Why I am Very Disappointed by the Upcoming V8 Removal of the Magnitude Component of the Gridcoin Staking WeightsteemCreated with Sketch.

in #gridcoin7 years ago (edited)

Rewarding hard science with cryptocurrency has always been my main draw to GRC. As a researcher and a scientist, after seeing endless Gigawatt hours wasted hashing for POW coins it was refreshing to see a coin that aims to help humanity with all that power. Mining Gridcoin, I can earn cryptocurrency while solving the unbroken WWII Enigma messages, or mapping the Milky Way, or helping find a cure for Ebola, etc.

For those unfamiliar with how Gridcoin functions, I would like to touch on that briefly. The 'miners' run the software package BOINC, which is a distributed computing application released by the University of California. The BOINC package distributes work from the server of a project of your choice, to your own computer. Your computer does the work, and sends the result back to the server, which rewards you with credits. The Gridcoin network pays you GRC based on the amount of credits you earn.

Currently, you get paid for accumulated research by staking a block on the Gridcoin blockchain, which is done using a weighted proof of stake system called DPOR (Distributed Proof of Research). This is a hybrid of a POR (Proof of Research) and a POS (Proof of Stake) system.

  • A POS system does not let you iterate over a nonce ad-infinitum until you find a valid block (like Bitcoin miners do). Instead, you try and create a block using each of the coins you own and a timer - the more coins you have, the higher the chance of staking a block. This essentially lets anyone earn interest on their holdings by staking a block to gather their accrued annual interest without doing any research. This system lets coin holders secure the network.

  • The POR system rewards 'miners' for their research magnitude, which is the rate at which they contribute to BOINC projects. A high magnitude increases your chance to stake, and earns you a payout proportional to your magnitude when you do stake a block. This system lets researchers secure the network.

The Gridcoin POS Kernel V8 at Next Mandatory (v3.6.0.1)

The next mandatory will eliminate the stake weight afforded to researchers based on their magnitude, which means staking is only done using each wallet's GRC balance using a POS system. You can read the full V8 changelog here. What this will mean for you in terms of your stake rate:

  • If you are an investor, your staking frequency will go up by a factor of just under 2.

  • If you are a solo miner with a low balance, you will never stake and therefore never be paid out for your research. If you do not stake research within six months of completing it, payment is lost. Currently you will need about 165k GRC to stake daily, 24k to stake weekly, and 1k GRC to stake every 6 months in the V8 system.

To put this in perspective, I am currently the biggest Gridcoin miner short of the pools. This change will see my stake frequency go from 15 blocks per day, down to one block every 5 days, on a balance of roughly 40k GRC.

Why Was This Change Introduced?

The removal of the magnitude component of stake weight has been implemented by the developer Tomas Brod in response to several exploits that were uncovered. Most notorious of these was the seizure of the millionth block on the chain by manipulating the implementation of the noobie boost. There was a further exploit identified that lets users amplify the magnitude component of their stake weight using large sets of wallet addresses. Both of these exploits enabled a user to stake more frequently than intended, but did not affect the overall payout that user received.

Regardless of the requirements for security, which are correctly being put first, this seems like huge change to make without community input.

My Concerns With the V8 Changes

Gridcoin has value due to its contributions to research, and the ability to purchase compute with GRC. This value is afforded to the coin by the miners. If we continue to remove power from the miners, Gridcoin gets closer to being another POS coin.

This change pushes miners into pools to ensure that they are able to stake their rewards, which introduces a variety of problems:

  • A user in a pool is unable to control their magnitude. As a result, the magnitude portion of the poll voting weight will be void due to the pool controlling almost all the network mag and so far abstaining from polls.

  • There is only a single active GRC pool. What if that pool ceases to exist? I am aware that there is intention for the pool to become open source, but this is not a simple process and there is no timeline for the open source release.

