<m​onero.arbo:matrix.org> I'm not disagreeing with that. I guess I'm more saying, fcmp verification times look scary as fuck

<t​orir:matrix.org> There's some added complexity with batching being possible with multiple transactions in some cases? It's not clear to me.

<o​frnxmr:xmr.mx> Byte for byte, its notbad

<o​frnxmr:xmr.mx> Looks the same as ringct to me

<o​frnxmr:xmr.mx> On ringct, i can sync ~1-1.5 blocks per second at ~200kb avg blocks. I verify abt 250kb per second on fcmp

<o​frnxmr:xmr.mx> The biggest difference is in tx generation

<o​frnxmr:xmr.mx> Takes 7mins to create a 128in

<m​onero.arbo:matrix.org> that's a weird comparison to use if TXes are much bigger. the amount of data needed to be processed will obviously go up in that case

<o​frnxmr:xmr.mx> They are onlt much bigger at low inputs

<o​frnxmr:xmr.mx> Like 7kb for a 2in (5x size), but a 128in is like 110kb (maybe 10-15%larger)

<m​onero.arbo:matrix.org> I'm looking at github but do you have a link handy for ringct verification times?

<o​frnxmr:xmr.mx> no, just syncing mainnet on the same device that i'm spamming fcmp

<o​frnxmr:xmr.mx> Post checkpoints, log-level 1 on mainnet shows me abt 1.5blocks per second

<m​onero.arbo:matrix.org> there's a cahrt somewhere from when ring size was last being discussed, I'm sure

<m​onero.arbo:matrix.org> there's a chart somewhere from when ring size was last being discussed, I'm sure

<o​frnxmr:xmr.mx> berman said he wanted to run the bumbers at some point, so i would assume they arent available

<m​onero.arbo:matrix.org> anyway I think a lot of the sync time is disk IO? cause cuprate is like twice as fast at initial sync than monerod and I don't think it's cause they verify transactions so much faster

<o​frnxmr:xmr.mx> Also need to test on hdd. Its hypothesised that fcmp will do better on hdd

<o​frnxmr:xmr.mx> Cuprate is much faster for various reasons, such as usinf 64bit accellerated libs

<o​frnxmr:xmr.mx> Theres a pr on monero to use the accellerated versions as well, increased sync 40% in my testing

<r​avfx:xmr.mx> For a thing, it should not have to forever seek to verify right

<o​frnxmr:xmr.mx> It doesnt have any effect on fcmp though

<o​frnxmr:xmr.mx> Wdym?

<r​avfx:xmr.mx> With current sheme dont you have to check the tx in the decoys (all of them) when you verify? Some new, some old...

<o​frnxmr:monero.social> Yeah, has to do db reads for decoys

<o​frnxmr:monero.social> Fcmp doesnt have to do that, which could lead to faster hdd sync

<r​avfx:xmr.mx> Orders of magnitude faster yes.

<r​avfx:xmr.mx> Monero should host a bootstrap from the beginning til the fcmp switch.

<r​avfx:xmr.mx> That should fix all sync issue for hdd I think

<o​frnxmr:monero.social> My hdd is failing, otherwise i'd check :P

<r​avfx:xmr.mx> With the current sheme your constantly seeking, so you have seek time + time for the thing to bring the sector to the head once it landed on the track, with can be instant up to one full rotation (5400.. 7200.. Rpm)

<r​avfx:xmr.mx> For each uncached decoys

<e​longated:matrix.org> Wil need to check for pre fcmp inputs though?

<o​frnxmr:monero.social> Eh.. my testnet doesnt have any/many of those :P

<r​avfx:xmr.mx> Yes, its why I proposed a bootstrap

<a​sdfqwfe:matrix.org> The US Gov has nuclear powered data centers. They will always be able to 51% attack cheaply.

<s​pirobel:kernal.eu> endor00: anything other than pow is just hype == anything other than pos that uses the ([CPU / asic / gpu + electricity] coin as eligible stake) is just hype

<a​sdfqwfe:matrix.org> That which can be asserted without the evidence can be dismissed without evidence

<a​sdfqwfe:matrix.org> PoW is hype

<r​avfx:xmr.mx> Should I get ready for POS and buy 2% of the supply?

<c​ountbleck:matrix.org> I'm not sure how comfortable I am with Monero's security being tied to holding Monero as opposed to expending something in the physical world

<c​ountbleck:matrix.org> Would PoS limit the sort of attacks that we're experiencing now?

