<o​frnxmr:xmr.mx> No, liquidity period. 80% of zano is staked, which is why the coin pumps on 0 liquidity

<e​longated:matrix.org> Small as in the community the number of miners the actual security cost

<d​iego:cypherstack.com> Which is why a block reward split of 80-20 miners-stakers might be feasible? Still a way to get money without staking while also having finality layer

you say that like we can flip a switch and turn off the network, hi585858

<e​longated:matrix.org> Zano is premined shitcoin, why are we discussing that ?

44 out last 100, indeed

<o​frnxmr:xmr.mx> I said 10:1, with only coinbases stakable

not asking to turn off the network

just sharing actual historical fact happening rn

<o​frnxmr:xmr.mx> Monero mined 100k coins in the first 3 days

<d​iego:cypherstack.com> yeah my numbers are not final, just representative. 10:! works too.

I am recording, trust me hi585858 :)

I hoard the data

<e​longated:matrix.org> So did btc ?

its impressive how they gain from 10 block to 80 block and

now*

<o​frnxmr:xmr.mx> So when you talk about premine, remember monero did not have a fair emission, crippleminer, fast emission, etx

another point is, would all coinbases be included, or only ones after certain date? or aged?

<o​frnxmr:xmr.mx> all coinbases, i think

which also brings issues about starting up PoS or finality layer that way

<e​longated:matrix.org> Crippleminer was bytecoin

<o​frnxmr:xmr.mx> No, it was monero

<d​iego:cypherstack.com> so spend a coinbase once and it becomes ineligible forever?

<o​frnxmr:xmr.mx> Yup, diego

<d​iego:cypherstack.com> potentially a problem with other coins, but with a tail emission not so much

<e​longated:matrix.org> Afaik it was bytecoin and hence the fork to xmr

<o​frnxmr:xmr.mx> Fork was due to premine lies

<o​frnxmr:xmr.mx> Monero forked, and had crippleminer and later proprietary miner that wasnt as crippled

the other goal is to see if they turn on their selfish mining today as they said

<d​iego:cypherstack.com> I am a Monero historian, having read all the things. Monero was cripplemined. Bytecoin people were salty that Bytecoin scam failed, and were saltier that Monero got taken from them, so they released a crippled miner under a different alias, and people used it.

<d​iego:cypherstack.com> da-data.blogspot.com/2014/08/minting-money-with-monero-and-cpu.html

<g​ingeropolous:monero.social> naw, bytecoin had the crippled miner originally

<g​ingeropolous:monero.social> and monero just inherited it

<g​ingeropolous:monero.social> cripple miner may have been part of the creating that backstory of bytecoin being around for years

<d​iego:cypherstack.com> Right. 2-3 weeks of cripple mining before the fix.

<o​frnxmr:xmr.mx> I'm trying so hard to hit 20mb blocks on fcmp. Highest is 15mb so far

<o​frnxmr:xmr.mx> hard to maintain due to lack of inputs / locked inputs. Started gaming the median by mining an 1/3 blocks as empty, but still run low on inputs before i can get 50 blocks at over 10mb 🙃

<g​ingeropolous:monero.social> this finality layer business would still leave us open to empty block mining and tx censoring etc

<o​frnxmr:xmr.mx> And open the door for someone to outspend us

<e​longated:matrix.org> How does tx censoring work ? You can know which decoy is being spent no Addr no amount

<o​frnxmr:xmr.mx> and, iiuc, the chain can halt if validators dont come to concensus

<e​longated:matrix.org> For how long ?

<e​longated:matrix.org> How does tx censoring work ? You cant know which decoy is being spent no Addr no amount

<o​frnxmr:xmr.mx> Until we hard fork

say they want to prevent p2pool miners from moving coins

<e​longated:matrix.org> Which pos was stuck ?

they can identify who it is due to how p2pool mines

<e​longated:matrix.org> Which pos chain was stuck ?

if they sweep outputs at once

here's some recent ones tied to a specific miner address p2pool.observer/sweeps

they might have alternate ways of doing so

<e​longated:matrix.org> P2pool is special case

yes, but they could have a view on an exchange transactions, for example

<o​frnxmr:xmr.mx> With a tfl? Idk

<e​longated:matrix.org> Okay, same can be said for pow miners ? Pools?

there's ways monero tries to prevent this and make it hard to know where a transaction comes from or who is

<e​longated:matrix.org> Idk what’s special for pos, same can be true with pow

but that's a different problem. if an attacker that can sustain 51% hashrate long term exists

and they know a txid

they can mine everything but that tx

<e​longated:matrix.org> With hybrid attacker needs to control both

and orphan any block that contains it

with finality?

<e​longated:matrix.org> Same for pow

if inclusion is decided solely by pow, no

<o​frnxmr:xmr.mx> There can be staking pools too

<e​longated:matrix.org> Yes same for pow or pos

<DataHoarder> yeah, the yields% that are usually done for by staking is ... waay too high to what monero, already on tail emission, is <<>> someone mentioned that ETH pays 2%

<d​atahoarder:monero.social> Yes I was always referring to pow, elongated. These messages made no sense on IRC as there was no threading there

For context how it looks in irc irc.gammaspectra.live/f8233335da0f5672/IMG_6843.jpeg

<e​longated:matrix.org> I think we need to hire some researcher for pos/pow to make it suitable for xmr

<e​longated:matrix.org> Maybe hire cypherstack to do some research, if they are interested

make ccs people will donate 100%

<e​longated:matrix.org> Diego Salazar:

POS is a change that totally moves the fundamental ways that Monero works. So assume there will be more friction or longer time before it can be done, compared to privacy improvements/changes in signatures and transactions, or even the PoW function change (which I wasn't here for originally!)

fundamental is security and privacy

communism is not

is this the offtopic channel

<e​longated:matrix.org> You can’t reject/move to hybrid, without any research

<r​ottenwheel:unredacted.org> nioc where are we and where is my beanie?

I am with Cat

<r​ottenwheel:unredacted.org> meow.

just about to give her her bedtime snack

she remains unconcerned

51% unknown on last 100 block

<d​iego:cypherstack.com> Hi its me.

<b​arthman132:matrix.org> Did they get it or did they just barely missed it?

<e​longated:matrix.org> Would cypherstack be interested in hybrid pow/pos research?

<d​iego:cypherstack.com> Yes but you need to pay in qubic

<d​iego:cypherstack.com> I jest. Gimme the money and off we go (we're already low key looking at bit)

<e​longated:matrix.org> Can you open a ccs? Or have a ccs on you own website

<d​iego:cypherstack.com> Oooh I can finally try out the whole Kuno thing, or whatever its called.

<e​longated:matrix.org> Kuno works

<d​iego:cypherstack.com> Although honestly I was going to float the idea by you all of a general CS retainer.

<d​iego:cypherstack.com> People and projects are trying to snap up our hours recently and last time I didnt give Monero first dibs on things people got mad at me.

<e​longated:matrix.org> CS retainer is not a bad idea, maybe discuss it in next MRL meeting ?

<a​rticmine:monero.social> This can be very useful in order to encourage self custody for staking

<o​frnxmr:xmr.mx> I'm against staking pools

<a​rticmine:monero.social> Why

<a​rticmine:monero.social> What attack vectors do you see

<o​frnxmr:xmr.mx> The idea of getting paid to deposit your $ on an exchange or pool - youre essentially no different than a speculative stock investor

<o​frnxmr:xmr.mx> Its custodial

<o​frnxmr:xmr.mx> Like eth. "32 eth min" so everyone just lets binance stake for them

<a​rticmine:monero.social> ... but is a self custody staking pool possible?

<o​frnxmr:xmr.mx> yea, and cold staking as well

<a​rticmine:monero.social> I see self custody of the stake as a way to mitigate POS attack vectors

<o​frnxmr:xmr.mx> academy.particl.io/en/latest/part-g…king.html#connect-to-a-staking-pool

<o​frnxmr:xmr.mx> > It is highly recommended to be careful about which cold staking pool you choose to stake your funds with. In fact, because you are effectively delegating your staking power to a third-party, staking from a cold staking pool tends to centralize the staking weight of the network.

<o​frnxmr:xmr.mx> particl.wiki/learn/staking/pools

<o​frnxmr:xmr.mx> Its similar centralization issue as mining pools

<a​rticmine:monero.social> What about a P2P self custody staking pool?

<f​iatmoneysucks:matrix.org> How it will works?

<o​frnxmr:xmr.mx> I dont think you'd need a pool for that, and personally i think stakers should have to participate and run their own infra

<o​frnxmr:xmr.mx> I really don't like the idea of "whale delegates view key of their ledger, and checks back in 20 years"

<a​rticmine:monero.social> So block singing for the staking pools

<a​rticmine:monero.social> Signing

<o​frnxmr:xmr.mx> In my brief time being around the particl community, ive grown a distaste for the way everybody in pos behaves (passively)

<a​rticmine:monero.social> We need to start thinking outside of the box here

<a​rticmine:monero.social> Like the Bitmain ASIC version of POS

<o​frnxmr:xmr.mx> my outside the box was: hot staking only (wallet needs to be online to sign blocks), coinbases only (initial stake comes from pow. moving the funds removes the stake, irreversibly).

<o​frnxmr:xmr.mx> Problem with coinase-only is that an entity like qubic could quickly obtain a majority% if they opt not to sell.

<o​frnxmr:xmr.mx> on the other hand, to stake, you'd be forced to contribute pow. You wouldnt be able to stake someone elses coins, or take out a loan to borrow coins against

<o​frnxmr:xmr.mx> The "staking pools" in this scenario, would be mining pools who offer extra reward if you dont withdraw your coins. So it doesnt eliminate pool staking (unless paired with signing the blocks)

<a​rticmine:monero.social> Block signing doesn't prevent pool mining. At least how I see it

<o​frnxmr:xmr.mx> block signing to prevent mining pools from staking your blocks

<b​arthman132:matrix.org> I mean there are douchbags in every community, but that is just how the world goes

<o​frnxmr:xmr.mx> My original idea was to split block production into 4min pow blocks and in pos blocks, with a 10:1 emission split. Reorging beyong 2 blocks on either side would require control of both sides.

<o​frnxmr:xmr.mx> i dont think splitting block production was received well either, but the pros are..

<o​frnxmr:xmr.mx> 1. Emission is still majority pow. Pos blocks are essential a soft finality

<o​frnxmr:xmr.mx> 2. P2pool payouts would be larger, due to increased pow block rewars

<o​frnxmr:xmr.mx> 3. Reorgs would be more difficult due to reduced block frequency

<a​rticmine:monero.social> If the signing requirements are only 1% the pools can stake the balance

<o​frnxmr:xmr.mx> 4 min pow and* 4 min pos

<a​rticmine:monero.social> I was thinking along the same lines with a 50%split block reward between POS and POW

<o​frnxmr:xmr.mx> i think 10:1 is nicer :P. Majority goes to pow miners, with pos being a bonus 10% for keeping your coinbases locked up to finalize the pow blocks

<b​oog900:monero.social> > Reorgs would be more difficult due to reduced block frequency

<b​oog900:monero.social> I still don't think this is the case when you take into account the amount of work reorged vs just the amount of blocks

<o​frnxmr:xmr.mx> 50:50 would probably be more attractive to only use pos, and turn off the pow

<a​rticmine:monero.social> The real question here is: Does the why is greater than the sum of its parts apply to security?