I think this is important to discuss as a community, and I wanted to make sure the facts are out there and recorded accurately for people to comment on. This is especially important as I get the impression that most community members have no idea this change is being brought in. It is live as of Gridcoin client version 3.6.0.1, and the new V8 staking rules will be activated from block 1010000.

There is discussion in the developer channels of introducing a new payment system for miners, but nothing has been set in stone yet.

What do you think about the changes brought about in V8?


Dutch2.gif

Sort:  

Having witness this debate play out on IRC over the last couple of days and having read the comments here, I started to think about to when I started GRC. It was already a pain in the ass to get the wallet and blockchain loaded. Then I had to send a beacon and wait like 1-2 weeks to get my first payout - and that was after buying a couple thousand coins.

Here's what I see as issues based upon what I'm reading:

  1. As a newbie, I wanted to stake at least once a week. Anything beyond that was scary to me. I think the idea that you could need 10K coins to even start tasking once a month or so, that might be pretty daunting and create a huge barrier to entry.
  2. GRC's purpose is to support BOINC - anything that creates barriers to entry to GRC creates a barrier to entry to new participants in BOINC.

Today is the first alternative I've heard presented that actually seemed to make sense: could we pay out research at each superblock? Could we perhaps have a discussion of what would be the pros and cons of doing this?

One thing, you might get more people to keep their wallets running as a side effect.

@erelas, I think many of us would see that as a very positive side effect.

could we pay out research at each superblock?

Not possible with the current SuperBlock/NeuralNetwork implementation, it would require significant development (we've got higher priority items in the roadmap) and would potentially negate one of the primary reasons for holding GRC (increasing payout frequency) without a replacement economic incentive.

It would also potentially increase the computational complexity of verifying superblocks as each payout would require verification (100's/1000's/10k+ POR payouts in a single block?) compared to only a single CPID requiring verififcation in each block right now.

So, what other options exist? I think a lot of us wold be in interesting in hearing that.

'If you are a solo miner with a low balance, you will never stake and therefore never be paid out for your research. If you do not stake research within six months of completing it, payment is lost.' I read that quite often and it could be misleading, you only lose work done not in the six month rolling window. So lets say you staked in 6 months and 2 days, you would lose the first 2 days of rewards owed. As the first month is mined at reduced Magnitude anyway due to RAC calculation, most miners would barely notice the loss of their first weeks rewards.

I agree we need to make Gridcoin as easy and as approachable as possible, and make mining rewards deliver to people quickly, but security comes first.

I dont agree with the statement that doing this makes Gridcoin closer to a PoS coin, you still get mining rewards in V8, you just dont get them quite so often.

I think we can vote on things, and should do, and I agree the Mandatory wasnt as well publicised as it could have been, but ultimately we are in a mode where someone started publishing outdated security issues with the coin and we dont want any more mud thrown in case it starts to stick. Lets patch this hole and find another secure way to mimic its effects.

EDIT: Maybe we should question the 6 month payout window, I dont know why 6 months is a magic number. Perhaps stake weight can be boosted after 5 months or something, Im not a dev so we have to let them think it over.

you only lose work done not in the six month rolling window

Good point - I probably did not make that as clear as it should have been. The first month may not be mined at reduced RAC though, as that only goes for brand new miners, which will mainly join pools anyway.

The miners that are going to be hit hardest will be the ones that are going at it solo already, but have not got a large balance of GRC.

