Topic: Possible Thoughts / Ideas for a new Coin. Real Improvements. Add your thoughts (Read 672 times)

It would be really neat if the mining QT would start advancing further to be more like CGMiner, for general public use and adoption.

There could easily be a tool to list all existing pools and allow the user to select which one they want and start mining with a few simple clicks.

A more advanced one could calculate profitability on mining per pool, mining pool up & down times, known outages etc. Bring data mining to the crypto world and use it for some advanced real-time analyics.

There could also be a built in 51% pool protection by switching users or preventing mining on specific pools by blocking them when they hit high 45% of the network hash rate.

The QT could also become an information hub  if the QT included a small web connected section for checking out latest updates, notes from the dev, or just news.

---

As for the coin itself the next coin improvement needs to be making it easy for transactions. The coin needs an advanced POS system.

The coin maybe able to improve transactions by separating payments into single and group transactions.

Group transactions would allow for coins to be pre-confirmed in small amounts of 0.1 or 0.01 ,etc depending on the price of the coin, and then use those small pre-confirmed transactions to group together to make 1 much faster transaction of any amount. Say you want to spend 1.5 coin on a hip new Beiber cassette, if the pre-confirmed amount was 0.01 coins then you would issue a group transaction of 150 pre-approved micro transactions.

This would require consumers to siphon off a bit of their coins prior to going out and using them so that they can be confirmed, stay confirmed, and then later be used for faster transactions. If the consumer wanted their coins back for trading purposes they could just move them back.

How about this idea . . .

Each block creates some coins that go towards a bounty fund.

People or organizations could nominate themselves to be bounty fund recipients. They could state publicly what they plan to do with the bounty funds. The community could evaluate their credibility.

Owners of the coin could vote as to who gets the bounty. Voting rights proportionate to ownership stake in the coin. Instant run-off voting.

This ensures that
1) The cost of benefit is shared equally among the coin owners
2) A fair public contest for the bounties
3) Non-performing bounty winners can be identified by the community and suffer in future elections - or coin owners could vote as to whether to release the bounty.
 

Sorry to cause confusion.

Read shakezula's guide!

Just to note on a couple things for clarity and to help


Correct on block timings;  though that exploit has been / is used today with current alts.  The NTP timestamp Fix from GG I point out in the thread above solves this. also, why it would be required.



CPU>GPU>FPGA>ASIC>? is not a flaw; it's an inevitability. 
BTC was CPU only when first released.  It was discussed in 2009 about using GPU; but was not utilized until later.  GPU for scrypt was, in fact, "cracked" to some degree Prior to LTC release.  Here again, it was not utilized until later when the coin had more size/age.  The problem is that there is no way, that I am aware of; to make a CPU only coin without making it fully ASIC and FPGA ready.  Even though, one could perceive that LTC failed with GPU's;  it is a large driving factor in it's growth, value and size today (Including alts).  Since the value of BTC has risen to a point where ASIC's and FPGA's are profitable to manufacture; miners are "extending" or "Continuing" their ability to mine with LTC (others too of course).  If we make an assumption that the world of crypto is to continue and be around for a few years;  it is very likely that LTC will also become FPGA / ASIC dominate.  Fortunately / Unfortunately (depending on your P.O.V.), GPU is basically the "bottom tier" of mining equipment.  I imagine it won't be long before BTC is entirely mined by corporations like Asicminer. 

If your with me at this point;  we are basically at the same point in time, at which, LTC was created.  There had just been a spike in BTC price, then.  There had been a "small" flood of alts, then.  People were looking to extend their CPU's, no longer feasible on BTC, then.   The value of BTC spiking; had shown legitimacy and that another coin had the possibility of "making it", then.
  To a large degree, this same scenario is playing out now;  only with GPU's needing a spot soon (from ASIC and FPGA) and LTC value rise showing the possibility of another coin "making it".   So essentially, to protect the "home miner", we need to be ASIC / FPGA resistant and GPU friendly;  which would be the modern equivalent, that LTC was at it's inception.  This is where the Super Scrypt algorithm comes into place.  Using the higher parameters should provide an equal (if not longer) time extension to GPU mining. 



I agree sadly, greed drives coins of "today" and people expect value for adoption.  However, greed could not have driven the first 2 years of Bitcoin; since it had no value and would have never started / made it to where we are now. 


Sorry to have popped in there.  Thank all for the thoughts so far! 

block rewards should increase with diff.

I'd like to see an altcoin that incorporates two ideas, which I consider to be Bitcoin flaws . . .