<a​rticmine:monero.social> Whole is greater than the sum of its parts between POW and POS

<a​rticmine:monero.social> For security

<o​frnxmr:xmr.mx> The diff would be double, so a 1 block reorg is essemtially 2, and if a pos block comes before you reorg, then youd have to reorg that too. Ignorinf pos entirely, its harder to reorg 1 10min block than 5x2min blocks, isnt it?

<o​frnxmr:xmr.mx> Or actuallt, that might be roughly the same, but youd have less reorgs on 10min blocks than on 2min ones

<a​rticmine:monero.social> The idea is mining on the POS block and staking on the POW block

<b​oog900:monero.social> The probability of reorging a number of block decreases exponentially the more blocks, so mining 1 10 min block is "easier" than 2 consecutive 5 min blocks

<a​rticmine:monero.social> So they alternate

<o​frnxmr:xmr.mx> Boog, corrected myself here

<o​frnxmr:xmr.mx> Yup. Diff targetting 4mins per side should have ~360 pow and ~360 pos blocks per day

<c​aptaincanaryllc:matrix.org> if more blocks makes it harder to reorg, how far can you push it?

<b​oog900:monero.social> it wouldn't be roughly the same but yes the amount of random alt blocks should decrease.

<b​oog900:monero.social> 1 block per hash would make it impossible without 51% :)

<b​oog900:monero.social> and assuming all blocks reach the whole network instantly

<b​oog900:monero.social> FWIW I was talking about when the attacker has less than 51%, with more than that they can reorg whatever they want (theoretically) with slower or faster blocks

<c​aptaincanaryllc:matrix.org> right, so with more frequent blocks, qubic reorgs may not have happened

<c​aptaincanaryllc:matrix.org> so increasing block frequency is a band-aid on its own? what are the downsides

<o​frnxmr:xmr.mx> No, with more frequent blocks, theyd have more reorgs

<o​frnxmr:xmr.mx> If blocks were 1 minute, assuming 100% luck, theyd have 14 block reorg instead of 7

<o​frnxmr:xmr.mx> If they had bad luck, they might get 10 blocks, good luck might get 18 blocks

<o​frnxmr:xmr.mx> more frequent blocks means you need consistent good luck. But each individual reorg is easier

<b​oog900:monero.social> you have assumed their share of the network has dropped by half

<b​oog900:monero.social> increased by double*

<b​oog900:monero.social> their share of the network is the same

<b​oog900:monero.social> they are still the minority hash power, this is all dependent on their luck

<b​oog900:monero.social> monero-project/research-lab #102#issuecomment-2402750881

<b​oog900:monero.social> to get 7 blocks at 20% of the network is 1.4% chance, to get 14 blocks at 20% of the network is 0.04571%

<o​frnxmr:xmr.mx> No, i assumed that the diffoculty has decreased by half

<b​oog900:monero.social> their share of the hash power is the same

<b​oog900:monero.social> it wouldn't be double because of the exponential but you get what I mean

<o​frnxmr:xmr.mx> Yeah, i'm sleeptalking (i dont sleep)

<o​frnxmr:xmr.mx> I see where i muddied my assumption

<b​oog900:monero.social> you would also need to take into account that with more blocks you get more attempts at reorgs

<b​oog900:monero.social> but because it drops off exponentially faster blocks would still be better for protecting PoW

<g​ingeropolous:monero.social> ugh im not seeing that. where in the paper does it directly talk about the different duration of blocks? the targed blocktime?

<g​ingeropolous:monero.social> ah nvm

<g​ingeropolous:monero.social> so 10 second blocks here we come?

p2pool 10 second blocks!

with uncles :)

<s​pirobel:kernal.eu> ArticMine: no. its better to get rid of PoW entirely. it serves zero purpose. the hybrid is only to keep the myth intact for a bit so people handle the transition better. It would be more honest to transition directly instead of putting whool on peoples eyes as you said. ofrnxmr staking from coinbases makes no sense and introduces needless fungibility problems.

<s​pirobel:kernal.eu> 1. in the case of Monero we want as low a stake as possible in any case to increase the privacy pool

<s​pirobel:kernal.eu> 2. safety is just a liveness problem with the added twist that a third party potentially lost money to a double spend. After the network is restarted and rolled back to a state before the attack, the validators can vote to slash the attacker stake or use it to compensate the third party, if this third party is sufficiently likeable.

<s​pirobel:kernal.eu> the downside of keeping pow around: 1. waste of money 2. slower finality 3. if an attacker spends tens to hundreds of millions on the stake, is he really going to stop at renting a few cpus as well? the attack cost against PoW is miniscule compared to attacking PoS. Also: if the CPUs are bought they can be reused in the next attack. With slashed stake that is impossible.

<j​ohnruth:matrix.org> Probability of consistent luck is lower. If you flip a coin five times vs ten times, getting all heads is 32 times harder in second scenario

<j​ohnruth:matrix.org> For those who are brainstorming ways to secure network from this attack; one way is to look how to make monero more resilient, the other is to poke holes in qubic. Sometimes, the best defence is good offence

<j​ohnruth:matrix.org> “There is no security on this planet, only opportunity”

Then you're just playing whack a mole every time a Qubic appears. Eventually a mole will eat through to the pantry.

<j​ohnruth:matrix.org> I strongly disagree, lived long enough

<j​ohnruth:matrix.org> I disagree. It’s like dealing with a bully. You show you don’t strike and many will try to take a jab at you. You rearrange somebody’s teeth once and nobody will fk with you

I think you are misapplying a lesson you have learnt before.

The immediate obstacle you will face here is that of attribution and reach.

A bully will be in range. Consider a hidden bully with a ranged weapon.

And no, saying "I go to the bully and proceed to rearrange teeth" is not a good answer.

The thing on the internet is that it is easy to "hide" and two opponents have different capabilities. Moreover, use of those capabilities may bring abot attention from a bigger bully, one you do *not* want to face off in a teeth rearrangement situation.

<j​ohnruth:matrix.org> Sounds like you never had to assert yourself or you like to live in fear, but this isn’t the point. It’s monero talk and I was making a metaphor

<o​frnxmr:monero.social> Plz explain how go rearrange someones face over the internet

<o​frnxmr:monero.social> How-to*

<o​frnxmr:monero.social> Or what the actionable implication was

<j​ohnruth:matrix.org> That was metaphor. I was talking about qubic having possible vulnerabilities or wholes that can be exploited

<j​ohnruth:matrix.org> Holes*

<j​ohnruth:matrix.org> The only thing I know is they have halving next week, their “Ai training” is vapourware there isn’t any actual Ai model being trained and they pay. They pay (with btc) for hash on mining rentals and they paid influencers and shai for hit pieces

<h​bs:matrix.org> which makes p2pool miners more exposed to QC risks

<t​orir:matrix.org> Previous message was a reply to DataHoarder:

<t​orir:matrix.org> > <D​ataHoarder> they can identify who it is due to how p2pool mines

<t​orir:matrix.org> (IRC can't see replies in Matrix)

<h​bs:matrix.org> My understanding was that a FL would require a limited set of validators, so not that many miners could participate and stake would be low per staker then

<a​ntilt:we2.ee> often, the higher your stake, the greater the probability to be selected as validator for an "epoch". But the curve can be made flat, too

<0​xfffc:monero.social> if anybody can, I need this: link.springer.com/article/10.1007/s10207-024-00849-5

<v​enture:monero.social> i think the opposite is true. if you have 30% hashrate / chance of winning a block, winning 2 consecutive blocks, would be 0.3^2, for 6, 0.3^6 and so on. so exponentionally harder with each block. if you have 720 blocks a day, it's therefore more likely to to eventually be able to strike, vs 360 blocks a day (= less frequent / slower blocks)

<g​ingeropolous:monero.social> yeah its a doozy to get your head around. searching around infinite space for a winning nonce. but last night it made sense to me . Take a 2 10 minute block vs 10 2 minute block analogy: The attacker's challenge is to secretly out-build the honest chain. It's vastly more difficult for them to get lucky enough to secretly find more than 20 blocks while the honest chain is finding 2<clipped

<g​ingeropolous:monero.social> 0, than it is for them to secretly find more than 2 blocks while the honest chain is finding 2.

<g​ingeropolous:monero.social> so how fast can we sling blocks around the network

tevador had mentioned that these aren't asics... but the bitmain X5 is still out there. And they could be the main thing on those rental sites (how else do you get 10 MH)

switching to RandomxV2 might go a long way with this situation

i was trying to find the code for that nonce distribution analysis to see how much of them are still on the network but couldn't find it

<j​ohnruth:matrix.org> eprint.iacr.org/2024/363.pdf

<j​ohnruth:matrix.org> arxiv.org/pdf/2502.17307

<j​ohnruth:matrix.org> jit.ndhu.edu.tw/article/viewFile/2543/2561

<j​ohnruth:matrix.org> oic-cert.org/en/journal/pdf/2/1/211.pdf

<j​ohnruth:matrix.org> scispace.com/pdf/blockchain-system-…for-double-spend-and-467y88mhpz.pdf

<j​ohnruth:matrix.org> ece.iit.edu/~yucheng/YCheng_PST22.pdf

<j​ohnruth:matrix.org> scitepress.org/Papers/2024/135245/135245.pdf

<c​aptaincanaryllc:matrix.org> I also want to know this

<b​oog900:monero.social> But you said it yourself, the probability of each individual reorg has dropped further than the increase chance you get because of more blocks.

<b​asses:matrix.org> idk if this was shared before

<b​asses:matrix.org> powerupprivacy.com/2025/03/05/moner…ty-one-percent-attack-analysis.html

<b​asses:matrix.org> calculates the approximate costs of 51% attack on Monero, research was published on March 05, 2025.

<b​asses:matrix.org> Rucknium

gingeropolous: mining rig rentals gives a description of the equipment that you are renting. I did not notice any bitmain X5

how many are left after so many burned up?

<r​ucknium:monero.social> captaincanaryllc gingeropolous See my rucknium.me/posts/monero-pool-transaction-delay

<r​ucknium:monero.social> >About 80 percent of blocks arrived at all five Monero nodes within a one-second interval.

<r​ucknium:monero.social> rando: Thanks. I had not seen that.

<b​arthman132:matrix.org> So around 12 to 18 million for one month to pull off. TBH that’s not very much in the grand scheme of things

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

<b​arthman132:matrix.org> Did qubic get ddosed?

<d​atahoarder:monero.social> They finished their marathon

<b​arthman132:matrix.org> Oh they did? Ok thanks

<d​atahoarder:monero.social> It happens for 24h in a row

<b​arthman132:matrix.org> Does their hashrate always drop to basically zero when they’re done?

when they are in their own mining phase they mine nothing in XMR

it comes in/out every hour for 30m

<l​ordx3nu:matrix.org> Is there any info showing how many blocks they found yesterday?

you can hover over the pie chart at the bottom of miningpoolstats.stream/monero

297 . dunno what the window is for the pie chart

<l​ordx3nu:matrix.org> I'm curious if they crossed 40%.

<r​ucknium:monero.social> Checked moneroconsensus.info lately? :D

gingeropolous: in the upper right it says last 1000 blocks

if you click that it will give you last 100 blocks

there yah go, so thats definitely covers the day

<r​ucknium:monero.social> Th line charts will show blocks on the main consensus chain only. That wouldn't include any honest blocks that got orphaned or any Qubic blocks that weren't broadcast because they lost the race.

<r​ucknium:monero.social> I mean, my line charts.