but you are a community member, and i agree with some bits of your opinion and comment here and want to point out dev or not , you , me CM and everybody else have the same voice... There is no prez there is no peon , your opinion and views along with everybody else as I want you to know and understand are equal dev or not and its a community coin not just the dev's. Yes security matters but when votes for projects are getting NO's over SSL after the vote and the community spoke and since boinc doesn't use ssl it just uses port 443 for plain text after attaching to a project and a whale telling me people are just worried about their credits ( still not ssl its plain text over port 443 per David Anderson on the boinc forums that some people are security paranoid , you must have missed the #gridcoin accused rain theft ( was not anyone stealing ) and missed #gridcoin-irpg and again more theft accusations because people do not understand irc connections are not single user accounts like facebook. So really man , I would love to hear more of your idea. This isn't going to work this way , I had no idea v8 just caters to investor whales and traders or people whom make their GRC off non mining and understand now why/who is supporting it and who is not. Greed vs Science. I did seti@home free on a k6-2 with 4mb stb riva tnt2 4mb with 4mb edo sodimm addon so i could watch the pretty openGL 3.5mhz frequency spectrum water fall screen saver via my 2 x sli voodoo 2's. 1wu every 72hr to a week for free. The 6mo window thought really sparks my interest if we could look into that , but then again what I want to know is what pepole getting as incentives or kickbacks to force new users to grcpool.com and this again as I keep calling it and now its a monopoly on new users being forced to join the pool in order to ( earn from completed wu's as we dont mine gridcoin ) but if extending past 6mo could change new users being forced to join the pool lets hear more , I am starting to think stake gets used loosely meaning " create a block that you get PoR from " where it was previously staking a block for interest , sorry @ 20k GRC i still get more from rain than I do interest so staking interest is not important to me , along with many other users, it seems to only matter to whales.

Loading...

If you are a solo miner with a low balance, you will never stake and therefore never be paid out for your research. If you do not stake research within six months of completing it, payment is lost. Currently you will need about 165k GRC to stake daily, 24k to stake weekly, and 1k GRC to stake every 6 months in the V8 system."

This statement remains undemonstrated and unsubstantiated. When v8 goes live, the difficulty of staking a POS block will decrease for those with medium balances. Currently we do not recommend attempting to go solo with a low balance, so I do not see how this is any different than the current situation.

Look up my username on gridcoinstats about 6 months ago, you'll see how this can affect. At my highest point of holding I only ever had around 10k in my balance, and to me as a casual researcher, that was a lot, I have less now, my own choice, sure, but even at 10k balance I was staking 150 GRC at a shot. That simply is not happening now, and will not be happening in the future as things stand and I understood @dutch's explanation correctly.
I don't consider myself a casual researcher now, as I routinely remain in the top 200's in some projects and have a higher "rank" in them thatn some names I recognize from here, yet, I never stake, or stake at terms longer than a week. V8 isn't active yet, it's gonna be a rough change to see.

I have less now, my own choice, sure

And that is your choice. By holding less coins, you stake less frequently. Welcome to Proof of Stake coins.

sighs That could have probably been stated better.

From scalextrix - "I dont agree with the statement that doing this makes Gridcoin closer to a PoS coin, you still get mining rewards in V8, you just dont get them quite so often."

A few comments down from that, from barton26 - "Welcome to Proof of Stake coins."

These little snips show how things devolve, imo. GRC has always been both PoW and PoS (PoR), and my question posted on the forum is still out there.

barton26, not making a big deal, but also being entirely honest, that came across a little snide sounding in my head. Meh, I take for granted that may be my issue, but if you don't know I took it that way, and it wasn't your intent and I let you know, then no harm no foul, I hope.

I still haven't seen much to dissuade me from taking @dutch's analysis at face value, perhaps I could get someone to actually define small miner, medium miner, etc., since these terms are being used to describe hypotheticals. I'm sure we're closer in numbers than this, but, it could be that I think a small miner is someone holding 1k balance or less and you consider it to be 100k or less. Let me know what it's to be considered and we can go from there.

Of course, holding less coins means staking fewer times - I'm currently involved in another coin that I stake eight to ten times a day, it's been good and I have significantly, as in years less, time in it than GRC. So I feel a baseline for defining these terms is good for conversation and information.

If I could, I'd put it up for a vote, before V8 went live, whether to leave it to activate, or to alter it asap so it activates but does so in a way that doesn't hinder new-solo and older/smaller solo miners.

The entire mechanism of Gridcoin was supposed to be to encourage research, I feel V8 moves us farther away from that and it should be altered before being allowed to activate.

