Author Topic: Pre Mainnet RTA/MVP launch Community discussion!! June 2019  (Read 490 times)

yidakee

  • Administrator
  • Newbie
  • *****
  • Posts: 37
  • Karma: +0/-0
Changelog:
13-06-19 - Added Proposal 4
17-06-19 - Removed proposal #3 - it is unfeasible for the next fork

_____
Dear Graft Community!

 It is with great pleasure that I am opening a round of discussion about the future of our project. RTA main net launch, or better, the Graft MVP (Minimum Viable Product) suit is close to being released. End of June / early July is the tentative date for final round of testnet, and upon pending bug squashing and tweaks - a hard-fork will happen to release it to main net.

 Now that the core team's vision of the Graft ecosystem is on the verge of materialising, it also now time for the Community to better self-organise and get more involved with the direction of the project's future. And for the non developers/coders to contribute outside of Github.

 It is my firm belief that neither Telegram nor Discord not are appropriate hubs to build a fruitful discussion, vote on issues, network and self-organize. Here, discussions are logged, can be segmented and reviewed. Post content is more thoughtful and meaningful. Polls work way better, and more democratic. Timezones are less relevant.

 I've received countless private message with numerous ideas and proposals on how to help the project thrive. We still have lots on the table, like go-to-market strategies, rebranding, redefining open-source models, governance, etc … all topics that need to be addressed for sure!

 But for this round, and after talking internally, it feels like we should take a first step to better self-organize and involve the Community - using this hard-fork as opportunity to take the first step in a new direction.

 As such, I would like to propose a healthy discussion on what we would like see changed. What needs to be changed, what shouldn’t be changed - and why. Then, create a poll and vote!

 Having said this, this is not a free card to change the core architecture. For example, many have proposed Supernodes get a cut from the mining emission, like classic masternode coins. That simply cannot happen at this stage, as it would require going back to the drawing board and deep recoding. Or eliminating the Tiers for a unified Tier structure - this clashes too with future service-types in the works.

But other things can be altered, as per the suggestions received.

 I'll kick off with some ideas thrown around in the chatrooms, here on the forum, and in private debates.

 If this shows to be a positive and thoughtful debate, we will select the best proposals and move onto a vote. This is your chance to make a difference!

Here are the a few plausible suggestions I received a few times from different community members.

 Proposal 1 - Reduce the emission curve again. We've done it before and it showed positive signs, but we got hit by crypto winter. Then FPGA's flooded the market, which forced us to create a custom PoW algo, which again, showed signs of improvement. Then Cryptopia got hacked and it now seems major sell pressure came from their liquidation. It seems (hope not speaking too soon) this terrible sell pressure is alleviating. By reducing the emission, just like BTC halving, miners would need to mine more for the equivalent cost of electrical power. Fundamentally, the base point of GRFT’s price equilibrium would rise. How do you feel about a another mining emission cut, and how much?

 Proposal 2 - Change Supernode Stake - Given that price crashed so bad, and ASIC and FPGA flooded the circulating supply, some community members have proposed a revision of the Tier requirement. By increasing the threshold for each Tier, current Supernode Ops would not be affected as the RTA stimulus can be adjusted to meet the projected return. However, there will be a higher barrier of entry for new SN Ops, which should promote buying pressure. Contrarily, node count will keep increasing too easily, driving ROI down consistently, to a point where there is no incentive to run a Supernode. Not to forget that delegated staking is planned, significantly lowering the requirements for participating in a Supernode and keeping things fair. The community member proposed a revision of the total current locked Graft into stakes, and recalculate the collateral for each to meet current node distribution to a total of 500 nodes.

 Proposal 3 -  by Serg on TG - Add two extra Tiers - T5 500k  and T6 1M - not yet formally proposed with logic behind it, waiting for him to post here. Placeholder ATM - Unfortunately, this proposal can't make it for the next fork.

 Proposal 4 -  by Raybl - Post#7 - Change the PoW algo to CN-GPU (or analogous) where to deter the FPGA miner owning 75% of Graft hashrate, thus bringing back more GPU miners and promoting better decentralisation.

Hoping to hear your thoughts and maybe more ideas!

Please feel free to propose, if it does not clash with architecture and we can fit it into the next fork, I will add the proposal to this post.

Discuss!
« Last Edit: June 17, 2019, 04:57:26 PM by yidakee »

imashowman

  • Newbie
  • *
  • Posts: 2
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #1 on: June 12, 2019, 07:07:17 PM »
Hello all!

I think proposal 1 is better than 2 for the permaholders like me. Im not seeking to buy more GRAFT, i already buy a lot at ICO prices.

