Topic: [ANN] Elacoin | Released | Fair Elastic Scrypt Mining | No Premine - page 10. (Read 133887 times)

the community is small enough, if the code is available, maybe you should release now and let people see the changes before they decide to join the band wagon

Indeed we are working on it. We've tested the code a bit today and will release tomorrow likely with appropriate repo's and what not. We'd love feedback on any bugs or whatnot. Stay tuned.

This is a benefit to everyone to revive an otherwise soon-dead coin, so any shenanigans are pointless on our or anyone else's side.

Good to see a few still working on this. I am sure Milkshake is around just under another name

yes no one suggest 25x or 50x reward increase. however, if your purpose is to FUD the coin to death, then keep trying. community concensus here, people. no central authority: remember what BTC is about! Milkshake hasnt been around since 4hrs after launch, so, he's pretty much gone.

The hashrate is low and the blocktimes are in the 2-4hr range, 60-120+x slower than intended right now.

Ultimately, the idea is scaling the reward with difficulty, as per the original release. However the original code didnt start doing this til diff=24 and we never got there. Also, it said "halving every 9 months (194400 blocks)" but the math in the code doenst reflect that. It's * 1/(n+1) where n is # of 270-day periods. (UTSL / RTFM!)

Right now with a diff of 4.35 things are slow. And the reward is 1.00.

Ultimately we've mostly agreed on this:

 - reward = difficulty, smoothed to ELAtoshis. this means if you throw 1MH/s into the coin, you get 1MH/s worth of coins out, always, forever, modulo the time-based scaling mentioned below
 - keep the 1/(n+1) 270-day periods scaling, smoothed out (ongoing smooth reduction, not a 'halving' event like with BTC)

- fixing current situ with hard fork:
  - set difficulty=1.1, giving a 1.0 ELC block now (due to 270d period scaling), ie exactly what the reward is right now
  - min reward based on diff is always 1 (then scaled by 270d-period scalings, keeps miners interested producing a minimal threshold hashrate for the coin)
  - allow a difficulty retarg in 1/4 the normal interval (we're jsut past 8*2160 '3-day' intervals now, so at 8.25 * 2160 = 17820th block retarg
  - another retarg at 8.5 * 2160 = 18360th block to allow proper readdjusting for new miners joining on (or lack of hash power to reduce diff more if slow)
  - one more adjustment retarg at 8.75 * 2160 = 18900th block for more adjusting
  - regular retarg schedule at 9 * 2160 = 19440 and every 2160 blocks ('3 days' if we hit a proper hashrate giving 2 min blocks)
  - keep the 4x and /4 maximum retarg scaling parameters (bet you didnt know about those! yes max 4x increase or decrease)
  - keep the max # of coins as is
  - hard fork around block #17760 (at 17747 now with 2-4hrs/block)

Theory behind this:

 - original idea of ELC was thus. Milkshake said 18 month periods (moore's law) but picked 9. Not sure why. Maybe he's bad at math. Changing it to 18 would be a major disruption to orig. schedule, so leaving it is best
 - many coins right now are just valued by those mining them. there's no global market of trading corn for ELC, or CNC or even BTC's right now. only miners and speculators are valuing them.
 - btc has serious delfation. there's no 'set right amount' of inflation that should counter it, but pegging a coin to the amount of mining power required to produce it seems fair
 - there will be improvements in technology (maybe not for scrypt?) - so we'll keep the 9 month 1/(n+1) reduction, which is LESS aggressive than BTC, LTC, etc, and will be a deflating force on the inflation of more miners/higher diff = more coins per block
 - anyone who says with any certainty "they know whats going to happen" and "the coin will inflate insanely!" or "delfate insanely! do we need another BTC?" is just wrong - you have no idea what will happen, neither do any of us. Right now, immediate and medium term APPEAL seems to be a good mode of addressing coin's creation/management, and pegging it to mining power seems equitable. If it inflates insanely, big deal, we're just as bad as any other coin (check out the trading rates for many new coins, i dont have enough decimal 0's around to price them).

The increasing diff/reward vs the decreasing 9 month 1/(n+1) will balance out the coin (not perfectly, but how could you without pegging it against a real world currency - and how you going to do that, flood the market with buys and sells with your $1million float in the same way real govts manage currencies?). Any more than that no one can say.