If I could, I'd put it up for a vote, before V8 went live, whether to leave it to activate, or to alter it asap so it activates but does so in a way that doesn't hinder new-solo and older/smaller solo miners.

The v8 changes were required due to critical network security risks associated with providing magnitude a stake weight. Once v8 is implemented and stable we can certainly prioritize the topic of POR payout prioritization on github/steemit/etc.

You should check out the github wiki's proposed incompatible changes, there are many proposed improvements which will directly benefit POR users.

The entire mechanism of Gridcoin was supposed to be to encourage research, I feel V8 moves us farther away from that and it should be altered before being allowed to activate.

Given the network security risk of sticking with the current staking mechanism (which includes mag in the staking mechanism), maintaining the network/blockchain on pre-mandatory versions (forked blockchain) will likely experience malicious network consensus interference.

Once the mandatory has passed, we can transparently discuss improvements for the POR mechanisms as a community & assign an appropriate priority (and bounty via dev program) for development (considering the other high priority development items in the pipeline).

I'm very interested in the we/community/... discussions. Where should I look to know what the communities priority list is. Point to issues lists (who decides on priorities?) and saying things will change is pointless. You're saying it will be worked on. But I have no idea if it will be first second or never. It's the opposite of fud.

Look up my username on gridcoinstats about 6 months ago
[ ... ]
at 10k balance I was staking 150 GRC at a shot. That simply is not happening now, and will not be happening in the future as things stand and I understood @dutch's explanation correctly.

Reductions in POR rewards are likely due to the substantial growth in team Gridcoin this year; thousands of more users have joined through pools or solo so the difficulty and competition to earn large magnitudes has increased. You need to competitively crunch, picking the projects with most compatibility with your machines and the least user participation/competition (crunch less popular projects).

The changes won't affect your payment amounts, rather the frequency at which you are paid will potentially decrease.

Be prepared, I honestly believe that you are about to see the first "I ain't been paid in 6 months", complaint, and the time to address that, imo, is before the complaint has a chance.

I have spent a, my wife wants to kill me, vast amount of time since...a certain incident, poring over Team Gridcoins members stats, looking at what they do, how they do it, etc. I am unconvinced that we can say it's due to a large influx of new users, sure, that's a factor, but not the entirety.

Please, take a moment to look at my recent stats, then ask me if I think I am competitively crunching, lol.

It costs more to stake frequently now than it would with V8. It is much cheaper to purchase GRC than it is to purchase the hardware required for a decent mag.

I agree. If your sole aim is to increase stake weight, that will become cheaper in V8.

I am mainly taking the angle of the miners that have existing hardware though.

Does that seem to be the consensus now? I'm having issues keeping up with disparate opinions in multiple places, so I am genuinely asking if that is being born out?

He is right. If your sole aim is to increase stake weight (starting from nothing - no hardware or GRC), that will become cheaper in V8.

I am mainly taking the angle of the miners that have existing hardware though. This is what I feel GRC and BOINC target - people who have idle hardware lying around already. They get hit by this change.

Yup, that is what I though, oh and color me affected. I mean for goodness sake, have you checked out the prices / availability of gpu's lately. If you count power, hardware, etc., no reasonable person is buying "turn your spare cpu cycles into SCIENCE!" That was great two years ago, when I didn't junk old hardware, GRC was that hardwares last hope of surviving.

These calculations can't possibly be correct. If you check my stats when I switched to PoS I staked on average 3.75 times per day with ~46k GRC. If we check the 4 days after and the 4 days before we get:

  • 4 most recent days as researcher, average mag 115: 2.5 stakes/day
  • Big jump to 16/7
  • 4 days as researcher, average mag 150: 5.25 stakes/day
  • 4 days as investor, mag 0: 3.75 stakes/day
  • 4 days as researcher, average mag 155: 3.75 stakes/day

I will switch to investor during the weekend and see how often I stake.