For me reducing the emission and ban FPGA will increase GRAFT price, so i agree with proposal 1. GRAFT inflation rate is too high and a high inflation never is a good thing for a coin

Ricor

  • Newbie
  • *
  • Posts: 2
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #2 on: June 12, 2019, 07:17:39 PM »
Concerning proposal 1 I'd rather see that the emission in part would be transferred from mining block rewards to SN rewards. This, when combined with increased SN tier requirements (proposal 2), would likely decrease the mine-and-dump effect on the price as the SN operators have a tendency to reinvest rather than sell their Graft (based on the steady increase in staked Graft and number of SNs). This adjustment would also have less of an impact long-term compared to halving the emission.

bigdan

  • Newbie
  • *
  • Posts: 2
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #3 on: June 12, 2019, 08:39:36 PM »
Hi everyone,  Proposal 1 makes a lot of sense to the investors who are hoping to cash in on graft should its price appreciate to their satisfaction but that's where i have a problem because it leaves the future of graft in jeopardy if we all/many of us are hoping to sell/dump our graft coins at some point for fiat I'd rather admonish the community to support it to become the payment processing system and currency we are all hoping for.

yidakee

  • Administrator
  • Newbie
  • *****
  • Posts: 37
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #4 on: June 12, 2019, 08:57:36 PM »
 Just to be clear it can either be

 Proposal 1 only
 Proposal 2 only
 Both Proposals
 Neither Proposal

 Other Proposals, mix and match

 Important is to get a sense of people sentiments and ideas … and to vote on them. The Community decides!
« Last Edit: June 12, 2019, 08:59:55 PM by yidakee »

JackJones

  • Newbie
  • *
  • Posts: 1
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #5 on: June 12, 2019, 09:28:03 PM »
Proposal 1 preferred BUT need to change algo again maybe moneros Cryptonight R with tweak. Proposal 2 only works short term. If graft succeeds we need 2000+ nodes to provide scalable service. So if we increase cost of ownership we reduce number of nodes available, so eventually we'll need to revert this change. In summary go with Proposal 1.

I like Ricors suggestion of sharing rewards between mining and supernodes. This would increase the value of supernodes and increase demand

yidakee

  • Administrator
  • Newbie
  • *****
  • Posts: 37
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #6 on: June 12, 2019, 10:12:14 PM »

 Please read the OP carefully. Sharing mining rewards with Supernodes at this moment in time off the table. It would require a deep recoding of the code, going back to the drawing board and re-do the entire architecture. I may be a possibility in the future, but that option right now is off the table. Supernodes were designed to receive rewards based on RL sales at the merchant POS.

Raybl

  • Newbie
  • *
  • Posts: 1
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #7 on: June 13, 2019, 01:35:36 PM »
Lowering the emission rate might have a short term effect on the price which is absolutely not guaranteed. The main effect will be a cut in the networks hashrate. Furthermore FPGAs will be incentivised, letting them overtake the network even more and the dumping continious. About 75% of the network is in the hand of one entity. That needs attention. Of course market needs to absorb less to maintain the price with a lower emission. Long term there is no real effect as the totall emission wont change. So reducing the emission rate today will increase the emission rate later. A long term effect would be a reduction of the max supply or a burning rate.
My proposal is implement checkpointing in the long run, let the SNs secure the network and distribute a high percentage to the SNs and the rest to the miners. I don't see any contradiction in letting the nodes earn from fees and block rewards. That would also incentivice holders.
For short term, change the algo to a fpga unfriendly one like cn gpu variant. 
I am not a fan of the tier structure at all. I don't see any benefit in it as shared nodes will likely lead to a balanced distribution between tiers resulting in equal profitablility between tiers. I also don't see beneficial effects in adding more tiers or changing the requirements that is buffered by the multi SN feature. The requirements for SNs is also related to the number of SNs required to serve the network. This should be taken into consideration if doing any changes. Skipping the tier structure at all and implement shared nodes is what I would prefer.
A resolve of the staking bug and infinite staking would be nice for the next fork.

Only a working product will bring real value!

yidakee

  • Administrator
  • Newbie
  • *****
  • Posts: 37
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #8 on: June 13, 2019, 02:30:01 PM »
Lowering the emission rate might have a short term effect on the price which is absolutely not guaranteed. The main effect will be a cut in the networks hashrate. Furthermore FPGAs will be incentivised, letting them overtake the network even more and the dumping continious. About 75% of the network is in the hand of one entity. That needs attention. Of course market needs to absorb less to maintain the price with a lower emission. Long term there is no real effect as the totall emission wont change. So reducing the emission rate today will increase the emission rate later. A long term effect would be a reduction of the max supply or a burning rate.