<b​arthman132:matrix.org> Ok it seems they’re starting back up again

yes, every hour, for 30 minutes

sort of, it's fuzzy based on their ticks

<b​arthman132:matrix.org> They had 3ghz all yesterday

<b​arthman132:matrix.org> They had 3gh/s all yesterday

<g​ingeropolous:monero.social> ok, so even with superfast blocks, the attacker could still make empty blocks, and make massive re-orgs occasionally

<b​arthman132:matrix.org> Honestly at this point we have to hope their halving on Wednesday saves us.

<l​ordx3nu:matrix.org> Yeah I guess I'm looking for blocks found during the start and end of their marathon. I know sech has posted his internal data before 👀

from what could see they crossed 39-40%

see I gather the data but I don't chart or process it in a proper way :D

<l​ordx3nu:matrix.org> Haha of course you do, you are datahoarder

<c​aptaincanaryllc:matrix.org> smaller blocks, fewer txs per block, same total confirmation lock time?

<c​aptaincanaryllc:matrix.org> responding to this

I love replies irc.gammaspectra.live/35f618ff56e832e6/image.png

sigh clipboard again

I should get around and make my new irc bridge proper for the rest of monero channels

it handles replies vastly better

<c​aptaincanaryllc:matrix.org> meaning assuming blocks half the current size (example), would require twice as many confirmations to unlock (same amount of time)?

<g​ingeropolous:monero.social> yeah. so instead of waiting for 10 blocks, you wait for 20

<c​aptaincanaryllc:matrix.org> 20 confirmations instead of 10. Wouldn't this make it more difficult (again) for adversary to reorg?

<g​ingeropolous:monero.social> yeah, afaiu

<g​ingeropolous:monero.social> same total proof of work performed, but the quicker blocks means the luck dragon has more chance to work with you (or against you)

<g​ingeropolous:monero.social> well in the case of like 30 second blocks, it would be 30 confirmations i guess

DataHoarder: yes please

<c​aptaincanaryllc:matrix.org> their chances of getting blocks goes down, and their difficulty of getting confirmations goes up

:D

the code is open source but it really isn't "production" ready. yet it has been working excellent over the past few months in p2pool channels

<g​ingeropolous:monero.social> i think the magic of the faster blocks is because the honest chain has a chance to find a new block faster. So the attacker has to always restart their attack to get their private chain longer

<s​pirobel:kernal.eu> + you get to keep the cpus.

<s​pirobel:kernal.eu> you dont lose anything

<s​pirobel:kernal.eu> in the case of pos the state is reversed and your stake is gone

<c​aptaincanaryllc:matrix.org> does anyone want to start testing feasible block size/speed changes? ofrnxmr

<o​frnxmr:xmr.mx> with fcmp its hard to create txs fast enough to fill blocks. i think, A faster block time would lead to the most well connected miners often being the first-seen block, and smaller miners being orphaned more often

<g​ingeropolous:monero.social> according to Rucknium 's work: "About 80 percent of blocks arrived at all five Monero nodes within a one-second interval."

<o​frnxmr:xmr.mx> Thats sub-300kb blocks

<o​frnxmr:xmr.mx> Every ~2mins

<a​lexanarcho:matrix.org> damn unknown at 45%+ right now

<o​frnxmr:monero.social> Meanwhile zcash has 75% on one pool

<e​longated:matrix.org> Honest pool they say 😅

<d​iego:cypherstack.com> An interesting attack to do on Zcash is if said big pool only put in non shielded stuff in the blocks forever.

<d​iego:cypherstack.com> Effectively making it Bitcoin.

<e​longated:matrix.org> Ztrash foundation probably owns that pool

<c​aptaincanaryllc:matrix.org> is it possible to add/send out padded sub-300kb blocks as fast as possible (at least 1 real tx per block) just to increase complexity for adversary, as well as the regular blocks at normal intervals

<e​longated:matrix.org> Is anyone working on checkpoints as a patch/poc?

<v​enture:monero.social> another avenue probably would be for everyone to adopt selfish mining as well and alter between mining secret side chain / withholding and broadcasting? fallacy of composition, if everyone selfish mines, nobody is :)

<v​enture:monero.social> *nobody gains an advantage

<n​oname-user0:matrix.org> it's already built in to the daemon. but issuing the checkpoint is done by core team manually right now with each wallet upgrade

<e​longated:matrix.org> So core just needs to update it every legit block?

or dns checkpoints

<n​eromonero1024:monero.social> Hey guys I want to float an idea. Would it be possible to integrate time as an additional blockweight factor into every block that a node sees. So for example, a node would calculate the weight of a block with some function: Weight=Difficulty+Time. And they'd calculate a chaintip weight as the sum of the weight of all blocks.

but that's opt-in

<n​eromonero1024:monero.social> This would imply a disadvantage for a 51% attacker, because now there's an additional time-weighting component that they have to fight. So hypothetically, they *could* still do long chain re-orgs, but they'd need more than 51% of the network hashpower to do it

<n​eromonero1024:monero.social> Again, each node would calculate their own view of the network independently, and choose to accept or reject a reorg, solely on the basis of their own calculation of relative weights between two competing chaintips

how do you know the proper time, except via the blockchain (which effectively provides an agreed time)

computer clocks are off hours sometimes

this is seen in p2pool where miners had random offsets on clocks

<s​yntheticbird:monero.social> yeah if monero had 10 minutes block time it would have been kind of interesting but many computers are off by multiple dozen of minutes and sometimes hours

<n​eromonero1024:monero.social> huh? I forwarded Bawdy's message from #monero:monero.social but it showed under my name

<n​eromonero1024:monero.social> the fk

there is no forwarding in IRC

<n​eromonero1024:monero.social> this is Bawdy's idea, not mine

<s​yntheticbird:monero.social> there is no forwarding in matrix either

<s​yntheticbird:monero.social> its a copy and paste shortcut

<n​eromonero1024:monero.social> oh

irc.gammaspectra.live/a77020023757948c/image.png

<n​eromonero1024:monero.social> XDDDDDDDDDDDDDD

I have seen a computer clock off by 2 years lol

The problem is how years and hours can sound similar, as are can and can't...

So... Instructions unclear, setting time off by 2 years...

<h​bs:matrix.org> Just off the press.... rekt.news/hashrate-heist-or-hype

<r​avfx:xmr.mx> Been more than one hour that qubic is it's offstate.

<r​avfx:xmr.mx> is in it's**

<r​ucknium:monero.social> {continued from MRL} I remember koe saying "[something] or get banned from this channel". Hard to search for "ban" in logs now because there has been so much talk about ban lists.

<r​ucknium:monero.social> Ah, here we are: libera.monerologs.net/monero-research-lab/20220617#c110125

<r​ucknium:monero.social> > control yourself or get banned from this channel

<r​ucknium:monero.social> > guess what? I checked logs just to see if you had anything constructive to say about the jamtis update. Instead I am confronted with this continuing garbage. Strike 1

<r​ucknium:monero.social> > There are no strict rules, but there is a line and you are way over the line.

<r​ucknium:monero.social> Ah, I really like this response from ooo123:

<r​ucknium:monero.social> > if it's possible to make a warn like "control yourself or get banned from this channel" and it's impossible to list all rules due to "There are no strict rules, but there is a line and you are way over the line." then it's called ping-pong where no one is responsible for everything. Please, prepare all rules the next time and warn me in case if I've been removed from your personal ignore lis

<r​ucknium:monero.social> Brings back memories 😂

<l​eonarth_:matrix.org> report commissioned by Qubic:

<l​eonarth_:matrix.org> shai-deshe.gitbook.io/parallel-thou…f-of-work/the-qubic-minority-report

<r​ucknium:monero.social> On a skim, seems to be a pretty objective analysis.

<r​ucknium:monero.social> > Given the data and analysis above, I can say with confidence that Qubic managed to obtain at least 28% of the current Monero hashrate. This is congruent with the 35% hashrate, as well as the perceived drop in average predicted hashrate due to increased orphan rates that follow from selfish mining.

<r​ucknium:monero.social> > Of the hashrate used by Qubic for selfish mining, at least 5% came from new sources of hash and not migrating Monero miners.

<r​ucknium:monero.social> > It is implausible that Qubic managed to obtain a majority of the actual hashrate, as applying Eyal-Sirer with a majority miner reduces to an indefinitely long double-spending attack.

<s​yntheticbird:monero.social> The interpretation and conclusion is pretty funny

There is no fixed rule set. It is impossible to create one in advance that predicts everything. It's like asking for before-the-fact overfitting of a data set that's not know prior. Basically, it's "don't be too much of an asshole, or you might get kicked". It is unfortunatly vague and varyig with time. If you're neurodivergent and can't deal with that vagueness, I'm sorry, but the best

thing that can be done is if you endeavour to stop when someone tells you're almost too much of an asshole.

(about rucknium's paste a few lines above)

<s​yntheticbird:monero.social> really reads like a "Qubic failed yet you guys panicked so monero failed"

<r​ucknium:monero.social> Right. I'm not sure ooo123 is in this room to hear that, but it would benefit him to take heed.

<0​xfffc:monero.social> any tl;dr

"Qubic failed yet you guys panicked so monero failed" <<>> Cat is aware and remains unconcerned

as usual I have little choice but to follow her lead

<0​xfffc:monero.social> time to buy some more XMR!

<n​eromonero1024:monero.social> Rucknium: (pardon the annoyance) forwarding from #monero:monero.social (idea by bawdyanarchist):

<n​eromonero1024:monero.social> Hey guys I want to float an idea. Would it be possible to integrate time as an additional blockweight factor into every block that a node sees. So for example, a node would calculate the weight of a block with some function: Weight=Difficulty+Time. And they'd calculate a chaintip weight as the sum of the weight of all blocks.

<n​eromonero1024:monero.social> This would imply a disadvantage for a 51% attacker, because now there's an additional time-weighting component that they have to fight. So hypothetically, they could still do long chain re-orgs, but they'd need more than 51% of the network hashpower to do it

<n​eromonero1024:monero.social> Again, each node would calculate their own view of the network independently, and choose to accept or reject a reorg, solely on the basis of their own calculation of relative weights between two competing chaintips

<n​eromonero1024:monero.social> it would only be the NTP time as seen by the local node.

<n​eromonero1024:monero.social> The idea would be "my node saw block {X} mined at NTP UTC {time}." Maybe the time-adjusted weighting of each block would follow a function with a horizontal asymptote for a "max possible weight" over time.

<a​rticmine:monero.social> Thank you for clarifying this. The real endgame with finality layers is a pure Proof of Stake solution.

<a​rticmine:monero.social> When it comes to honesty I strongly believe that late is always better than never. So again I thank you.

<s​yntheticbird:monero.social> ArticMine got kidnapped and replaced by a gov puppet confirmed

<s​gp_:monero.social> Spirobel has always been pro proof of stake in these discussions. Choosing a finality layer does not force a removal of pow later, that's just false. If pow is removed it'll be for its own merits not because something is an inevitable "endgame"

<r​ucknium:monero.social> neromonero1024: Thanks. Some variations of that have been proposed. The idea has pros and cons.

<s​gp_:monero.social> There definitely is an argument for PoS only, but I don't think most people here agree with it

<s​pirobel:kernal.eu> i am not pro proof of stake or pro proof of work I am pro logic and reason

<a​rticmine:monero.social> This endgame point was made in Zcash

<a​rticmine:monero.social> What happens if the finality layer is attacked?

<s​yntheticbird:monero.social> very logic and reasonable answer. Thx for your contribution spirobel

<s​gp_:monero.social> I did some napkin math on the security budget increase with more costly transactions. If the average cost was raised to $1 (it would take actve management to readjust to fix it at this price), and if the transaction volume remained unaffected, that's a roughly 20-25% increase in rewards and thus hashrate

<s​gp_:monero.social> Which is non-trivial but also probably doesn't get us totally in the clear

<r​ucknium:monero.social> IMHO, you don't necessarily need much active management. Just pin it to network difficulty and you would get a good moving approximation. But like I said in the meeting, the necessary magnitudes aren't there.