I based all those numbers off the wallet's staking estimates, which correlated nicely with Vortac's approximate stake rate (and he is primarily an investor in terms of GRC/mag ratio).

Based on your numbers, you essentially got no boost from your mag to help you stake in the current stake engine...?

EDIT: I just realised you are effectively an investor too. Most miners are not sitting on ~50k GRC. You will not be affected heavily by V8 as you are probably near the sweet spot where increased balance stake rate and removal of mag stake weight cancel.

Writing it here as well: my 115 mag and 34 UTXOs has the same stake weight as 4000 GRC. I would say this mostly affects users with a lot of UTXOs.

dutch, I really appreciate the way you explain things, no offense but it seems to work along better than others explanations with my brain, lol. You should hear how I have to turn things around for my wife.

I've never really looked into the big fish of GRC, looked at their stats etc., because what would be the point, GRC is supposed to be altruistic, and now I have. I'm about to pass RHalford in one project, but he's made no movement there in a long time, merely "invested" and sat in that position. I really should have given that a think when I started, but, I was more excited to do research for research sake, and others are playing a different game. Ok, well, game on.

GRC is supposed to be altruistic

That may be the perception by some, but this has never been an official stance within the Gridcoin network. For-Profit and non-scientific projects are likely going to be a reality in the future.

he's made no movement there in a long time

Similar to the vast majority of the BOINC network, they have become inactive in crunching due to real life situational changes.

I was more excited to do research for research sake, and others are playing a different game.

You can do that without gridcoin, the introduction of gridcoin in the picture means you're playing the same game.

Me : GRC is supposed to be altruistic

cm-steem :That may be the perception by some, but this has never been an official stance within the Gridcoin network. For-Profit and non-scientific projects are likely going to be a reality in the future.

I have typed my response to this, like four times now, thank God for backspace, because I feel like I am forced to pretend that I did not hear that.

Security in this world (unfortunately) has to always be the top concern. I don't love the concept of losing out on my magnitude stake weight, but if that's the price that must be paid to avoid security FUD dragging Gridcoin down, that's acceptable to me.

I think there are some really good ideas being tossed around that I believe we should be strongly considering. I've heard that one option is to basically pay everybody their mining rewards at the SuperBlock. IMO, this is a crucial change if magnitude is no longer part of stake weight. Think about it: if there's no relationship between magnitude and stake weight, then why should mining rewards only be paid when someone stakes a block? Makes no sense. Instead, just have the neural network track everyone's owed mining rewards and if you're owed at least 100GRC, you get a payment at the next SuperBlock. Basically along with a SuperBlock being minted would be hundreds of mining payments. It would be doable, especially if you limited the minimum payment to 100 GRC which means very few miners would be getting paid daily but most would be paid at least once every couple weeks.

Bump the threshold up to 500GRC if you want to really minimize those payments and keep an incentive for people to join pools if they still want daily payouts.

You could as well pay out after a certain amount of time instead of coins. More predictable and fair for everyone.

I would probably vote yes on the ideas in your second paragraph, and even mostly agree with bumping it to 500. Hell, I've got pools now that pay out only at much more (they are also more profitable) as a minimum. It should though be voted on I think after the proposal was hashed out.

IMO I don't think we should have a threshold for payout. I would prefer something like STEEM where you can click claim rewards.

I can only agree ! And once in 6 months staking is nevertheless a too big timeframe for getting rewarded. We definitely should look for something that is working quicker than that. The pool topic should really be emphasized nevertheless grcpool is doing a great job it is currently a probable single point of failure and out of risk purposes this is a nightmare in itself. Please all an everyone who's got an idea to that come join the gridcoin slack channel and discuss with us. We already got some things sketched but like @dutch says there's nothing written in stone yet...
Thanks for the post !

The pool is open source, it's only a matter of time before many pools are available (especially when the marketcap increases).

The current solution to improving staking frequency is to acquire more GRC (they're still undervalued at the moment, imo).