<c​ountbleck:matrix.org> How big of a whale would you have to be to disrupt the network?

<s​pirobel:kernal.eu> yes. it would be an order of magnitude more expensive. we had the discussion the in the mrl meeting

<c​ountbleck:matrix.org> Ah, link?

anyone know when the mrl meeting logs will be posted?

<b​arthman132:matrix.org> I get that man, but only huge cryptocurrency’s like bitcoin could survive something like that. There is compelling evidence that suggest it’s not possible anymore to do a 51% attack anymore for Bitcoin, because the cost to do so is outrageous and it would require a total of 7 million asics, which is more than currently available on market.

<b​arthman132:matrix.org> cointelegraph.com/news/bitcoin-ethe…rcent-attacks-coin-metrics-research

<r​avfx:xmr.mx> Yeah, the rich would own the thing. Meaning you have to have more money than them to 51% it

<s​pirobel:kernal.eu> libera.monerologs.net/monero-research-lab/20250813#c557536

thanks

<r​avfx:xmr.mx> Except if the rich are malicious (state actor buying there way in, faster than setting up mining rig and no need to micro manage them)

<a​sdfqwfe:matrix.org> I'm no expert, but I think to attack PoS, you have to PUMP THE BAGS of the existing holders

<c​ountbleck:matrix.org> It looks like PoS discussion didn't begin at that point

<r​avfx:xmr.mx> Exacly

<a​sdfqwfe:matrix.org> So the 51% who sold... They leave with a fat pile of BTC/fiat

<r​avfx:xmr.mx> Easier for state actors, harder for the common asshole

<a​sdfqwfe:matrix.org> And they can just take their bag and spin up a new PoS privacy chain

<a​sdfqwfe:matrix.org> Rinse and repeat

<a​sdfqwfe:matrix.org> The key insight is this

<t​orir:matrix.org> I think if Monero does Proof-of-Stake, we need to invent a way to disincentivize exchanges from staking their customer's coins, or we just get back to the large mining pool problem except now it is centralized exchanges beholden to government.

<a​sdfqwfe:matrix.org> In pure PoW the cost to attack and the cost to defend are 1:1

<a​sdfqwfe:matrix.org> PoS is asymmetric. It favors defense

<a​sdfqwfe:matrix.org> You have to pump their bags

<r​avfx:xmr.mx> The question is who is going to get an easy grab of the network?

<a​sdfqwfe:matrix.org> Coin distribution is more of a problem if you're doing pure PoS from the Genesis block

<a​sdfqwfe:matrix.org> Less if you migrate from PoW

<r​avfx:xmr.mx> At the end, cex and some elite/state will own xmr, sound good

<r​avfx:xmr.mx> As long as they need xmr

<s​pirobel:kernal.eu> no inflation

<t​orir:matrix.org> In PoW, you need to pay ongoing maintenance costs to continue holding your "share" of the network defense cost. In PoS, the ongoing cost is lower, which means people will stake and never have an incentive to unstake. Could we add some sort of meaningful maintenance to staking?

<r​avfx:xmr.mx> Why would the exchange not stack it. If its not some way to control the network

<r​avfx:xmr.mx> They are already online 24/7

<r​avfx:xmr.mx> It cost nothing for them, reward or no

<s​pirobel:kernal.eu> give them the tx fees. that also gives them an incentive to make sure people actually use the network

<t​orir:matrix.org> Transaction fees don't scale. In practice we know users will migrate to coins with lower fees.

<t​orir:matrix.org> Bitcoin is going to have that problem once the block emission runs out.

<r​avfx:xmr.mx> Bitcoin rely on TW/h

<s​pirobel:kernal.eu> the less incentive the better. only truly committed participants should be staking. the others can be bribed in any case. (similar to opportunistic hashrate switching to qubic or getting rented by them)

<r​avfx:xmr.mx> That have a cost

<r​avfx:xmr.mx> Stalking have no cost (for exchange s)

<t​orir:matrix.org> Transaction fees work as a cost to attack the network via spam attacks. They don't work well as a financial incentive to secure the network, not when users will switch to another network to have lower fees.

<s​pirobel:kernal.eu> financial incentives dont in the form of staking rewards or mining rewards dont really secure the network.

<s​pirobel:kernal.eu> financial incentives in the form of staking rewards or mining rewards dont really secure the network.

