Create 9 - Change mining algorithm to RandomX

[doctor-gonzo]

A mining algo change would go a long way towards re-vitalizing the mining community. Many people built rigs or pointed rigs at QRL initially but have been crushed by ASICs / FPGAs and no longer participate.

[surg0r]

Since mainnet launch (approximately 1 year 8 months ago) mining has become concentrated in the hands of a small number of miners with exotic hardware and the network node count has gradually dwindled.

A switch to randomX should (initially) promote more distributed mining and encourage community members to run a mining node.

As such I personally support this QIP.

[cyyber]

It is very clear by looking to the nicehash Cryptonightv7 rent cost, that some kind of unknown device exists that can mine Cryptonightv7 at much lower cost, which is not known to public. Due to this low cost mining, miners are selling the QRL at any price above their cost, which could be the major reason why QRL price fell at this level. Thus change of the mining algorithm in the upcoming hardfork is not just necessary but very much critical.

RandomX algorithm suites very much to this purpose as its more tilted towards CPU rather than GPU. A CPU with higher memory can outnumber the hash rate compared to GPU. Moreover, it will not be easy for anyone to run botnet mining without being noticed, as this algorithm itself consumes heavy amount of CPU Memory.

I totally support this QIP.

[Ottslayer]

Curefrankosflue from the discord qrl chat was early on suspecting the effect of ASICS on the qrl environment, and I am totally on board with his reasoning.

I support this QIP

[Doneyy]

Sounds like a good plan. Better sooner than later. I support this QIP

[atoma01]

This change seems to be better than the last one https://github.com/theQRL/theqrl.org/blob/master/_qips/migrate_to_cryptonote_v8.md. As Cyber said " its more tilted towards CPU rather than GPU." Also like the fact its been audited. I support this QIP. However keep in mind that a loss of total hashrate can be interpeted by the market to be negative i regards to imutabillity of the Chain. So i think its important to motivate the community to support the price, run full nodes, mine and hold Quanta.

[IMac318]

I support this QIP. If QRL was previously following Monero, the mining algo would need to be changed every 6 months with questionable benefit. With RandomX, there is supposed to be 3+ years until specialized hardware catches up, which would hopefully give more than enough time for POS to be developed and implemented.

[vasanthr880]

I support this QIP, as I understand it helps holders to mine with our modest GPU/ CPU

[StrikeAttack]

Great reasons have already been left. I won't create any more noise. After a review of the RandomX protocol, and PoC building a miner, I support this QIP.

[mg16spider]

Switching to random X protocol could make QRL mining more competitive. As mentioned by others, I also believe the change can have a positive impact on QRL market cap. The heart of a successful coin is the mining community. Creating a fair and competitive space for miners can create exposure and growth. I support this QIP.

[cooper7777]

I support this QIP. But i think, that the change of algo should be done before the upcoming harfork, that planned on Jan or Feb. We don't have so much time to allow FPGAs to dominate.

[mastercyb]

I support an idea of urgently changing mining algorithm and I like RandomX. But if QRL adopts it attack surface appears. Huge Monero hashrate can be used by pool operators for double spending. What do you think guys on that? What about adopting more GPU oriented but unique ProgPow which has been created, but not being used yet?

[surg0r]

I support an idea of urgently changing mining algorithm and I like RandomX. But if QRL adopts it attack surface appears. Huge Monero hashrate can be used by pool operators for double spending. What do you think guys on that? What about adopting more GPU oriented but unique ProgPow which has been created, but not being used yet?

A 51% attack may be mitigated by setting a 'reorg limit' below what exchanges track for deposit confirmation. Reducing the limit is a planned part of the next core client hard fork upgrade.

[mastercyb]

I support an idea of urgently changing mining algorithm and I like RandomX. But if QRL adopts it attack surface appears. Huge Monero hashrate can be used by pool operators for double spending. What do you think guys on that? What about adopting more GPU oriented but unique ProgPow which has been created, but not being used yet?

A 51% attack may be mitigated by setting a 'reorg limit' below what exchanges track for deposit confirmation. Reducing the limit is a planned part of the next core client hard fork upgrade.

That does not add security for the chain

[robypsi]

I support this qip, particularly in combination with emission reduction to 2 qrl.

[atoma01]

I do no longer support this QIP. I think a move to PoS should take priority. See reasons by discord user II here:

«QRL needs to get off the PoW nothing-at-stake weakness & vulnerability. Subsidising a few dozen miners who have no allegiance and can switch their mining equipment on & off the chain at will & precluding the large majority of users from being into securing by PoS, in PoW there will be always manipulation, selling to recover all sorts of costs, without ever benefitting the QRL network nor bolstering its security assumption.

PoS has never had the alleged Noting-at-Stake Vulnerability, it's a PoW maximalists narrative that exactly accuses the counterpart of their own fundamental flaws.

PoW was never needed in QRL, and it has only damaged the community by constricting the network rewards to a few miners, most of which also sell continuously, be that to recover costs or to add enough sell pressure.