While the pool is 'open source' it is VERY hard to get running and there is a lot of work left to do until someone can set it up without help from Bgb. Darthur went through the process and it took him days and a lot of Bgb's input, even though he has decades of experience under his belt with programming.

Projects usually last longer than a few days to reach a production state, especially experimental/beta software. As more people work on the pool, it'll get better & easier to run a new one.

Fair point, maybe expecting near plug-and-play functionality is too high a goal.

For the short term, yes it's too high a goal. Anyone interested in running a pool needs to get their hands dirty & not expect a production level pool kit.

I am sorry if this sounds....bad tempered, please don't take it that way. I stated in the forum that you should shut off crunching in the short term, mine something profitable, buy GRC to achieve a higher balance and then continue on. I was, debated on that statement. Now I see that statement is, effectively being repeated and this back and forth is making me nutso.

It is very easy to say to someone, to make more "money" you have to spend more "money", we've all heard that and some of us can not, will not, don't want to hear how easy it is to fix with money, the problem.

I don't see how your post relates to the programming/running of a gridcoin pool? Did you reply to the wrong comment?

cm-steem : The current solution to improving staking frequency is to acquire more GRC (they're still undervalued at the moment, imo).

Worded differently, that means, "They're cheap atm, buy now". That involves throwing money at the problem to fix the issue on an individual basis, not fixing the problem itself persay.

I think what we are seeing is the sometimes odd ordering of comments, they are in one place at posting, someone votes on it and it can be moved in position, sort of placing it out of a topically logical order.

How does this affect solo vs group mining? Why would one solo mine vs group mine and what is the switching point?

Solo investors will stake more frequently, the pool will likely stake less frequently however they no doubt have enough grc in reserves to stake with in a timely fashion.

Switching from pool to solo -> the point is to help secure the network yourself, getting voting rights, maximum decentralization.

If you're in it to mine and earn coins none of that really matters. the voting power is minimal compared to the big guy. Users won't mine if they don't think it's secure. Decentralization follows the same reasoning. What's the reward for helping securing/decentralization with V8?
Edit: is that the 1.5 % apr?

Group mining gets you paid more frequently in V8, assuming the pool will hold a large reserve balance to stake. Solo mining will see a vast reduction in payment frequency.

I guess more users care about frequent payment instead of 1.5% apr. At least I do with a coin I find harder to trust due to several factors. That way I can move earnings somewhere I trust more. End result: even less secure coin.

At least I do with a coin I find harder to trust due to several factors.

Cryptocurrencies are trustless by design.

End result: even less secure coin.

I don't believe that if you use a pool or sell your grc that contributes to a less secure gridcoin network.

Every currency only works because of trust. I don't think my personal preference makes a difference. But if everybody turns to the pool it would.

Every currency only works because of trust.

Perhaps FIAT currencies where you trust a bank to exchange cash for the real monetary value of the notes, but in crypto trust does not enter the equation.

Sure it does.
What will happen if people do not trust Bitcoin anymore? Well they won't buy it anymore en people owning will be selling. No demand and plenty supply means the price will drop en the coin will be worthless.
The only difference between fiat currency and crypto currency is that fiat currency is centralized government controlled whereas crypto is decentralized en controlled by the community. They are both based on trust.

That's nonsense. Why are shops using bitcoin and not some minor altcoin? Even when looking at it as an asset you're wrong. Trust is one of the major aspects.

Cryptocurrency is not trustless. Blockchain is trustless. Cryptocurrency working depends on lots of trust, from the dev(s) down.

Thanks a lot for the great summary. Looks like I will have to stay in the pool for a lot longer, in order to have frequent rewards.
Don't get me wrong, I love the pool and all the good it gives to us newbies, but I was planning to be a mature miner at some point.

Coin Marketplace

STEEM 0.28
TRX 0.14
JST 0.036
BTC 61568.14
ETH 3403.94
USDT 1.00
SBD 4.51