<r​ottenwheel:unredacted.org> nice. looking forward to dumping all XMR and passing all sites to the next guy. 👍

<r​ottenwheel:unredacted.org> all in on ARRR.

<s​yntheticbird:monero.social> rotten, spirobel has been spamming the same retarded, substance-less take for multiple days now. Please wait for actual researcher and devs to discuss this

<r​ottenwheel:unredacted.org> absolutely despicable. PoS in XMR.

<n​eromonero1024:monero.social> DERO FTW

<a​ntilt:we2.ee> BFT != PoS ... please, there is more to it tha stake

<s​yntheticbird:monero.social> Spirobel and the few newcomers agreeing with him aren't indicating the general temperature towards a PoS migration

<r​ottenwheel:unredacted.org> I'll become Reuben's next lost cousin.

<r​ottenwheel:unredacted.org> doesn't mean they aren't all considering it in the first place. again, disappointing.

<o​frnxmr:xmr.mx> Speaking of dero, any thoughts on blockdag like dero or kaspa

<s​yntheticbird:monero.social> well then get disappointed its only natural these points are being brought up into the discussion despite being of unreasonable effort/benefit ratio

<r​ottenwheel:unredacted.org> you don't tell me what to do, kitty cat.

<s​yntheticbird:monero.social> alright sorry rouepourrite

<s​pirobel:kernal.eu> dont know if this is meant to be sarcastic or not. I spelled out the logic very clearly and so far received zero convincing arguments against my points. The only article laying out the cost of an attack on monero says its 10-20 million to buy CPUs powerupprivacy.com/2025/03/05/moner…ty-one-percent-attack-analysis.html which is like buying a form of stake.

<r​ottenwheel:unredacted.org> as for spirobel's nonsense... we know since the monerochan sticker on a piece of paper in Afghanistan. 🤣

<l​ordx3nu:matrix.org> Wtf I love pos now

<r​ottenwheel:unredacted.org> "I have achieved more than the entire US DOJ!!!"

<r​ottenwheel:unredacted.org> yes!

<s​pirobel:kernal.eu> nah people agree with me. stop being an asshole. and cut your pseudo sarcastic bs. give a real counterargument or fuck off.

<s​pirobel:kernal.eu> its cheaper to attack bitcoin than ethereum

<s​pirobel:kernal.eu> if pos is the devil proof them wrong.

<h​bs:matrix.org> Then fees are in XMR unfortunately, not in USD, so anything equating 1 USD today may be really costly fiatwise in a few years. Even 1 USD is a no go for use as digital cash, way too high

<s​yntheticbird:monero.social> spirobel alts*

<s​pirobel:kernal.eu> honestly bird there is no reason to start chirping like this

<s​pirobel:kernal.eu> not cool

<d​iego:cypherstack.com> Actual counterargument to full PoS? Ok. I don't like how it embraces weak subjectivity and how entering the ecosystem essentially becomes a permissioned thing.

<d​iego:cypherstack.com> With PoS, there is no way to obtain the coin other than buying it from someone who already has it.

<s​yntheticbird:monero.social> All you had to do is not to copy paste your same brainless take 10 times in meeting

<s​pirobel:kernal.eu> yes you need to buy a stake in the cpu+electricity coin instead.

<d​iego:cypherstack.com> I think this is not just a slight net negative, but a large one. The weak subjectivity thing is objectively true in its introduction, but debatable by people as to whether it matters or not (I think it does)

<s​pirobel:kernal.eu> guys honestly you can hate on me, but i am just bringing the message

<d​iego:cypherstack.com> With PoW it can be mined. You can join a pool and contribute hashes and get some XMR. Not a ton, but some.

<s​gp_:monero.social> Widely available by design

<b​awdyanarchist:matrix.org> Damn guys. The title here is "Research Lounge." Not argumentative flaming lounge. Shouldn't we be a bit more netural in tone and stick to numbers, analysis, etc?

<d​iego:cypherstack.com> Don't know why you guys are egging each other on when most of you all are known to get trolly and pissy when there's an opinion you don't agree wiht.

<d​iego:cypherstack.com> What I said, but nicer, yeah

<d​iego:cypherstack.com> Don't know why you guys are egging each other on when most of you all are known to get trolly and pissy when there's an opinion you don't agree with.

<s​yntheticbird:monero.social> ok whatever, let the endgame boy spam his take full of numbers and analysis

What do you mean by "weak subjectivity" ?

<s​pirobel:kernal.eu> guys read the article. you can keep the cpus after the attack and try again and again just like qubic is doing now. if governments really wanted to destroy us they cut do this practically for free. they even get to keep the cpus. bitcoin could be attacked as well. the era of pow is coming to a close and you can either accept it or keep insulting me. i dont care. its really not my problem.

<s​pirobel:kernal.eu> let the bird chirp

<d​iego:cypherstack.com> moneromooo: blog.ethereum.org/2014/11/25/proof-…take-learned-love-weak-subjectivity

<s​yntheticbird:monero.social> someone look at me in the eyes and tell me this is an argument

<s​pirobel:kernal.eu> zero substance response bird

<d​iego:cypherstack.com> basically there's a minor element of trust with PoS that is not present in PoW, since PoW has an objective metric (most hashes)

<a​rticmine:monero.social> Check the last Monero Research Lab meeting minutes.

<a​rticmine:monero.social> In my view hidden agendas have no place in a technical research discussion. When this occurs when Monero is under attack, I believe it is responsible for me to call this out.

<a​rticmine:monero.social> The Qubic attack has magnified my concern

<d​iego:cypherstack.com> While I agree with this, fwiw, I don't share the sentiment that a finality layer necessarily errs towards full blown PoS (which I don't care to see in Monero)

<d​iego:cypherstack.com> that said, I acknowledge that as an opinion, that people are free to agree or disagree with

<s​pirobel:kernal.eu> i dont see how pow is fundamentally different than pos. you buy a stake of about 20 million worth of amd cpus and then you get to own the network. So pow is just a very ineffective and costly proof of stake network.

<s​yntheticbird:monero.social> The difference being that most people even poor people can access the network

<s​yntheticbird:monero.social> PoS block that

<s​yntheticbird:monero.social> PoS raise an exagerate minimum floor of investement to participate in the network making it permissioned

<s​pirobel:kernal.eu> poor people that have amd cpus. give me a break

<s​yntheticbird:monero.social> Again, someone tell me this is an argument

<d​iego:cypherstack.com> From a social and economic game theory perspective, perhaps. But from a BFT perspective they are different.

<d​iego:cypherstack.com> PoW has an objective way for nodes and miners to decide the validity of competing chains

<a​ntilt:we2.ee> btw >Given the data and analysis above, I can say with confidence that Qubic managed to obtain at least 28% of the current Monero hashrate. shai-deshe.gitbook.io/parallel-thou…f-of-work/the-qubic-minority-report

<d​iego:cypherstack.com> Blockchain is NOT just economics. It's NOT just tech. It's NOT just cryptography. It's NOT just game theory. It's all of these things together. And more complicated, it's how all of these things interact.

<d​iego:cypherstack.com> Comments like this are extremely reductive to the complexities, and honestly shows a lack of holistic understanding about how blockchains work.

<d​iego:cypherstack.com> In other words, I agree with your assessment from an economic game theory perspective, spirobel. But disagree from a BFT (tech) perspective

<s​yntheticbird:monero.social> Welcome, take a seat and a bag of popcorn, this has been the level of take spirobel has spammed in yesterday meeting

<d​iego:cypherstack.com> It's ok, I'm smart enough for all of us.

<d​iego:cypherstack.com> (I say this in jest, before people get to thinking I'm full of myself :P)

<a​rticmine:monero.social> It is because of borrowing a stake at effectively no cost. One cannot borrow CPUs at scale for free.

<a​rticmine:monero.social> My biggest concern with POS is staking by the likes of:

<a​rticmine:monero.social> 1) Pirate at40

<a​rticmine:monero.social> 2) MtGox

<a​rticmine:monero.social> 3) QuadrigaCX

<a​rticmine:monero.social> 4) FTX

<a​rticmine:monero.social> ....

<n​eromonero1024:monero.social> spirobel: just for the record, man, are you actually serious about PoS? or just joking?

<s​pirobel:kernal.eu> to be clear: there is no hidden agenda. Its just the result of thinking this through to the end. if the cost of attack currently on monero is 20 million and you get to keep the cpus after the attack. then whats the point? powerupprivacy.com/2025/03/05/moner…ty-one-percent-attack-analysis.html proof of stake attack would cost 100 million on the low end in stake and the <clipped message>

<s​pirobel:kernal.eu> stake would get slashed once the network is restarted. with pow you get to keep the CPUS and you can try again and again until eternity. so why would we keep this meaningless ritual around that just adds sell pressure and makes finality take longer?

<s​pirobel:kernal.eu> Comments like this are extremely reductive to the complexities, and honestly shows a lack of holistic understanding about how blockchains work.

<s​yntheticbird:monero.social> wrong clipboard

<s​yntheticbird:monero.social> i hope

<d​iego:cypherstack.com> I'm not saying you don't know about blockchains. I'm dating your reductive comments aren't doing you any favors here.

<s​pirobel:kernal.eu> where can you borrow monero for free? CEX barely have any monero and they would never partake in an attack on the system. and if they did: just restart the network and slash their stake, whats the problem? again: you can take away the stake, but you can take away the cpus after a hypothetical attack happened

<d​iego:cypherstack.com> > <@spirobel:kernal.eu> Comments like this are extremely reductive to the complexities, and honestly shows a lack of holistic understanding about how blockchains work.

<d​iego:cypherstack.com> I'm not saying you don't know about blockchains. I'm saying your reductive comments aren't doing you any favors here.

<s​gp_:monero.social> Issuing a reward to stakers gives XMR holders opportunity cost for not staking fwiw. It can discourage the use of a custodian

<d​iego:cypherstack.com> Fwiw, I think PoS proves itself more every day with an actual big coin like Ethereum using it. One of the primary criticisms of years past was things weren't fully understood about the attack surface because of its newness

<o​frnxmr:xmr.mx> Spirobel is as serious as serious can get

<s​pirobel:kernal.eu> I am willing to change my mind if there is a good argument. I am not pro or against anything.

<d​iego:cypherstack.com> I gave one that went unanswered?

<d​iego:cypherstack.com> I don't like the idea of subjectivity (even weak) in Monero. I don't think that it part of the ethos

<o​frnxmr:xmr.mx> I never said pro or con, i said serious, as in, not joking/trolling

<s​pirobel:kernal.eu> ???

<c​haser:monero.social> spirobel: on what weak subjectivity is in proof of stake: blog.ethereum.org/2014/11/25/proof-…take-learned-love-weak-subjectivity

<c​haser:monero.social> Diego Salazar: it's very counterintuitive, but proof of work also requires weak subjectivity. I recommend this as an enlightening read on the topic: forum.celestia.org/t/nakamoto-conse…al-coordination-and-subjectivity/28

<d​iego:cypherstack.com> Aha! A chance to expand my knowledge. Thanks, will read.

<s​pirobel:kernal.eu> i am sorry but this weak subjectivity sounds like a word that vitalik came up with while he was on the toilet.

