its great that there's so many new coins, but someone needs to setup a good altcoin directory cuz I cant even keep track of what new coins EXIST never mind what their features are - and then, most importantly - what the unintended side effects are. That last thing is the biggest problem with most alts.


Elacoin/ELC - block reward proportional-to-difficulty coin! http://elacoin.com
mat5x (OP)
Full Member
***
Offline Offline

Activity: 181
Merit: 100


View Profile
June 09, 2013, 04:49:11 PM
 #11

This is probably a stupid idea, but how about a coin that starts with a decent difficulty, say 5 or 10 and then doesn't have a retarget, just a static difficulty. You would have to keep interest in the coin or it could end up taking a long time to solve blocks and do transactions. Like I said, this is a stupid idea as a currency because there's no way to know how long blocks and transaction times will take. But, as far as not being subject to crazy difficulty swings, a static difficulty would take care of that part of the coinchoose raping.

So then when the network hash rate multiplies by 1000x, you could have (depending on the intended hash rate) blocks each second? Hello orphans

Mebezac: I thoguht of this exactly - as thats what ELC is effectively. Static reward, static diff. Problem as Vinne81 says, it's network issues. If the protocol could handle 1000s of blocks per second, then we'd have no issues, but we cant. The competitive nature of the protocol to produce a winning chain is an elegant way to solve concurrency conflicts, but costs us in response times.

People think a tiny coin like FASTCOIN is ok with 7 second blocktimes, but wait til the network diameter gets a little bigger. Orphans benefit bigrigfarms as they can churn out not just new blocks but new CHAINS faster than the rest of the network. Luck itself causes mini 51%-attacks for everyone naturally, but it's exponentially higher for the more MH/s you have (and vanishingly small chance for you who has 0.1% of the hashrate).

Vinne81: this is why we designed ELC, tho I think the block retarg (nInterval) is too long. A sliding window with shorter #blocks would be better, but i have other ideas to improve on that even further - even with a sliding window, value is transferred from future miners to WaveMiners during low diff.

My idea is hyper radical, next post.

Elacoin/ELC - block reward proportional-to-difficulty coin! http://elacoin.com
mebezac
Member
**
Offline Offline

Activity: 112
Merit: 10



View Profile
June 09, 2013, 04:51:32 PM
 #12

This is probably a stupid idea, but how about a coin that starts with a decent difficulty, say 5 or 10 and then doesn't have a retarget, just a static difficulty. You would have to keep interest in the coin or it could end up taking a long time to solve blocks and do transactions. Like I said, this is a stupid idea as a currency because there's no way to know how long blocks and transaction times will take. But, as far as not being subject to crazy difficulty swings, a static difficulty would take care of that part of the coinchoose raping.

So then when the network hash rate multiplies by 1000x, you could have (depending on the intended hash rate) blocks each second? Hello orphans

Exactly, like I said, a stupid idea, but one that  takes care of crazy difficulty swings. As others have said, I think people blindly following coinchoose have every right to in a free market. I'll just stick to mining litecoin and bitcoin

BTC: 168d57nW72Y6DidPgE88iL7vYmpQD45dYK|LTC: LLPttXuFF2uTo2CWeEwXU5CwcUcgQ4NGd3
anderl
Hero Member
*****
Offline Offline

Activity: 714
Merit: 500



View Profile
June 09, 2013, 04:51:46 PM
 #13

an unskilled worker blames his tools.  not sure what you mean by "coinchoose resistant"  seems like you don't understand economics or programming.  another nonsense thread on bitcointalk.

Do you understand cryptocurrency?
Since when is it normal to have the network hashpower increase by 500x in a few minutes?

bitcoin and asicminer is a good example.  "since when is it normal"  not sure that equates to cryptocurrencies but it happens alot in regular life.  securities markets are a good example.  company issues a press release and millions or billions of dollars flow into the stock.  are you trying to support an idea to make markets resistance to buying and selling type mechanisms?  Again nonsense thread.
sal002
Hero Member
*****
Offline Offline

Activity: 490
Merit: 500


View Profile WWW
June 09, 2013, 04:52:15 PM
 #14

FYI - YaCoin is CoinChoose resilient.  Guess why?
mebezac
Member
**
Offline Offline

Activity: 112
Merit: 10



View Profile
June 09, 2013, 04:57:06 PM
 #15

FYI - YaCoin is CoinChoose resilient.  Guess why?

I think I remember seeing it listed once very shortly and the profitability was like 300% but maybe I was just seeing things.

BTC: 168d57nW72Y6DidPgE88iL7vYmpQD45dYK|LTC: LLPttXuFF2uTo2CWeEwXU5CwcUcgQ4NGd3
sal002
Hero Member
*****
Offline Offline

Activity: 490
Merit: 500


View Profile WWW
June 09, 2013, 04:58:44 PM
 #16

FYI - YaCoin is CoinChoose resilient.  Guess why?

I think I remember seeing it listed once very shortly and the profitability was like 300% but maybe I was just seeing things.