1. CPU-Only Mining. So that it is easy to convert electricity into coin - reduces reliance on exchanges, helps widespread adoption. (Litecoin tried to do this, and failed, and seemingly succeeded as a coin anyway).

2. Adjustable Inflation Rate. Block reward is set by vote - voting rights awarded in proportion to coin ownership -> Coin owners can be relied upon to vote for an inflation rate that maximizes coin value -> increasing coin value drives adoption.
   

how about no more new coins.

Thanks.

You've got some good ideas here. Also good roundup of all the issues to consider with a new cc. I like your divisions of 'birthing, survivability and long-term growth'.

>Variable block reward.
That sounds very innovative and solves some problems, but I wonder if it doesn't create an incentive for miners to falsify their block timings.

>A GPU miner program would need to be altered to allow these parameters.   In a perfect world, one of the community involved would have the miner ready upon coin release.
Interested to know why you think supporting GPU mining is a good idea. Litecoin was started with GPU-hostility in mind . . just the implementation wasn't good enough. Efficient GPU mining narrows down the mining pool, and hampers widespread adoption.

>provided greed, profitability or instant return is left second to design.

I think greed drives the adoption of new coins. Help early adopters get rich - early adopters will help your altcoin. Maintaining and increasing the value of the coin should be a primary concern.

My hope here is to merely spark ideas, or better, a discussion on how we (as a community) can develop a legitimate, survivable and working new Crypto-coin.  There are many great people, thinkers, coders, hackers etc. on this site.  There should be no reason, we as a whole, cannot develop a good coin........provided greed, profitability or instant return is left second to design.

I've been watching the alt scene for a bit and pondering various issues (mostly birthing) with these.  As well, I've been digging around various older failed coins from years past; just to learn more.   I've tried to compare why certain coins failed and the few have continued.    I believe there are a few things that can be implemented into an Altcoin that should help birthing, survivability and long-term growth.
 
I have NO personal interest in developing or being involved in a coin myself.  These are just merely suggestions / other past "fixes" etc that may prove useful today.  This is taking from the assumption that a person is looking to make a "Real crypto" not a pump and dump.  As well this is taking into assumption of said coin possibly being (eventually) a 3rd or 4th running coin.  E.G- Copper relation

It seems we can bulk most coin fails into a few categories;  Pre-mine, failed launch and exploits.  
I did not include "clone" coins as a category for failure; as it seems they failed primarily from exploits or pre-mine rather than being a clone itself.

Another key point, if we are to consider crypto successes BTC and LTC;  NEITHER of these coins were developed or released with Profit, Quick return, or dollar value in mind first.  Specifically with BTC, it had a high chance to never reach even 1 dollar in value; yet people mined and supported.  It seems now, people place dollar value or quick return AHEAD of design, innovation, development, philosophy of the coin, etc.  Ultimately, if this is ones' mindset, there is no need for improvements or innovations;  however I hope that is not the case, with the community at large.


It seems there are a few "given's" that will have to be accepted with any new release;
-Insta-mining will be attempted
-pump and dump is first assumption
-large early hash rate fluxuations
-extreme likelyhood of exploitation attempts
-high public doubt of scam intent
-high 51% attack vulnerability depending on algorithm


Ok so onto specifics.

The obvious;
-Pre-launch thread announcement that is separate from the "ANN release thread" and a minimum of 1 week lead time from launch.
-Have a public vote for release Time in the pre-launch thread.
-Have a Testnet operational for the coin during the pre-launch window.  This will let dev's hash bugs and users to get their systems setup and ready. (similar to LTC)
-Have Linux, Mac and Windows downloads available at the beginning of pre-launch
-Have a website and pool(s) available at the beginning or durning pre-launch window.
-No Pre-mine!  Genesis Block(s) only.  Pre-mine was not used in either LTC/BTC (yes genesis).  



Thoughts on General aspects;
(numbers variable to some degree of course, not a big deal for most items.  Current numbers are appx. 36% more than LTC, so 2/3 relative value)
-110 second block time (1 min 50 sec). Satoshi chose 10 min for "long-term" latency issues.  General discussion area also talked of lowering BTC time.....2 min is about lowest.  Lower times increase vulnerability.
- 8 Confirmations  You can read up on orphan rates relative to confirmations.  but here, 8 will be more secure (relative to hash power req'd to orphan) than BTC or LTC AND less total time. (Total Confirm time: 14:40) Main reason for 110 sec blocks.
-785 Blocks per day (appx)
-120 Million Coins total
-Halving @ 1,145,433 Blocks (appx 4 years)
-50 Coins per block ***Ancillary Variable, see below
-39,272 Set coins per day
-Difficulty re-target every 2356 blocks (appx 3 days)  
-Primary difficulty use standard *4/4
-Starting difficulty of 1 **see algo below

Suggestions for Innovations / Improvements:

-Correct off-by-one difficulty re-targetting (not been done that I know of) Not an issue for BTC anymore, any startup coin should fix this.....