<s​pirobel:kernal.eu> the game theory behind PoS is stronger than PoW

<s​yntheticbird:monero.social> I JUST CAN'T make the lack of seriousness of this guy more explicit

<b​awdyanarchist:matrix.org> I am a bit new to the discussion in the forum here but are we discussion large scale long term consensus changes for new blocks by moving to a staking mechanism? Or are we only discussing the finality mechanism?

<b​awdyanarchist:matrix.org> It seems that, hypothetically, a finality mechanism could just be basically an addon. If the PoS finality mechanism fails, then we just end up where we already are - with no finality, but still a functional chain. Perhaps we could use the 10-block lock as a fallback in the case of delayed finality?

<c​haser:monero.social> spirobel: Vitalik came up with a *lot* of the game theory behind modern proof-of-stake consensus. and I can assure you that weak subjectivity is a real thing.

<s​yntheticbird:monero.social> Please note this discussion outside of meeting and is therefore unlikely to lead to any development decision. But yeah right now people are discussing non-PoW consensus mechanism

<s​gp_:monero.social> Gridlocked finality would basically require a hard fork to make the network functional again

<s​gp_:monero.social> Sure, people could accept non-finalized payments but they shouldn't. That would be very risky

has the report been posted? shai-deshe.gitbook.io/parallel-thou…f-of-work/the-qubic-minority-report

<s​yntheticbird:monero.social> DataHoarder yes

<s​yntheticbird:monero.social> also in #monero-research-lab

<s​pirobel:kernal.eu> also consider the other part of the equation: you can just get a loan. perform the attack and sell the cpus again. the cpus dont lose their value. or even better: make up some mlm ponzi about mining agi until 2027 with cpus (in the ai garden called aigarth) and get lots of suckers to max out their credit cards or use buy now pay later

<d​iego:cypherstack.com> I actually disagree with this. It's a semantic thing.

<d​iego:cypherstack.com> I especially disagree with the idea that PoW subjectivity is no less subjective than weak subjectivity.

<d​iego:cypherstack.com> A chain may not know if the chain is the real chain unless they ask 100% of peers (functionally impossible), but they can always objectively reconcile known chains.

<d​iego:cypherstack.com> I think a PoS proponent is better served by arguing that the subjectivity in PoS is so weak as to functionally not matter rather than try to convince that PoW is also weakly subjective to the same degree.

<s​pirobel:kernal.eu> i prefer game theory over philosophy lessons. I saw no counter argument on game theory. that is the only thing that matters when push comes to shove

<s​pirobel:kernal.eu> put a clear dollar amount on the attack costs. the only article we have so far just lays out the cost of CPUs ( that can be borrowed and resold)

<s​yntheticbird:monero.social> Red Herring

<s​gp_:monero.social> spirobel you're being purposefully dense. If an exchange has >34% of the staked assets in PoS, they can halt the network at ***no*** direct cost. That's a clear regression

<n​eromonero1024:monero.social> * PoS makes the network permissioned... no one can get XMR without buying XMR from someone else

<n​eromonero1024:monero.social> * rich gets richer without any cost (compare cost of hosting a node vs mining)

<n​eromonero1024:monero.social> * eventually, they will control vast majority of the coin supply... makes government's job much easier

<n​eromonero1024:monero.social> * Monero's emission curve wasn't the most fair... I'd imagine coin distribution isn't much "decentralized" either

<n​eromonero1024:monero.social> * if the government greenlights such a project, if it comes out, it will be a big publicity issue... another deterrent

<n​eromonero1024:monero.social> IMO, governments are better off trying to fk with projects like Monero through regulations ("ThInK oF tHe ChIlDrEn") rather than directly attacking by buying/renting $100 million of AMD CPUs

<s​pirobel:kernal.eu> sgp_: of course. the cost would be them getting their stake slashed after the network restarts. its just a liveness problem that can be fixed with them losing 100% of their capital

<s​pirobel:kernal.eu> we cant do that with qubics cpus

<d​iego:cypherstack.com> I'm actually kind of speechless that you think objectivity vs subjectivity is philosophy.

<d​iego:cypherstack.com> Regardless, I gave a good reason to not go full PoS, you don't think its a good reason (or even acknowledge it as A reason), so there's no reason to converse further on it as we disagree here fundamentally.

<s​pirobel:kernal.eu> thats the difference

<s​gp_:monero.social> requiring hard forks to get the network back on track isn't ideal, even if it is a remedy. It's centralizing, messy, and still results in significant downtime, distrust, etc. Clearly I'm not totally against PoS in all forms, but I know it's not black and white, good vs bad

<s​pirobel:kernal.eu> i have all kinds of deep thoughts on philosophy I enjoy reading Nietzsche, Hegel etc ... (I am German after all) but I don't think this is at all relevant to the conversation of PoS vs PoW that is just about game theory and which one is more secure

<b​arthman132:matrix.org> Monero being made illegal is unlikely at this point. I did think it was more likely a couple of years when it was still relatively new, but it just seems their strategy is just to pressure exchanges to delist it.

<v​enture:monero.social> THAT reminds me. your extreme standpoint caters to the hegelian dialect. problem, solution, synthesis

<w​eaverethan:matrix.org> until thats not working anymore

<v​enture:monero.social> THAT reminds me. your extreme standpoint caters to the hegelian dialect. problem, over-reaction, synthesis

<b​arthman132:matrix.org> It’s a risk you guys have to take would you rather lose to qubic or potentially a powerful government

<s​yntheticbird:monero.social> DANCE

<s​yntheticbird:monero.social> DANCE spirobel

<s​yntheticbird:monero.social> the three typing notifications must appear at the same time

<w​eaverethan:matrix.org> if they ban it so what?

<s​pirobel:kernal.eu> sgp_: we have pos already. the stake is defined in how many CPUs you own. Its relatively cheap to acquire (just 20 million to control the stake for a 5 billion dollar market cap project). there is also the downside that the stake cant be slashed or forked away from should the owners of this stake misbehave.

<w​eaverethan:matrix.org> cia will always have a use case for monero lol

<o​frnxmr:xmr.mx> At 2.5gh, ryzen 3900 powered @1k per rig a coupke yrs ago, i had it ar 140m for the full nethash. Not 20m

<o​frnxmr:xmr.mx> Btc at 16b for hardware costs at 215eh (at the time)

<b​arthman132:matrix.org> Exactly the network can still be bought out with pow. It’s just that POS is more direct

<s​yntheticbird:monero.social> matrix.monero.social/_matrix/media/…ero.social/yQbBMAMfxlPXgqmrAzJItnHF

<s​pirobel:kernal.eu> monerobull agreed with me too yesterday

<o​frnxmr:xmr.mx> For servers. It was 215m for the hardware

<s​yntheticbird:monero.social> nah they agreed on specific point sir

<s​pirobel:kernal.eu> source for the number was this article powerupprivacy.com/2025/03/05/moner…ty-one-percent-attack-analysis.html

<l​eonarth_:matrix.org> qubic's not mining anymore

<b​arthman132:matrix.org> do you have a better solution for monero problems lately?

<o​frnxmr:xmr.mx> That doesnt include mobo, power supple, etc?

<b​arthman132:matrix.org> I saw that they likely just taking a break I guess, they did that earlier today

<o​frnxmr:xmr.mx> Just straight up cpus?

20:43:53 <m-relay> <l​eonarth_:matrix.org> qubic's not mining anymore

they got stick

stuck* ticking

<b​arthman132:matrix.org> Wait really? What happened to them? How did they get stuck

they are trying to solve via consensus via discord by issuing manual votes (and their central arbitrator might need to remove computors)

<o​frnxmr:xmr.mx> ```

<o​frnxmr:xmr.mx> Hashrate per unit: 98.24 KH/s

<o​frnxmr:xmr.mx> Price per unit: $528

<o​frnxmr:xmr.mx> Units needed for 51% attack: 25,043

<o​frnxmr:xmr.mx> Total hardware cost (CPUs): $13.22 million

<o​frnxmr:xmr.mx> Power consumption (at 225W per CPU): 5.63 megawatts

<o​frnxmr:xmr.mx> ```

<o​frnxmr:xmr.mx> I dont think this includes anything but the cpu..

<b​arthman132:matrix.org> Wonderful news

3 hours ago they went down while their nodes still are active

<o​frnxmr:xmr.mx> Rhats supposed to be `CPU Option (AMD EPYC 7B12)`

<l​eonarth_:matrix.org> what's 13mil for a 3-letter, they spend that on a weekend

they just can't get out to the next ticks every so often, meaning they don't switch mining phases

<o​frnxmr:xmr.mx> 100kh for $500? Sound wrong

no transactions either, as their consensus is stuck

<b​awdyanarchist:matrix.org> Just a little tidbit for those who believe that PoS is inherently "rich get richer." Well designed PoS systems (like Eth) are provably and robustly fair. I'm not advocating block production move to a PoS mechanism. But I would like to dispense with the false argument that all PoS systems inherently cause centralization of supply.

<b​awdyanarchist:matrix.org> arxiv.org/pdf/2103.14713

<o​frnxmr:xmr.mx> is is so rubust, that you need 140k to be a validator

<b​arthman132:matrix.org> Great news anything to slow them down is always welcome

<o​frnxmr:xmr.mx> Eth* is

<o​frnxmr:xmr.mx> And if you dont have 140k, you can join a centralized staking pool on binance, and pay them a percent of your coins so they can further try to own the market

<w​eaverethan:matrix.org> PoW + federated signatures?

<o​frnxmr:xmr.mx> If you dont have 32 eth, and (iirc) eth validator hardware is higher spec than my mining rigs

<s​yntheticbird:monero.social> federated?

<w​eaverethan:matrix.org> A small set of known, widely trusted entities co-sign blocks. Even if hashpower is captured, attacker can’t rewrite without those keys.

<w​eaverethan:matrix.org> sense POW and POS can both be bought out

<w​eaverethan:matrix.org> with low market cap its always a risk

<b​oog900:monero.social> you can do that a better way by having a max reorg depth with trusted nodes that can handle anyone who tries to split the network by publishing a chain right on the edge of the boundary

<b​oog900:monero.social> that way the trusted nodes do nothing except for if there is an attack

<b​awdyanarchist:matrix.org> In the sense that if you own X% of the stake, you can expect to get X% of rewards over long timeframes, yes Eth is provably fair. Again, I'm not saying that we should copy their model, or change block production to match Eth. What I am saying is that a PoS layer very likely *can* be designed such that supply doesnt centralize over time.

<o​frnxmr:xmr.mx> Rewards != fair

<o​frnxmr:xmr.mx> You srent the validator , binance is

<n​eromonero1024:monero.social> what you're saying is, basically, if someone stakes 1% of total coins staked, he'll get 1% of the all block rewards

<n​eromonero1024:monero.social> basically, in PoW terms, you'll get paid according to your hash rate

<n​eromonero1024:monero.social> I use "rich getting richer" in the context that the rich can earn significant portion of stakes, thus keep significant portion of the block rewards for themselves

<n​eromonero1024:monero.social> same also goes for PoW but the added friction of infrastructure + maintenance acts as a deterrent

<o​frnxmr:xmr.mx> The stake is custodied by binance

<o​frnxmr:xmr.mx> Just like zano users coins were staked by tradeogre, you might get rewards, but the stake is not yours

<s​pirobel:kernal.eu> the rewards should be fair. tail emission is there to offset the cost of the miners. proof of stake does not have this cost. the transaction fees are enough of a reward. inflation has to be zero. There will only ever be 18 million monero.