PoS would give immediate benefits to ALL participants, and while avoiding recurrently expensive real world dependencies around mining equipment that also entail real attack vectors, PoS would be run on any participants actual computing devices, mostly laptops and other mobile devices, securing the network more than sufficiently with the current distribution.

There would not be any need to seek investments that spend on mining equipment or buy hashing, liberating all amounts of money to be DIRECTLY invested in having actually Something-at-Stake in the QRL-network itself: QRL-Coins in the Coinbase. NO hashpower games, no chain switching.

PoS networks aren't less secure than PoW, and QRL has been distributed enough to function within its security assumptions, far better with a PoS than with PoW, which excludes practically everyone from the rewards.

Adoption is also happening by including as many as possible in the network growth, easily & effectively securing through a staking algorithm & incentives.

PoW inflates costs, attaches unnecessary real-world expenses & dependencies, invites hashpower gambling schemes & centralisation, creates a false security.

PoW is what kills QRL. Don't expect anything else from doing the same idiocies.

What's better, subsidising a few dozen PoW miners with no commitment guarantee or have 1000s of regular community members each adding to the security while getting also a small regular incentive through PoS; and much more decentralised, by staking wallet user numbers & locality?

QRL doesn't need RandomX nor RandomY mining algorithm switches, it won't change anything fundamentally wrong with PoW; and even lowering the emission would be just another futile attempt in the wrong direction of how this blockchain could evolve.

Furthermore, PoS has also other advantages in regards to voting & governance mechanisms, finality, and scalability.»

And my last comment on the Quanta emission rate QIP:

https://github.com/theQRL/qips/pull/21#issuecomment-555873942

[atoma01]

In regard to PoS this talk also some interesting viewpoints https://coinhub.news/cs/article/ethereumfoundation-the-case-for-proof-of-stake-by-emin-gun-sirer-devcon4

[cryptofuture]

I prefer RandomX even over long waited POS. As basically mining on the cpu is seems very cool

[lilyanatia]

Huge Monero hashrate can be used by pool operators for double spending. What do you think guys on that?

RandomX has parameters that can be configured to make it incompatible with Monero.

What about adopting more GPU oriented but unique ProgPow which has been created, but not being used yet?

ProgPoW has much weaker ASIC resistance than RandomX.

[curefrankosflus]

I support this QIP due to the my assumption that most hashrate now is generated through hardware (FPGAs/Asics) for CNv1 that it not publicly accessible. I think POS will not be around for another year at least. and as it has been quoted many times, the wheel doesn't have to be reinvented. QRL could just shadow Monero until POS is properly developed.

[surg0r]

I now vote we fork as soon as is safe to RandomX.

[Ottslayer]

totally on board with a soon as possible/safe fork to RandomX... we need a safer algo until POS

[AleGiovanardi]

\i agree with this pathway. QRL needs to establish a more intimate connection with average miners and GPU users. This will broaden our mining audience, thus QRL public recognition and usability. Network will also be strengthen by this new algo, because if QRL has the ability to change thier algo to stay up to date with techonological advances in crypto mining this is only a plus point for our community and the tech behind it. How can you really think to be "afraid" and against mining algo change if needed (as it is right now) and still pretend to be an avantguarde technology? Being on edge of tech development mean primarly the ability to shape according to the times and situations.

[StrikeAttack]

I agree with this QIP. We need to remove all highly specialized hardware from the hashrate and put mining back into the hands of our community and maintain better focus on decentralization. This is the best PoW option available before moving to PoS.

[Coin-Runner]

I agree with this QIP. The fall in node counts to 11 is an all-time low and not indicative of a healthy mining landscape. Pleased to see it has @cyyber's support. Is there a commitment that the change can be made in the next HF, scheduled Jan/Feb 2020 ?

[mastercyb]

I do agree that we have to switch ASAP

[Kryptoxic]

I fully support this qip as well. A lot of people have a CPU and by switching to RandomX, anyone that has a CPU can mine QRL which makes the network healthy.

[mastercyb]

Guys! Lets discuss, what we can do to roll out RandomX sooner. I can offer 10k quantas as a bounty for either implementation or security check

[surg0r]

@xhipster we have RandomX on devnet already. Public testnet will be rolled out imminently. Team expects a Feb hard fork date. Exchanges need 2-4 weeks to upgrade prior being most of delay now going forward.

[mastercyb]

Cool news. The reason I am rising the question is because I can not find new algo in the scope here. Although I can see it in the last update. (strange enough but I am pretty sure I was not able to find this in the first version of update, It was changed?). Amazing that the scope has been updated

[surg0r]

Scope needs a minor amendment to explain the two upcoming forks instead of one. Well spotted we'll get this fixed today.

Updated: https://github.com/theQRL/theqrl.org/pull/148

[Ottslayer]

I think the status (awaiting hard fork) should be updated.. same with the status of multi-sig :)