Yep.  Can't mine it with a GPU so it was delisted
mat5x (OP)
Full Member
***
Offline Offline

Activity: 181
Merit: 100


View Profile
June 09, 2013, 05:00:46 PM
 #17

My idea is this for a CCresistant coin:

- scrypt
- 1 min nominal blocktime.
- retarg every block
- sliding window of 2 blocks avg diff calc
- reward = difficulty
- slow value decay of *T/(1+t) or gentler of value to combat inflation for scryptASIC release in future (t = time to date avg of last 200 blocks (avoid timestamp manip), T = constant Time factor)
- max incr/decr 16x or /16
- if x16, then block time drops to 1/4 of normal (15 seconds). value 1/4'd. still high enough to avoid too many orphans from goodluck series of <1s blocks.
- if /16 then block time drops to 1/4 of normal (15 seconds). value is 1/4'd.
- similar scaling for x8 or /8, then blocktime 1/2 etc
- 1/4 do blocktime 2/3rd or something. (ideally 1/sqrt of increase but i dont want too much floatmath in the code, who knows what kinda IEEE implimentations various hardware has)

The 1/4 blocktime is a good idea to allow a recalc on difficulty faster after a large change. Additionally, it makes blocks so fast, it makes it hard for hoppers to come in and out of the coin exactly on time to cause a reinforcing frequency wave like TRC has (which is over many minutes or hours for difficulty manipulation).

The 1/4 time thing effectively seperates luck from actual hashpower. Could even go 1/8th perhaps (if peopel put up with fastcoin's 7s, then for the purpose of a retarg its worth the cost of luck intefering abit).

The 2 block average really hurts luck's chances of causing two x16 increases in a row randomly.

One caveat of course is that the average time of any number of blocks isnt static (tho, check BTC< it isnt either! blocks in last interval were <10minutes by a large margin, moving from 12.2 to 15.6 difficulty!)

I still need a rundown on what the frequency of lucky blocks are producing 1/n or lower block time of normal (ie <=1/4 of normal time, how often?). My stats arent strong enough here. Lilhelp?

Elacoin/ELC - block reward proportional-to-difficulty coin! http://elacoin.com
mat5x (OP)
Full Member
***
Offline Offline

Activity: 181
Merit: 100


View Profile
June 09, 2013, 05:06:28 PM
 #18

FYI - YaCoin is CoinChoose resilient.  Guess why?

Yep.  Can't mine it with a GPU so it was delisted

Whats your reasoning behind that sal002?

YACcoin is a great thing tho, and Subtrata was 100% right about it. It encourages botnet mining, so that means when every other coin is arrested and jailed off the face of the earth, YACoin will still be around.

to a previous replier regarding "resisting market tools to assist trading":

As for market valuations, there's no reason to value a coin on what the profit of the NEXT BLOCK is. I dont see how that's intelligent either. Anyway people can do what they want, but i think you'll see that the value supposedly created by CC's predictions has NOT materialized at all. As I say, do some calculations. See the hashpower waves that come in from CC, calculate how many BTC in value that cost in hashpower, and see if the market has paid that back out (modulo people holding it). I think you'll find that the value is being 'held' by 90% because its just not materializing. Soon as any flash mined coins are dumped the value plummets because it was never actually there in the first place.

Also, we can do whatever we'd like, and I dont see how a coin that doesnt manage to allow itself to be gamed by huge hashpower waves is a detriment to the community. Explain how that works out.

Elacoin/ELC - block reward proportional-to-difficulty coin! http://elacoin.com
area
Full Member
***
Offline Offline

Activity: 177
Merit: 100


View Profile
June 09, 2013, 05:58:16 PM
 #19

Also, we can do whatever we'd like, and I dont see how a coin that doesnt manage to allow itself to be gamed by huge hashpower waves is a detriment to the community. Explain how that works out.

I don't think that's what anyone's saying. A coin that manages to solve this problem certainly brings something new to the table, and PPCoin's proof-of-stake system, if it takes over from proof-of-work eventually as planned, probably gets there.

I am glad that this conversation is happening, however, as instigating it was part of my goal of releasing CryptoSwitcher rather than keeping it for personal use.
mat5x (OP)
Full Member
***
Offline Offline

Activity: 181
Merit: 100


View Profile
June 09, 2013, 07:27:52 PM
 #20

Also, we can do whatever we'd like, and I dont see how a coin that doesnt manage to allow itself to be gamed by huge hashpower waves is a detriment to the community. Explain how that works out.

I don't think that's what anyone's saying. A coin that manages to solve this problem certainly brings something new to the table, and PPCoin's proof-of-stake system, if it takes over from proof-of-work eventually as planned, probably gets there.

I am glad that this conversation is happening, however, as instigating it was part of my goal of releasing CryptoSwitcher rather than keeping it for personal use.

Hilariously you may have succeeded in your ironic goal: to raze the field of toxic ALT polluting our pure BTC/LTC/??? harvests. Lol.

'eventually' taking over unfortunately is the only thing PoS _CAN_ do, you have to wait til there's enough stake evolved. And you cant rush these things, or its too easy to get a high stake, then there's nothing seperating it from regular early mine win which destroys ALT.

At least all ALT now have an increasing reward release with block #. A side effect of ELAcoing ill have it take credit for, partially.

At any rate, consider the 1-block retarg idea... yah?

Elacoin/ELC - block reward proportional-to-difficulty coin! http://elacoin.com
Pages: [1] 2 »  All
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!