<s​pirobel:kernal.eu> opportunistic hashrate or stakers can be bribed

<s​pirobel:kernal.eu> lets say you are a whale right now, if you wanted to contribute to the security of the network you would have to sell part of your xmr to mine at a loss. From this most interested in the security of the network participant's perspective no inflation PoS would be better. Currently we have proof of stake that you have to pay for instead of receiving a reward. (and it still works bec<clipped message>

<s​pirobel:kernal.eu> ause people believe in monero so much)

<r​avfx:xmr.mx> Stalkers can serve there own interrest. And its easy way in

<r​avfx:xmr.mx> Blaclrock can scoup 10% of the supply tomorrow

<e​longated:matrix.org> Atleast they will not lend it to market makers 😅 and short it to ground

<r​avfx:xmr.mx> Problem with the pow model is only the price

<e​longated:matrix.org> Blackrock is not interested in your tiny coin

<s​pirobel:kernal.eu> would be more expensive than buying the hashrate

<r​avfx:xmr.mx> If xmr was at its just price, we would not have issue

<e​longated:matrix.org> Current pow is the problem

<t​orir:matrix.org> There are potential threat actors willing to attack Monero at a loss. Qubic is believed to be mining at a loss. A government might also be willing to perform an attack against the coin. Giving defenders a financial incentive to secure the network probably helps more than giving them nothing.

<r​avfx:xmr.mx> Hashrate require time, expertise, employees, etcetc

<r​avfx:xmr.mx> Pos require pressing a button on kucoin

<t​orir:matrix.org> There are coins in existence with no inflation and PoS-like consensus such as Nano.

<e​longated:matrix.org> Qubic has no cost, they are burning vc funds and greedy miners who are paid in their shitcoins

<e​longated:matrix.org> Atleast they will be staked and not sold by market makers

<r​avfx:xmr.mx> Qubic is a ponzi per definition

<t​orir:matrix.org> I think the solution most likely to be adopted for Monero is PoW + no emission PoS finality layer.

<t​orir:matrix.org> With PoW still providing emission.

<e​longated:matrix.org> Pow+pos for finanality , pos needs incentive

<t​orir:matrix.org> It doesn't in Nano.

<e​longated:matrix.org> Nano is useless

<t​orir:matrix.org> Qubic is more useless than Nano yet still is managing to cause us a massive headache. Let's consider what we can learn from "useless" coins.

<r​avfx:xmr.mx> Ponzi work wonferfully til they dosent

<b​asses:matrix.org> just rich CEOs

<r​avfx:xmr.mx> Future XMR controller

<r​avfx:xmr.mx> What if they vote to return to decoys and limite it to 4 decoys? For compliance

<b​asses:matrix.org> if we had a CEO things would have been easier 😔

<t​orir:matrix.org> The actual hardest part of mitigating 51% will be coming to consensus on a solution, I see.

<b​asses:matrix.org> also a lot of venture funding and gov funding like zcash

<r​avfx:xmr.mx> If only xmr was at its just price of at least 4k per coins

<r​avfx:xmr.mx> We would not have that discution

<3​21bob321:monero.social> Rock,scissors,paper

<e​longated:matrix.org> It will never reach there with current randomx

<r​avfx:xmr.mx> Sure, centralizing mining with make it better

<r​avfx:xmr.mx> somehow

<b​arthman132:matrix.org> matrix.monero.social/_matrix/media/…matrix.org/vofnWwkJUHxIBvKhJdPsUlsR

<b​arthman132:matrix.org> Is mexc allowing back monero?

<e​longated:matrix.org> Just getting reducing effect of botnets will do much better. Bump up ram requirements it’s not a specialised hardware

<r​avfx:xmr.mx> Bump up ram is to prevent specialized hardware to heat it all

<r​avfx:xmr.mx> yep

<e​longated:matrix.org> Yah or just continue being rags with botnets

<r​avfx:xmr.mx> I mine right now

<r​avfx:xmr.mx> I would mine qubit or some other shit if Monero required ASIC

<e​longated:matrix.org> You can’t buy 16gb ram ?

<r​avfx:xmr.mx> OH, then I missread you

<r​avfx:xmr.mx> So you mean, bumping the mining so it require more ram?

<e​longated:matrix.org> Yes

<r​avfx:xmr.mx> That could work.

<r​avfx:xmr.mx> Same process but requiring a lot more.

<e​longated:matrix.org> Don’t want to work ? Want freebies like botnets? Stay poor with price

<r​avfx:xmr.mx> How botnets affect the price, they provide hashrate.