My proposal is implement checkpointing in the long run, let the SNs secure the network and distribute a high percentage to the SNs and the rest to the miners. I don't see any contradiction in letting the nodes earn from fees and block rewards. That would also incentivice holders.
For short term, change the algo to a fpga unfriendly one like cn gpu variant. 

I am not a fan of the tier structure at all. I don't see any benefit in it as shared nodes will likely lead to a balanced distribution between tiers resulting in equal profitablility between tiers. I also don't see beneficial effects in adding more tiers or changing the requirements that is buffered by the multi SN feature. The requirements for SNs is also related to the number of SNs required to serve the network. This should be taken into consideration if doing any changes.

Skipping the tier structure at all and implement shared nodes is what I would prefer.
A resolve of the staking bug and infinite staking would be nice for the next fork.

Only a working product will bring real value!

 Very good points Raybl - thanks for that. yeah, last time we had this mining discussion the community was in favor of CN-GPU indeed, or equivalent algo that closes the gap between FPGA and GPU's.

 I will add this as proposal as it is a viable option for the fork. Regrettably, SN rewards direct from the minting process is not an option at this point, hopefully this door may open in the future.

 Regarding the tier structure - I also agree the 4 tier system is marginally functioning as envisioned. Not necessarily because of the multi-SN support, as soon as RTA in enabled this setup will surely suffer a tremendous hit. If you add delegated staking to the mix, the 4 Tier system will quickly reach equal ROI, providing no extra incentive for a T4 vs T1 (except VPS cost). I'm just not sure if such a big change is plausible in such a short time, or if the team would veto this proposal. We dont have a true governance model yet for such big decisions, that may clash with the vision and high-level architecture, but maybe Slava or Dan could shed a light here.
« Last Edit: June 13, 2019, 02:41:01 PM by yidakee »

JChemist

  • Newbie
  • *
  • Posts: 4
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #9 on: June 14, 2019, 07:12:08 AM »
 Proposal 1 - Reduce the emission curve again. We've done it before and it showed positive signs, but we got hit by crypto winter. Then FPGA's flooded the market, which forced us to create a custom PoW algo, which again, showed signs of improvement. Then Cryptopia got hacked and it now seems major sell pressure came from their liquidation. It seems (hope not speaking too soon) this terrible sell pressure is alleviating. By reducing the emission, just like BTC halving, miners would need to mine more for the equivalent cost of electrical power. Fundamentally, the base point of GRFT’s price equilibrium would rise. How do you feel about a another mining emission cut, and how much?

Primary goal should be greater number of miners and mining pools.  Getting rid of FPGAs would support this goal by making it more economical for gpu miners.    That said, there are too many GRFT floating around.  This can be solved many ways  but if there is no proper solution, you gotta bite the bullet and reduce emissions so that current holder value is not diluted so much by emission.   A better solution is create more demand but we have a rough patch with all the competition and uncertainty it creates.   Obviously, if GRFT had one big partner or one big investor, it might be enough to change the situation w/o emssion cuts

 Proposal 2 - Given that price crashed so bad, and ASIC and FPGA flooded the circulating supply, some community members have proposed a revision of the Tier requirement. By increasing the threshold for each Tier, current Supernode Ops would not be affected as the RTA stimulus can be adjusted to meet the projected return. However, there will be a higher barrier of entry for new SN Ops, which should promote buying pressure. Contrarily, node count will keep increasing too easily, driving ROI down consistently, to a point where there is no incentive to run a Supernode. Not to forget that delegated staking is planned, significantly lowering the requirements for participating in a Supernode and keeping things fair. The community member proposed a revision of the total current locked Graft into stakes, and recalculate the collateral for each to meet current node distribution to a total of 500 nodes.

Not necessary with muti-sig nodes, the return is achieved by buying more nodes.    I do not like the current situation of aggregation of nodes and that of delegated staking.   Need to ensure true distribution of nodes among many holders.  I believe the number of SN holders is much less than the number of nodes.   I think you need to anticipate further delays as safety margin and how do you maintain network when return is less than cost.   At the very least you need to maintain value by less dilution.   If at all possible, create SN mining.  I think there is some type of loss of some privacy issue but that can be an upfront tradeoff only for those that stake for mining.    I suppose this is a huge programming cost and should not take away from current roadmap.


 Proposal 3 -  by Serg on TG - Add two extra Tiers - T5 500k  and T6 1M - not yet formally proposed with logic behind it, waiting for him to post here. Placeholder ATM

Get a partner and up the programming pace.   GRFT may have a good idea and good team but faster progress.  I saw note about SAFE Note on Telegram.   I have no idea if mergers is possible but other teams are going to face reality of more competition, more copy cats, and limited funds.     