-Lower Acceptance window!!!  This is and has been an issue for recent altcoins and forks.  Lowering the acceptance window, alone can help reduce exlploits and 51% attacks;  but is only part of the solution.  With the above suggested block time;  I would recommend a window of 20-30sec. Must be < 1min. to properly use the variable block reward shown below.

-NTP Timestamp fix from GG.  In order to properly utilize the point above and below this is required!  This patch virtually eliminates "Time Fixing"......another issue with some recent coins.

-Variable block reward.  Though it may not appear as such, this is one key component to prevent exploit, insta-mine and 51% attacks.  It does REQUIRE the NTP fix from GG to prevent time fixing though.  The system here can be thought of as a 1 block re-target (ONLY in terms of reward!).  The applied formula would be fairly simple:  (Block Find Time / Block Target Time)*Block Reward.
   For example, if we assume above 110 second block and 50 coin reward;  A block found in 30 seconds would pay (30/110 =.2727*50) 13.636 Coins for that block.  In contrast, a block found in 200 seconds would pay (200/110=1.818*50) 90.909 Coins for that block.  This method should have NO cap on +/- adjustment as it is ONLY for a block payout not actual difficulty.  
This ALONE can add several huge features to a coin.  
*Exploitation and 51% resistance via diminishing returns. There is no way to prevent exploit or 51% attempts; however this should "de-incentivize" an attack dramatically (to the point of not profitable).
*Helps prevent Insta-mining early at low difficulty via diminishing returns.
*Helps reinforce and stabilize hash rate to appropriate block times for any given difficulty.
*ADDS incentive for miners to support the network during starting / slow times; or even if difficulty has been "left high". (e.g. from higher block reward with slow "under hashed" finds / times)
*Does NOT remove underlying RNG from block finds;  only institutes proportional payouts of said luck.
*Improves Coin generation stability over long periods for miners; relative to their speed.
*Bypasses many of the issues caused with long difficulty re-targets and newer coins. Thus allows longer targets,  like 2356 as suggested and a *4/4 without issue.

-Super Scrypt Algorithm.  Though not technically a new name from Scrypt.....Yes...it is time.  Scrypt Parameters (4096,2,1)  
GPU is pretty much out of the game for BTC and if LTC's value increases well over the next year or so;  GPU will be out for it as well.  ASICS and FPGA's have and will come into the scene as profitability and market capitalization increases.  Now is a good time to implement the next level of Scrypt.  
*A GPU miner program would need to be altered to allow these parameters.   In a perfect world, one of the community involved would have the miner ready upon coin release.
*This "SHOULD" put the Cache required per thread above what current ASICS and FPGA's utilize;  without eliminating GPU processing as we have now.
*Offers the next level of ASIC and FPGA resistance.
*Would be a slower H/s than LTC given equal GPU's


-Bounty creation.   I do not really consider this a requirement in the least and a good supported coin will not need it.  However, I understand that this is a huge meme and carries some weight....So...
A bounty could possibly be derived from several sources without the need to pre-mine / insta-mine.  
*Dev created pool(s) available at launch can initially use all pool fee's towards a bounty wallet.
*Dev/ support team created lottery, gambling, etc type site can initially divert all revenue towards a bounty wallet.
*Donations could be accepted on forum threads and on website.  
*A donation pool could be created initially, where all proceeds would go towards the bounty wallet.  Miners could then "donate" hash power
*Dev/support team can run some of their own miners temporarily and devote proceeds to a bounty wallet.
*Dev/support team could invest some of their own crypto funds into a bounty wallet.
***Any bounty wallets should be publicly shown / tracked as well as payouts.  

I think overall a coin like this could be supported, grow long-term, be fair and offer the highest exploit and 51% resist (of any coin so far).  These are just some thoughts.  It would be nice to hear what others thought or ideas others may have.  It would seem as though, if we as a community helped come up with ideas (not thinking of profit first);  there indeed could be another really successful coin.

What do you think?  What could be improved?  What could be added?  Would it work?  

Anyway, I guess these are just some thoughts.  Sorry to be long.