<r​avfx:xmr.mx> Sure, bumping the botnet out could bring more honest actors but we might also endup with less hashrate

<r​avfx:xmr.mx> The reward coupled with electricity cost is what bring hashrate

<r​avfx:xmr.mx> 0.6 xmr at 250$ is different to 0.6 xmr at 4k

<r​avfx:xmr.mx> at the end, just an increase in price would put the botnet into the insignifiant bracket

<e​longated:matrix.org> Botnet owners don’t support price, it costs them almost nothing to mine xmr ; they do spend dump at any cost and it affects price

<e​longated:matrix.org> If someone has paid for their rig they won’t sell at any price

<r​avfx:xmr.mx> because the botnet change according to how much shit they can infect / propagate

<m​onero.arbo:matrix.org> even if every miner dumps, 100k a day in sell pressure shouldn't be that impactful

<r​avfx:xmr.mx> they don't really care about the price, for them the most they infect the better

<e​longated:matrix.org> Costs almost nothing to mine, zero cost

<r​avfx:xmr.mx> it cost nothing to botnet but do you think that if price was 4k, the botnet would infect like 20x more machines

<r​avfx:xmr.mx> because if you take the account of electricity cost / reward you would see that price alone would bring a crapton of hashrate

<e​longated:matrix.org> They have been doing it for years, someone buys a few million in xmr, just to come back in a week to see someone mindlessly selling at any price

<s​herry:unredacted.org> xmr support by lots of botnet/low performance devices, change to something with 16gb RAM, maybe kill xmr network healthy

<e​longated:matrix.org> It’s not healthy, it’s rotten

<r​avfx:xmr.mx> it would nuke the botnet and at minimum make xmr more vulnerable until replacement hashrate is brought online.

<r​avfx:xmr.mx> Sure, botnet hashrate would be replaced by more honest hashrate if the ram requirement was higher (again, reward & electricity cost is pretty much the only thing that matter)

<r​avfx:xmr.mx> botnet gone == extra reward buffer

<e​longated:matrix.org> With current state of xmr, legit miners are buffer

<r​avfx:xmr.mx> have to be careful with that migration, it's just what i'm saying.

<r​avfx:xmr.mx> bumping the requirement will probably shut the botnet instantly but it will take more than instantly to brind replacement hashrate

<r​avfx:xmr.mx> have to be careful with that migration, it's just what i'm saying.

<r​avfx:xmr.mx> bumping the requirement will probably shut the botnet instantly but it will take more than instantly to bring replacement hashrate

<e​longated:matrix.org> With current randomx, you need pos for actual security

<s​herry:unredacted.org> actually, nobody knows how many hashrates support by botnet, 30%? 70%?, if just 30%, ok for replace to a bigger algo, but if 70%...

<s​herry:unredacted.org> more serious to next step

<r​avfx:xmr.mx> Can they slowly increase the ram requirement, that would be ideal. That would scrap out the botnet.. slowly while healty hashrate replace it

<m​onero.arbo:matrix.org> botnets are good actually?

<s​herry:unredacted.org> i think botnet is bad

<m​onero.arbo:matrix.org> hashes are hashes

<r​avfx:xmr.mx> We don't know what there up to too much

<r​avfx:xmr.mx> but they do provide hashrate.

<r​avfx:xmr.mx> if they are proving 70% and we nuke them in one day then it will be way easiler to attack monero till the replacement hashrate is brought online

<r​avfx:xmr.mx> And yeah, I don't think we can even know how much hashrate there prividing, so a gentle scrapping would be ideal... imo...

<e​longated:matrix.org> For the operators, not price or confidence of ppl investing in mining or coin

<m​onero.arbo:matrix.org> Explain again how kicking honest miners off the network helps against a 51% attack again

<m​onero.arbo:matrix.org> Explain again how kicking honest miners off the network helps against a 51% attack

<e​longated:matrix.org> Botnets are not honest

<m​onero.arbo:matrix.org> If you produce valid blocks are aren't employing selfish mining techniques, that's an honest miner

<e​longated:matrix.org> Just needs a bigger botnet to mine for cubic

<e​longated:matrix.org> Free zero cost mine anything

<r​avfx:xmr.mx> Yeah, it would also kick people wou have less ram.

<r​avfx:xmr.mx> 16GB "should be ok" but 12GB is better.