<b​awdyanarchist:matrix.org> The important point is that, a large staked amount doesnt end up gaining a higher percentage of the staked supply than what they started with; thus, it doesnt necessarily have to be inherently centralizing.

<n​eromonero1024:monero.social> I get the "provably fair" part but not sure I'm getting the "a large staked amount doesnt end up gaining a higher percentage of the staked supply than what they started with" part

<n​eromonero1024:monero.social> let's try with this hypothetical scenario

<n​eromonero1024:monero.social> let's say I stake 1% of total coins staked, I get rewarded 1% of the blocks produced

<n​eromonero1024:monero.social> let's say the block reward accounts for 0.00001% of total coins staked

<n​eromonero1024:monero.social> now, I have 1.00001% stake

<n​eromonero1024:monero.social> continue the cycle, now you get "rich gets richer"

<b​awdyanarchist:matrix.org> Just read the paper

<a​rticmine:monero.social> Everyone of the examples I cited borrowed Bitcoin. for free. Had Bitcoin been POS serious attack could have been possible against Bitcoin.

<a​rticmine:monero.social> Slashing rules can be worked around by a determined attacker particularly one that is desperate

<j​effro256:monero.social> I will say though, rich people *can* usually afford to have a higher percent of their total capital locked up in non-liquid activities like PoS staking. So while the ROI% is relatively equal for all stakers regardless of their size of stake, a rich person can afford to put up more stake as a % of their wealth than a poor person. As a (not so) extreme example, someone on the povert<clipped mes

<j​effro256:monero.social> y line living paycheck to paycheck has 0 long-term assets, and thus 0 stake. Someone making $70,000 a year with house payments will be able to stake some, but not as much as someone making $70,000 with no house payments

<j​effro256:monero.social> Of course, there's also the argument that rich people can afford to put a larger portion of their capital into mining farms, so there is an analogue for PoW, but I'm trying to make the point that it can also be centralizing

<o​frnxmr:monero.social> Also that rich people dont have to use their own $ for buyinf monero

<o​frnxmr:monero.social> They can borrow against other assets as long as the interest is less than the potential return on the xmr+stake

<j​effro256:monero.social> You could also borrow against existing assets to buy mining hardware, but that's inherently a depreciating asset

<o​frnxmr:monero.social> renting hashrate is a burn, since you can only recover the xmr produced

<j​effro256:monero.social> Whereas with PoS, you would be borrowing XMR against other assets to earn a ROI% on that XMR, which itself may rise in value

<j​effro256:monero.social> But both of these are only theoretical AFAIK since banks don't like the risk appetite of crypto at the moment, especially with new bonkers AML BS being pushed nowadays

<t​esttank:matrix.org> The argument Pow vs Pos sound like: let people from third world mine monero or make it so that Monero can’t be attacked by a random shitcoin.

<b​awdyanarchist:matrix.org> Borrowing implies that you'll be paying a percentage on what you borrowed. So that already starts to offset your potential profit. Typically borrowing percentages are actually quite high. Much higher than I imagine any staking system would pay.

<s​pirobel:kernal.eu> ArticMine: some even held large amount of pos coins for free. did they do an attack against these systems? would it make sense under any circumstances to do such an attack? no. slashing cant be worked around. Because the network can continue without their stake. If they print themselves money in a double spend they can have fun on their version of the world while the rest moves o<clipped message>

<s​pirobel:kernal.eu> n with 100% of their capital slashed. again we cant do this to qubics cpus. they can try again and again

<s​pirobel:kernal.eu> jeffro256: we also saw this with the weak subjectivity argument earlier. these arguments can be made both ways. the reason for that is that pow is inherently pos. the stake in the case of pow is the amount of cpus you bought. (the downside is that those cant be slashed / removed in a fork if this actor misbehaves. the stake is also very low compared to the marketcap of the coin.)

<o​frnxmr:monero.social> There is a difference between active cost of production, vs doing nothing

<o​frnxmr:monero.social> Pow isnt simply "buy rig". You have to run the rig. Pos, you dont even have to be online. You can just delegate your $ to someone else (like a cex)

<o​frnxmr:monero.social> Move $ from bank to exchange, check back in 4 years -> profit

<s​pirobel:kernal.eu> meanwhile: proposal to rent hashrate with gf funds

<o​frnxmr:monero.social> i wonder how some of these merchants can offer over 200mh total

<j​effro256:monero.social> It depends on the current state of the economy and if you can convince the creditor that staking XMR is low risk. Right now the borrowing rates from anyone would almost certainly be higher than staking rates, but if a future where XMR gets bigger adoption, and perhaps in a highly supply-side-inflationary economy, this might not be true

<j​effro256:monero.social> But big institutions and "rich" people will always get lower rates for any loans

<s​pirobel:kernal.eu> ofrnxmr: if there is no inflation you would have to be online. if there is no nominal inflation incentive there is no reason to delegate which is what we should aim for. (also for memetic reasons to piss of the bitcoiners)

<s​pirobel:kernal.eu> i would argue no inflation pos is better for the pow coin with inflation. inflation hits the lower classer more

<s​pirobel:kernal.eu> i would argue no inflation pos is better for the pow coin with inflation. inflation hits the lower classes more

<t​esttank:matrix.org> You could make the argument that POS gives more chance to the little guy to defend the network, given that it’s much less resource intensive

<s​pirobel:kernal.eu> *i would argue no inflation pos is better for the poor than pow coin with inflation. inflation hits the lower classes more

<o​frnxmr:xmr.mx> It doesnt though

<j​effro256:monero.social> The reason to delegate is so that you don't get slashed for going offline

<o​frnxmr:xmr.mx> Resource intensiveness is why bitcoin isnt mined by microstrategt

<o​frnxmr:xmr.mx> microstrategy would likely be btc's largest staker. US government would stake as well

The current block subsidy of <1% per year doesn't even compensate for lost coins, so there is no inflation in practice.

<s​pirobel:kernal.eu> there are protocol without active slashing. multiple options have been mentioned in the github issue. this was the paper mentioned as a preference eprint.iacr.org/2024/677

<s​pirobel:kernal.eu> there are protocols without active slashing. multiple options have been mentioned in the github issue. this was the paper mentioned as a preference eprint.iacr.org/2024/677

<h​ardenedsteel:monero.social> did u guys consider how community would react to the solutions purposed? PoS and ASIC way?

<b​awdyanarchist:matrix.org> At best you might be able to posit that, in the event of truly large scale adoption, the cost of borrowing will eventually converge as equal to staking revenue. So a zero profit proposition. Until then, in pretty much all cases, borrowing percentages significantly exceed anything that would be proposed as a reasonable staking revenue.

<h​ardenedsteel:monero.social> shai-deshe.gitbook.io/parallel-thou…eport#interpretation-and-conclusion

<s​pirobel:kernal.eu> any staking revenue would be nominal in any case. thats why I argue for no inflation. Inflation was a justification to cover miner costs. But that argument does not apply for stakers. They get the utility of providing security for the network. That should be enough. We see people mining at a loss to protect monero from qubic. That is like proof of stake where the staker has to pay<clipped message>

<s​pirobel:kernal.eu> to provide security.

<d​iego:cypherstack.com> Nice was looking forward to this

<s​pirobel:kernal.eu> asics are like permanently locked stake. it can be resold but it loses all its value once all the networks using its hashing algorithm lose their value. If there is a major draw down in bitcoin / as halvings progress and miners give up the price of asics tends to zero. at that point there is just the electricity cost left. even now older less efficient miners are much cheaper than<clipped message>

<s​pirobel:kernal.eu> the new ones but still provide a lot of hash rate. the efficiency does not matter for a 51% attack as much as it does for an honest miner bound by electricity costs. it is inevitable that over time this flips over as the costs to acquire enough asics to perform the attack decline.

<s​pirobel:kernal.eu> we will see bitcoin getting into this situation in the next 5-10 years max

<s​pirobel:kernal.eu> asics are not an option

<d​iego:cypherstack.com> Agree

Reacting to "The Qubic Minority Report": I fail to see how being ASIC friendly would prevent this attack. The security budget would be the same.

<s​pirobel:kernal.eu> that means there is only one other option left. I was a proof of work maxi btw: xcancel.com/spirobel/status/1541794519406022658 no shame in changing an opinion after learning more about something.

<h​ardenedsteel:monero.social> bold claim, would like to learn more

<s​pirobel:kernal.eu> even right now. if you look at the math of buying older asics you could pull a qubic on bitcoin for a few billion dollar.

<s​pirobel:kernal.eu> only getting worse

<d​uggavo:matrix.org> Does that Shai guy still have any credibility?

<t​esttank:matrix.org> Asics aren’t as easy to rent as CPU’s + if you have an Asic to mine a specific crypto it means you have bigger investment tied to it

<h​ardenedsteel:monero.social> no one has infinite money, the cost is the difference

this is what happens when """centralized network""" goes wrong, they stop ticking irc.gammaspectra.live/da780b2686b9a7ae/image.png

they ended up asking for votes to skip ticks or resync via discord

<t​esttank:matrix.org> So you wouldn’t want to hurt the network

<t​esttank:matrix.org> (I’m not saying im for asics)

<s​pirobel:kernal.eu> it is a miniscule amount compared to the market cap of bitcoin. if you find enough places to short it without getting detected it could be profitable very soon. we can do the napkin math together

<s​pirobel:kernal.eu> price per TH for the older once is 3 dollar

<s​pirobel:kernal.eu> price per TH for the older ones is 3 dollar

<s​pirobel:kernal.eu> bitcoin hashrate is one billion terra hash

<s​pirobel:kernal.eu> so 3 billion for the old miners to get this entire hashrate

<s​pirobel:kernal.eu> eth is more expensive to attack

<t​esttank:matrix.org> I think most people in the community want the “big boys”, the devs, to have an open discussion about what would be the benefits/drawbacks of POS without any prejudice

<a​sdfqwfe:matrix.org> Very interesting point.

<a​sdfqwfe:matrix.org> So for $3B you could buy enough ASICs, to attack Bitcoin... But how much would the electricity cost?

<s​pirobel:kernal.eu> here is a paper that looks at this in detail if you want to do more than napkin math papers.ssrn.com/sol3/papers.cfm?abstract_id=4727999#maincontent

<o​frnxmr:xmr.mx> 140b

<o​frnxmr:xmr.mx> In 2023

<o​frnxmr:xmr.mx> 16b**

<s​pirobel:kernal.eu> this number is only going to drop as halvings continue and old miners hit the shelves.

<a​sdfqwfe:matrix.org> One problem with ASICs is that you have to actually buy them. Easy for governments to track, ban, imprison.

<a​sdfqwfe:matrix.org> Deal breaker. Hard pass.

<a​sdfqwfe:matrix.org> One problem with ASICs is that you have to actually buy/ship/store them. Easy for governments to track, ban, imprison.

<a​sdfqwfe:matrix.org> Deal breaker. Hard pass.

<h​ardenedsteel:monero.social> there are many governments, bitcoin mined used to be in china

<m​onero.arbo:matrix.org> that's why my suggestion is to 50% merge with something like LTC that already has widely distributed ASICs, plus you're still not entirely subject to LTC's security

<m​onero.arbo:matrix.org> plus with Doge being merged, LTC effectively has tail emission

<a​sdfqwfe:matrix.org> Where can I read more about merging?