 Proposal 4 -  by Raybl - Post#7 - Change the PoW algo to CN-GPU (or analogous) where to deter the FPGA miner owning 75% of Graft hashrate, thus bringing back more GPU miners and promoting better decentralisation. 

I would rather you just cut the reward.     If anything create RTA mining (w block validation) to offset the FPGA.   Efforts to circumvent ASICs, then FPGA is too much.  Just change the game instead of trying to outsmart the miners.

RoggyDog

  • Newbie
  • *
  • Posts: 1
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #10 on: June 14, 2019, 02:38:35 PM »
I like the tier structure as it is. What I would like to see is some sort of mechanism to limit the number of nodes so when Graft becomes popular it doesn't get flooded with big money players, making profits for early adopters vanish. As the price/popularity of Graft increases I think the value of higher tier nodes will increase because it will become increasingly more expensive to have a T3 or T4 node. Something along the lines of...
1) Enable infinite staking with ability to remove your stake with maybe a delay to enable network to find another node to replace yours.

2) The network only allows for a certain number of nodes based on transaction volume with some percent above that to account for growth/holiday shopping. Like maybe 30-50% above what current transaction volume is. This number of nodes could be calculated possibly weekly going up or down as network needs it. This would maintain consistent profits for supernode operators. If transaction volume decreases the number of nodes would only decrease if someone unstakes their node and the network wouldn't replace them.

3) Create mechanism for future wannabe node operators to lock funds and have a running SN that is only queried by network maybe once per day to verify wannabe node is online and current (but this wannabe node doesn't do RTA transactions, it just sits there updating it's copy of the blockchain while waiting to be activated by the network, maybe mining while it waits). The network uses this info to select from a pool of wannabe nodes whenever the network needs to increase node quantity. Selection would take into consideration earliest funds lock, most consistent uptime, and running up to date version. This way early adopters, long-term holders get preference over someone with lots of cash just jumping in and flooding the network with nodes.

No idea how feasible those ideas are, I just don't want to see my commitment to Graft become worthless when it becomes popular and greedy traders come flooding in to buy up a ton of nodes and reduce their value to almost nothing. I can only afford a few nodes, unlike others out there with big money that have pretty much ruined most of the worlds societies with their greed and get rich at any cost without doing any of the actual work.


JChemist

  • Newbie
  • *
  • Posts: 4
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #11 on: June 14, 2019, 11:43:52 PM »
Limiting nodes to those active is a great idea.  This is incentive to keep the investment.

My only small concern is that there should be a quality standard for VPS in the future.   Current situation is OK but those pushing the limits of low end VPS for actual network seems not good.   At that time, GRFT should be worth some $ and upgrading VPS should not be an issue.

Ricor

  • Newbie
  • *
  • Posts: 2
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #12 on: June 15, 2019, 05:59:29 AM »
Proposal 4 is highly important and will soon become urgent to address. Currently there are two miners (likely FPGAs and likely operated by the same person) on the Graft community donation pool with 2.46 MH/s and 966 kH/s hashrate repsectively. The Graft network has a total hashrate of 6.5 MH/s. As FPGAs is hardware that isn't publically available (in contrast to CPUs and GPUs), the bitstreams aren't open source (unlike e.g. xmr-stak) and a few entities dominate the network they currently pose an enormous threat to decentralization - and is likely connected to the price dump as history has shown that small miners are more prone to hold their earnings compared to large farms. Shifting to CN-GPU or CN-R would remedy these issues.

yidakee

  • Administrator
  • Newbie
  • *****
  • Posts: 37
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #13 on: June 15, 2019, 09:14:57 AM »
 I cannot stress this enough - despite many valid ideas presented, please keep in mind that this discussion is for feasible changes that can be made in time for the next fork.

 Radical code changes are simply impossible. PoW change is a possibility, so is changing other parameters like staking requirement, emission or others ..

 Please keep it realistic for this discussion to be fruitful - 100% of engineering time is dedicated to launch our MVP come next fork, not the change the code for new features

JChemist

  • Newbie
  • *
  • Posts: 4
  • Karma: +0/-0
Re: Pre Mainnet RTA/MVP launch Community discussion!! June 2019
« Reply #14 on: June 16, 2019, 10:18:46 AM »
Re practical    Then I vote for block reward reduction.   The reward should be dropped by 75% and Increase every block until back to 500 and then start drop again.  Something that parallels planned benchmark and demand.

w regard to algo change.  its a game.   The only thing that beats FPGA is ASIC and ASIC have dropped in price such that instead of GPU, people can invest buy a cheap second hand ASIC.   SHA256