<r​avfx:xmr.mx> a LOT of honest miners use that traditionnal "two 8GB dual rank stick" so they have 16GB today

<r​avfx:xmr.mx> And Ryzen run a lot shittier if you stuff the four slots so it would force them to replace 16GB for 32GB. It will incure a cost per 20Kh/s

<e​longated:matrix.org> Mine whatever pays more

<r​avfx:xmr.mx> is 20KH/s is worth 150$?

<r​avfx:xmr.mx> considering it make less than 10 a month, that is the question

<m​onero.arbo:matrix.org> you just described most crypto miners

<r​avfx:xmr.mx> Most botnet are "probably" on shit hardware and office shit that have 8GB or less. Again, probably

<e​longated:matrix.org> Good luck

<r​avfx:xmr.mx> Just rising the bar to 6GB should scrap a ton of them

<r​avfx:xmr.mx> I would say if they can make code that force the algo to require 1GB more, every months, until we are at 12GB, could be the way.

<r​avfx:xmr.mx> The only honest people that will get anoyed by that, are the one that mine on the computer they use at the same time

<r​avfx:xmr.mx> 12GB out of 64GB is nothing, but 12GB out of 32GB might be a lot depending of the users

<e​longated:matrix.org> They are not going to even increase it by 1mb, they will give rpi excuse to keep mining themselves on botnets

<r​avfx:xmr.mx> do people even mine with that garbage grade hardware?

<e​longated:matrix.org> Ask lyza and crew defending their botnet algo

<r​avfx:xmr.mx> lol

<r​avfx:xmr.mx> for the price of an orange pi, you can get an used thinkcentre with 8 or 16GB ram, that have like 6x the oomph of a orange pi (plus it come with a metal case, 3 display output, 6-7 usb, 1-2 nvme and sata).

<r​avfx:xmr.mx> rpi/opi are such an overpriced scam

<r​avfx:xmr.mx> except if your plan is to make a computer for you car or to send it to space or something

<e​longated:matrix.org> Botnet loving crew is responsible for current state of xmr and will be dead in price and security in couple of years

<e​longated:matrix.org> Anyways I am out, this is MRL; I can’t do daily botnet rants

<s​herry:unredacted.org> lol btw my laptop only 8G RAM XD

<s​herry:unredacted.org> with core 2 duo CPU

<s​herry:unredacted.org> if change to bigger algo, i cant mine xmr on this laptop for fun! (just kidding

<r​avfx:xmr.mx> Bymping to 12GB would remove only 4kh/s from my stash

<e​longated:matrix.org> Yes you are mining so much, thanks without your laptop mining for masters; xmr wouldn’t be where it is

<g​ingeropolous:monero.social> i love that I can write into the bots "cryptographic whatsits" and it responds with "oh you mean commitments"

<e​longated:matrix.org> PoS + PoW Hybrid Would Work for Monero

<e​longated:matrix.org> In a hybrid design, both miners and stakers must agree on a block.

<e​longated:matrix.org> Example:

<e​longated:matrix.org> • PoW miner creates a block →

<e​longated:matrix.org> • PoS validators sign/finalize it →

<e​longated:matrix.org> • Block enters the chain only if it passes both checks.

<e​longated:matrix.org> This means an attacker would need:

<e​longated:matrix.org> 1. >50% of the hash rate and

<e​longated:matrix.org> 2. >50% of the staked XMR

<e​longated:matrix.org> …which is exponentially harder than just dominating PoW.

<p​reland:monero.social> (If the goal is to completely control the chain, then yes.

<p​reland:monero.social> If the goal is only to disrupt or stall the network, 50% in either would be enough to prevent consensus

<p​reland:monero.social> Back to hiatus)

<e​longated:matrix.org> You need both for 51% attack

<e​longated:matrix.org> You may halt it for sometime but reorgs are not possible

<e​longated:matrix.org> You may halt it for sometime but deep reorgs are not possible

<b​arthman132:matrix.org> That sounds like a good idea

<b​arthman132:matrix.org> How do you guys feel about a hybrid like system?

<m​onerobull:matrix.org> i like it

<h​bs:matrix.org> It all depends how validators are chosen, how many there are and what their power is. Introducing an additional layer should be thought out as to not weaken the resistance of Monero to censorship/coercion. As of now I see it as abandoning decentralization, but happy to change my mind as details unfold

<m​onerobull:matrix.org> i like the 1 week epoch idea, not sure about PoS-ing every block for instant finality.