EDIT:

We didnt set diff to 1.1, we let it scale down by up to /16 (which is what it will likely be in 3 blocks at the fork). However, reward is 1.00 fixed until block 18360, then reward = diff * scaling on the 270d period applies. 1.1 woulda been a slow block rate unless we got major hash power in which we werent certain of.

At the 0.27 difficulty, the reward of 1.0 will be higher than usual but is designed to re-attract miners to the coin. However, its only 270 coins worth (so dont whine about total devaluation of the coin

I'll tell them that you guys want to see the source first. How much time do you think would be good?

However, as soon as the source goes out, anyone can compile and run it. It's as good as released. Actually, anyone can release any code they like whenever they want as long as they have enough hash power on board(not hard to do with the current state of things >_<)

Changes: I don't think we've settled on anything in particular. we've discussed things, I don't think any of us agree on multiplying the reward 50x or 25x(or anything close). mat5x has smoothed out the 9 month divider so it doesn't happen all at once. TBH the code is not that hard to pump out. It's deciding what to do and then doing it. I've been trying to get the windows QT client to compile(done), but I'm not particularly sure what the others have been up to. Mat5x has been working on the source,  another guy(don't know his name on the forum) has been doing some testing. Really, we need people to mine the coin. So transactions will move. So people will use it Some sort of marketplace would be nice too.

You can come into #elacoin on freenode if you want to chat. That's where we've been doing all of our discussing.

And the actual modified source code for review prior to being implemented?

Any news about hard fork?

The problem is, something has to happen to the coin or its dead seeing how there are other viable options out there.
This is IMO.

Changing the block reward to 25x higher than it was when all ELC mined so far were minted, somehow preserves the value of current coins?  I fail to see the logic, particularly as a holder of a fair quantity of ELC.

Personally, if a proposal is fielded that proposes to make my ELC worth about 4% of the current value, I would tend to consider that "killing the value of current coins."

The goal is to give reward but not too much because the coin launched with what... 1 coin  per block. so if you do the bigger number... it really kills the value of current coins.

By that reasoning, why not make it the following?
int64 nSubsidy = ((250 + (diff / 24)) * COIN) / (1 + (nHeight / 194400)); // 250 ELC per block

Unless anyone can explain why changing the block reward from 1 to 25 is good while changing it to 250 is bad?  How about 2500?  25000, perhaps?  If there is no concern at all for the value of coins mined by participants in the Elacoin community prior to your proposed change, then is the actual magnitude of the change of any concern?

Note - I still feel, as I stated earlier in the thread, that this is a case of miners voting to increase their own rewards at the expense of anyone that mined ELC prior to that point (and that includes me).  And by "voting", it appears that actually means "enforced by 51% attack against anyone that doesn't go along with the plan and switch to this new version of the client" if I read back just a few posts.

gain a little more interest. Plus keep it somewhat intresting for those that stick around after the bots come though and rape the shit out of the coin and move on once the difficulty gets too high

What is the reasoning behind changing the block reward?

really have no idea how but i've noticed that if you cant get a stable connection when you start up the client, it wont sync

why does that slow down block updates on the chain for a client that's behind?

that is due to the low hash rate and very few (if any) nodes.

Probably as soon as we get elacoin-qt to build for windows, which I just managed but without upnp support. I'm having trouble testing it though because my wallet(not the one I compiled) is having trouble syncing...(IOW, my built wallet won't sync but neither will my downloaded wallet, it connects to peers but gets stuck on n blocks)

that's great, when can we expect a release?

right so the hard part has been community building. i dont see you in #elacoin.

and I have collected about 20MH/s worth of peopel who are willing to work on this. and sure, you can fork any coin anytime you wanna throw huge hashes at it. early coins are vuln at low diffs of course but that goes for every single coin out there. I dont see a way to work around this easily.

At least it's activity in the coin. And watching TRC you cant just make the retarg more frequent or people will abuse it as continues on TRC to this day (tho as I say, I dont know if this is abuse or a feature of the coin now)

We're close to being ready, got a testnet going here going to get some testers on it.

uhh, if the dev disappears the community can continue, we dont need him. as for being attacked, this coin has only 2.5 mh/s left, jus pointing one rig at it, i can force a fork. what we need to do is encourage people to download the updated client then start mining, the chain is the same, only updated parameters. your balances are saved if you back up your wallet.dat


anyone with coding skills can change it, it's just up to the community to choose who's changes the will follow.