<m​onero.arbo:matrix.org> for basics you can check out: gate.com/learn/articles/merged-mining/858

<m​onero.arbo:matrix.org> You could also look into Namecoin, Dogecoin, or Tari (merged with XMR) if you want examples

<m​onero.arbo:matrix.org> Tari has split mining (rn 33% merged with XMR, 33% GPU/ASIC, 33% independent Randomx) like I'm suggesting as well. Each chain has a 6 minute target block time and independent difficulty, so overall block target remains 2 minutes

<s​pirobel:kernal.eu> merge mining with doge to save pow

<m​onero.arbo:matrix.org> it's merge mining with LTC. Doge is just already merged with LTC

<a​sdfqwfe:matrix.org> Is there a way I can export the contents of this chat so that I can command-F/search for stuff?

<m​onero.arbo:matrix.org> so effectively you would get all three coins from scrypt mining

logs of this chat are kept from IRC. I think there's an archive page

<a​sdfqwfe:matrix.org> I don't want to ask redundant questions 😇

<a​sdfqwfe:matrix.org> Thank you data hoarder! Very appropriate name lol.

for once it's not mine!

<a​sdfqwfe:matrix.org> What about using Eth as a finality layer as a quick fix?

we are looking for a well thought out fix

all ideas welcome

<a​sdfqwfe:matrix.org> Are there any good resources describing how similar contentious decisions have been made?

<3​21bob321:monero.social> Maybe pin Luke's github issue they made in the room

<3​21bob321:monero.social> Which has the options

<a​sdfqwfe:matrix.org> Monero isnt a democracy. It's not voting of shareholders.

<a​sdfqwfe:matrix.org> It sounds like governance by shouting contest, but that's a too uncharitable.

<a​sdfqwfe:matrix.org> Monero isnt a democracy. It's not voting of shareholders.

<a​sdfqwfe:matrix.org> It sounds like governance by shouting contest, but that's too uncharitable.

<m​onero.arbo:matrix.org> what we're looking for is *ahem* "rough consensus"

<m​onero.arbo:matrix.org> whatever faction can make their diea happen without a chain split (:

<m​onero.arbo:matrix.org> whatever faction can make their idea happen without a chain split (:

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

<a​sdfqwfe:matrix.org> # the other consensus problem/question

<c​howbungaman:matrix.org> Would any monero sages in this room be down to jump on tomorrow’s MoneroTopia to discuss all the options 🙏 Similar to what Luke did last week. Obviously we will open it up to everyone but would nice to have someone with a good handle on all the options being discussed to lead the discussion 🙏

<a​sdfqwfe:matrix.org> Aka freedom of association. Anyone can split the chain, but nobody has to follow you.

<3​21bob321:monero.social> leetarxiv.substack.com/p/smart-attack-on-elliptic-curves-for

<3​21bob321:monero.social> "Smart Attack on Elliptic Curves for Programmers"

<h​bs:matrix.org> I hear people at OFAC celebrating something....

<k​ayabanerve:matrix.org> rottenwheel: Sorry, I'm only just now poking around. May I ask for a brief list of your issues? I understand if you outright refuse PoS regardless but I do want to see there's a conversation which ends with you not absolutely leaving due to any notion of 'PoS'.

<k​ayabanerve:matrix.org> So far, reading up, I'm seeing (in general):

<k​ayabanerve:matrix.org> - Can't validate without buying XMR to join the network (can't sync the network without connecting to someone with the blockchain, either, but okay)

<k​ayabanerve:matrix.org> - 'Rich gets richer', yep, people who have servers earn revenue

<k​ayabanerve:matrix.org> - Centralizing to wealth. We are already at tail emission?

<k​ayabanerve:matrix.org> I'd also like to note we'd be targeting a minumum of hundreds of nodes, runnable at home, behind Tor.

<d​ezinfik:matrix.org> i agree with kayabanerve

<d​ezinfik:matrix.org> we should do it asap

<k​ayabanerve:matrix.org> (And as a finality layer under my advocacy, so we still have a PoW-method of advancing the chain and earning emissions)

ASAP means probably 2 years

<k​ayabanerve:matrix.org> Also, I know you mentioned Piratechain as a joke, yet Komodo's whole pitch was how they're a pseudo-PoS finality layer thanks to Bitcoin

<d​ezinfik:matrix.org> monero can no longer rely solely on the idea of being a people’s coin maintained only through decentralized mining by everyday users. it has been attacked, with coordinated attempts such as a "51% attack" and block reorganizations, showing that there are actors with significant resources willing to take control. if nothing is done, we risk losing security and true decentralizati<clipped message

<d​ezinfik:matrix.org> on. a new approach is needed to reduce the possibility of centralization while preserving privacy and network resilience

<r​ottenwheel:unredacted.org> it's simply antagonistic to whatever monero stands for. leave PoS for EVM chains. use tevador's bandwidth, make it mineable on raspis only, I don't know. anything but PoS. this is just me though. I hold -999 power in here. it could go full blown PoS if it were up to my decision, as I do not sway anyone one way or the other.

<k​ayabanerve:matrix.org> (where Piratechain is a Komodo-ecosystem project and claims to be invincible to reorganizations due to using Litecoin as a finality layer, though I'm unsure how practical that is)

<k​ayabanerve:matrix.org> *how true that is

<r​ottenwheel:unredacted.org> rolling XMR out on PoS is like saying, let's merge mine XRP and Dentacoin with bitcoin.

<k​ayabanerve:matrix.org> tevador: Ethereum as finality? No, I _could_ do that in a few weeks.

No, our own TFL.

<k​ayabanerve:matrix.org> rottenwheel: You don't have to talk to me. You don't have to talk to me here. I respect your history enough I'd like to talk to you. Can we take a step back or shall I accept it's not happening tonight?

<r​ottenwheel:unredacted.org> oh, I already took it back, this is just the oversimplified big no from me... but sure, what else would you like to hear from me?

<k​ayabanerve:matrix.org> (and by talk, i don't just mean hearing that PoS is anti-thetical to Monero. I consider decentralization and privacy most important, and despite having historically hated PoS as centralized, I'm here now with the argument it'd be more decentralized)

I don't think anyone is advocating to make Monero dependent on ETH/BTC blockchains.

<k​ayabanerve:matrix.org> Also, sorry if the above comes off as pandering/condescending. I legitimately want to make sure we're on the same page as best I can.

<r​ottenwheel:unredacted.org> well, now that's interesting. 🤔 how?

<k​ayabanerve:matrix.org> tevador: Our own TFL? Oh yeah, even if the community wasn't where it is now, 1.5y for full dev and testing with a _reasonable_ community process.

<r​ottenwheel:unredacted.org> tevador nah, that's just sarcasm. 👍

<k​ayabanerve:matrix.org> We have 30 mining pools and 1000 miners on p2pool?

<k​ayabanerve:matrix.org> How decentralized is that, even before we start looking at the distribution of mining power among pools? We have 2 above 25%?

<k​ayabanerve:matrix.org> And yes, the miners can move to another pool, and yet Qubic who is _promoting itself as hostile_ has the hash-power it does.

I personally don't understand how miners can run qubic's malware on their machines.

<k​ayabanerve:matrix.org> I firmly believe a finality layer, where defined validators achieve consensus on blocks, would improve security. I believe if we had 300+ PoS validators, able to be ran at home, all behind Tor, we'd be in a much better position than we are today.

<r​ottenwheel:unredacted.org> yeah, no, mining as it stands is not as decentralized as one would like. someone asked in pass somewhere else... where do the staking rewards will come from if we take PoS route? 🤔

<k​ayabanerve:matrix.org> *2 above 25% each

<r​ottenwheel:unredacted.org> tail emission?

<k​ayabanerve:matrix.org> I'm just saying. I don't think PoS is against Monero: I think centralization is. I think a PoS finality layer would be more decentralized than Monero's current PoW, or even its PoW before Qubic.

<k​ayabanerve:matrix.org> I'd advocate 20-50% of the block reward go to PoS validators.

<r​ottenwheel:unredacted.org> hm, I see...

<k​ayabanerve:matrix.org> (also, that's because I believe we can get 300+ validators behind Tor at home)

<r​ottenwheel:unredacted.org> so 20-50% out of the set 0.6 XMR per block, is that it?

<k​ayabanerve:matrix.org> I hated PoS 8 years ago. It's no longer 9 years ago.

<k​ayabanerve:matrix.org> *no longer 8 years ago

<k​ayabanerve:matrix.org> 😅

<k​ayabanerve:matrix.org> That'd be my vote. I think 20% more than covers server costs, if whoever implements it does their job correctly, and also compensates for the _effort_ which needs to be put in to run and maintain a PoS validator.

<n​oname-user0:matrix.org> this is the only real valid argument against pos that ive read here.

<s​yntheticbird:monero.social> maybe you should read last MRL meeting log then

<k​ayabanerve:matrix.org> I did like the argument, I believe by sgp_ but I'd have to search the logs to check, that holding on exchanges is now explicitly a loss in potential profit.

<k​ayabanerve:matrix.org> I also believe the exchange risk is non-trivial. Since our largest exchange is US regulated (and I personally would trust not to stake user coins, but why am I premising a technical argument on 'US government won't allow it'), and with the rise of DEXs, I believe it's a better position than we face with Qubic.

<n​oname-user0:matrix.org> probably can be mitigated in a way so pos block production can continue albeit slower until staking difficulty adjusts. there are different types of staking algorithms we need to be aware of. im just thinking of the old peercoin base staking model.

PoS doesn't need that much subsidy, should not be more than 10% IMO.

<s​gp_:monero.social> Yeah, I mentioned the potential opportunity cost for money not earned for staking

<k​ayabanerve:matrix.org> Especially since the finality _halts_ with malicious behavior (and a mild synchrony assumption). It doesn't just wipe out the last week and leave us picking up the pieces which have been blown apart like a brick of C4 was detonated.

<k​ayabanerve:matrix.org> tevador: Eh. amount of validators * (hardware costs + annoyance costs)

<r​ottenwheel:unredacted.org> kayabanerve: what is your ELI5 (Explain like I'm 5) concise, layman terms for this 'final PoS layer' thing?

<r​ottenwheel:unredacted.org> How do things work? You only can cram it in a handful of sentences, and no super nerdy astro math, please. :P

<k​ayabanerve:matrix.org> When I did napkin math, I hit ~15% there, and have been saying 20% as I believe it's a better discussion point. There's some game theory to 50% though.

<k​ayabanerve:matrix.org> rottenwheel: that I have a CCS to write a book explaining it in its entirety, which you can preorder now for just $19.95! that's right, just $19.95!

<r​ottenwheel:unredacted.org> lol

<k​ayabanerve:matrix.org> order now and we'll throw in these _free_ monero bookmarks!

<k​ayabanerve:matrix.org> One sec, and I'll write a proper paragraph, but the point of my CCS is so we can have an educated discussion

Miners have all that cost multiplied by their stake (hashrate). For validators, their cost is the same for any staked amount.

<r​ottenwheel:unredacted.org> are you really asking me to pay $20 to read a book that will exactly trigger me!?

<n​oname-user0:matrix.org> I really think people should not worry about rich getting richer type arguments or that the assumption monero has big whales sitting on huge supply are waiting to halt the network as soon as stake is implemented...

<r​ottenwheel:unredacted.org> What a seller!

<s​gp_:monero.social> I think the other major risk, besides halting, is that a PoS cartel forms and colludes with one mining pool to only allow their blocks to be finalized

<r​ottenwheel:unredacted.org> I need 3!

<k​ayabanerve:matrix.org> Because I can personally explain this to you now, and whoever else happens to be here, but I can't spend all day personally explaining it to every member of the community lol

<r​ottenwheel:unredacted.org> Ya, I get that, indeed.

<k​ayabanerve:matrix.org> sgp_: That's easily detected and runs the risk of a social slash, yada yada ya

<s​yntheticbird:monero.social> interesting

<r​ottenwheel:unredacted.org> and what is this famous tevador bandwith PoW thing?

<k​ayabanerve:matrix.org> And we can say we shouldn't consider a social slash because non-autonomous, but us here living and breathing Monero, discussing its future, is proof Monero goes past the blockchain alone

<s​gp_:monero.social> Yeah I wonder how the slashing would actually happen for that in practice. At least it could be detected, but then what

<r​ottenwheel:unredacted.org> it's sounding like we'll be deciding whether we do PoS or bandwith PoW, no?

<n​oname-user0:matrix.org> that risk is so much lower than one random entity with no stake going after 51% of the network via pow directly

<r​ottenwheel:unredacted.org> Or is there any other strong thing going on that I'm unaware of.

<k​ayabanerve:matrix.org> So, finality layer. Proof of Work block comes in. A defined validator set, however defined, observes the block. They run a consensus protocol to agree they all see the same block. If they see distinct blocks, the consensus protocol outputs THE block they'll move forward with. This produces a finalization (like a bunch of digital signatures) and now, Monero nodes will no longer re-<clipped mess

<k​ayabanerve:matrix.org> organize _off_ of that block.

No, PoS would be just for finalization. Blocks would be produced with PoW.

<r​ottenwheel:unredacted.org> tevador no, I worded that poorly. What I meant to say is that it seemed like we'd be down to decide whether we improve what we have at the moment with whatever bandwidth PoW thing you came up with, or kayaba's PoS thing.

<k​ayabanerve:matrix.org> We can _immediately_ shim this with Ethereum. Monero miner mines blocks. Monero miner publishes _the entire block_ onto the Ethereum L1. When syncing a Monero node, you fetch the data from Ethereum. As long as Ethereum doesn't reorg, the order blocks were posted onto Ethereum will be stable, and everyone will see the same Monero blockchain (with no reorganizations).

<k​ayabanerve:matrix.org> That isn't to say we should use Ethereum. It's to provide more background to the idea.

This is the PoW proposal btw: monero-project/research-lab #98

<o​frnxmr:xmr.mx> Defined by coinbases 😝

<n​oname-user0:matrix.org> decentralized cpu mining worked until now only because there was no economic reason for someone to gather a huge supply of generalized compute before.

<n​oname-user0:matrix.org> now that ai uses general compute, there is a super huge risk that's developing against what we thought was a probably fair distribution and distributed consensus mechanism

<k​ayabanerve:matrix.org> A finality layer is just a magic box which decides what block _we do not reorganize off of_.

<k​ayabanerve:matrix.org> It can be defined by Bitcoin. It can be defined by Ethereum. It can be defined by the Monero core team. It can be defined by you personally, if we pledge our souls to you and your dark abyss.

<r​ottenwheel:unredacted.org> Thanks, tevador. Will have a read later.

<k​ayabanerve:matrix.org> Or it can be a decentralized set of validators _the Monero project decides_. The question is _how_ do we decide?

<k​ayabanerve:matrix.org> The best answer, so far, IMO, would be PoS

<k​ayabanerve:matrix.org> but buy my book now to read the chapter on poS and two _other_ selection mechanisms not yet discussed! Only $19.95!

<s​yntheticbird:monero.social> Ai doesn't use general compute sir

<k​ayabanerve:matrix.org> /s :P

<s​gp_:monero.social> I prefer that we let me decide, and me alone 😎

AI does not use CPU compute, that's just qubic's BS. Most AI runs on GPUs.

<s​yntheticbird:monero.social> this

<s​yntheticbird:monero.social> and RandomX is shit on GPUs

<s​pirobel:kernal.eu> I responded to both ArticMine and sgp_ on that. you can scroll up and read it. A modification of this argument would make more sense: if there is a significant reward through staking exchanges will bag this reward or stake for their users and take a large cut. Which leads to centralization and gives the CEX money it does not deserve. That is why I argue for zero inflation if we use PoS.

<n​oname-user0:matrix.org> it can also be looked at in reverse... it's not that pos or pow is the finality, but that they work in conjunction to ensure a lower risk to a single entity controlling the production of blocks and inclusion or exclusion of transactions on the network

<k​ayabanerve:matrix.org> The risk is two-fold: 1) The finality layer stops issuing finalizations. Then we're still on PoW, but services probably shouldn't handle those blocks because they aren't finalized. Services should only handle finalized blocks.

<k​ayabanerve:matrix.org> 2) The finality layer finalizes two blocks. Then we have a net split. The easiest solution for Monero is simply to wait sixty seconds after seeing a finalized block. If it doesn't see _another_ finalization, presumably there isn't one.

<k​ayabanerve:matrix.org> (we can define any latency parameter. the point is, we can just wait for however long that the lack of successful network propagation is negligible)

<n​oname-user0:matrix.org> no, it totally does with the unified memory architecture computers coming out

Any AI code running on CPUs would be absolutely crushed by GPUs or FPGAs.

<k​ayabanerve:matrix.org> There's also a thing where instead of waiting for one finalization, you wait for two? And that gets you a much higher confidence percentage? It's some research from Ethereum.

<s​yntheticbird:monero.social> niche hardware unused in data center which is the primary hashrate source

<n​oname-user0:matrix.org> yes, it runs better on gpus.... but it does run using general compute

<k​ayabanerve:matrix.org> TL;DR There's two risks. One can be mitigated by just waiting a bit longer. The other isn't a network halt, but does mean services should likely halt.

<a​tomfried:matrix.org> No it doesnt at scale

<k​ayabanerve:matrix.org> But also, the 10-block lock? We can replace that with a finality-block lock.

<u​nt0ld:matrix.org> Time to spread this post:I think more fees + GUI wallet mining by default would help a lot. Shill this to devs.

<u​nt0ld:matrix.org> The casual user should go out of their way to disable mining and higher fees. Not the other way around. The Monero software should be a bit selfish and try to support its network. The wallet must by default:

<u​nt0ld:matrix.org> - Set up as a pruned node (public node if HDD, full node if >750GB SSD free).

<u​nt0ld:matrix.org> - Mine, even a little.

<u​nt0ld:matrix.org> - Don't have slow fee selected by default.

<o​frnxmr:xmr.mx> Whats the 101 on blockdags like kaspa? literally asking for a friend

<k​ayabanerve:matrix.org> So if we get finality before the next block is mined, we can now spend Monero after 1 block, not 10.

<n​oname-user0:matrix.org> guys you are arguing that something like qubic wont exist when it clearly does.

<n​oname-user0:matrix.org> and that you are saying there's no risk that a bigger qubic wont exist ever?

Qubic doesn't exist because of AI. This attacks has nothing to do with AI.

<s​yntheticbird:monero.social> another random account spawning to spread fallacies

<a​tomfried:matrix.org> No we are arguing that the AI mumbojimbo is a scam

<s​yntheticbird:monero.social> interesting

<s​yntheticbird:monero.social> they are raining these days

<k​ayabanerve:matrix.org> Oh. Risk 3: The PoS validators can finalize their own PoW blocks instead of the PoW blocks produced by mining pools. This would be observable, due to the reorganizations it'd cause, and isn't _likely_ IMO because we could simply hard fork to remove the PoS validators (and all their stake), reverting to pure PoW. The PoS validators would be risking their entire stake for a few PoW block rewards

<s​pirobel:kernal.eu> tevador: its hardcore how people could ever fall for this. also the whole framing with "the agi grows in the ai garden" is so absurd.

<k​ayabanerve:matrix.org> Also, for BFT PoS? There's almost no reported attacks in history AFAIK? Maybe none at all?

<n​oname-user0:matrix.org> i run ai on my general compute servers at home (not at scale) btw

<k​ayabanerve:matrix.org> PoW is very accessible to attack. Even if PoS may also be in theory, it hasn't hit practice for whatever reason.

<s​yntheticbird:monero.social> who cares

<k​ayabanerve:matrix.org> (BFT PoS, not Peercoin bs)

<k​ayabanerve:matrix.org> Eh, I shouldn't hate on Peercoin that badly...

<n​oname-user0:matrix.org> because you guys are ignoring the fact that general compute is relevant to ai

<s​yntheticbird:monero.social> Thanks for outlining the risk

<k​ayabanerve:matrix.org> Anyways, the finality layer would have to have validators, Monero would need them to be decentralized, I believe PoS would be fine if hundreds of nodes, runnable at home, behind Tor.

<s​pirobel:kernal.eu> me too for inference. but they want to train the model with CPUs. if they really believe that just buy intel and amd stock and short nvidia lol

<n​oname-user0:matrix.org> i have them switch off of ai from this week and running gupax instead

<a​tomfried:matrix.org> Yeah what are you training? Nothing. Your are only using it for inference

<k​ayabanerve:matrix.org> I like tevador's PoW proposal, I'm unsure it's best to implement now (a concern I keep repeating, sorry if doing so to an overkill degree), but and hark sgp_'s comments it's per-miner, not per-hash.

<k​ayabanerve:matrix.org> ofrnxmr: Block DAGs are just blockchains with a fast block time which incorporate uncles in order to avoid how a fast block time causes constant re-orgs as people find blocks at the same time.

<s​gp_:monero.social> I still think the biggest impact from that PoW tweak will be removing small miners :/

<k​ayabanerve:matrix.org> They can be argued for lower latency/throughput to some degree? But not for security

<n​oname-user0:matrix.org> youtu.be/0YV8aZAC9s4

<n​oname-user0:matrix.org> please don't assume you cant do inference or fine tuning on general compute.

<n​oname-user0:matrix.org> yes im not building a new openai sized llm but i do use it for training

<s​pirobel:kernal.eu> yeah. the argument of CEX attacking the network just does not make sense. imagine a CEX using its stake to do a double spend. okay lovely exchange have fun with the token the community forks of with all your capital removed.

<k​ayabanerve:matrix.org> rottenwheel: Those are my arguments for a finality layer. I think it'd be more decentralized and not impact the privacy of Monero users. While validators will have their stakes published, who they are, on a blockchain level (it'd trace back to an output, and with FCMP...) and on a network level (Tor), would be secure.

<k​ayabanerve:matrix.org> I understand if I can't change your mind about PoS. I'd hope we can agree anti-PoS isn't core to Monero: decentralization is.

<k​ayabanerve:matrix.org> (and also, oh fuck there are so many centralized PoS schemes)

<k​ayabanerve:matrix.org> (this just isn't one of them)

<n​oname-user0:matrix.org> 100%

<n​oname-user0:matrix.org> +1

<s​pirobel:kernal.eu> yeah its moving there. also with the new 395+ from amd with higher memory bandwidth. But their plan is to mine agi until 2027 on cpus

<k​ayabanerve:matrix.org> preorder the book now to receive an autographed copy for just $1995 more! that is not a typo, that is one thousand, nine hundred, ninety-five dollars more!

<r​avfx:xmr.mx> PoS on monero can be fine I mean, privacy work, so if they don't know who send to who or how much they can't collude to censor

<k​ayabanerve:matrix.org> (if I don't keep the comedy routine going, I'm going to go insane from repeating this discussion every 6 hours for the past 4 days)

<k​ayabanerve:matrix.org> The largest issue will be because it's private, we can't analyze how decentralized it is other than the facts reported to us: N validators each with S stake.

<k​ayabanerve:matrix.org> It's exactly what stops us from asking how much the exchanges have right now.

<o​frnxmr:xmr.mx> the best thing about being insane, is you dont know youre insane

I think most miners who mine 24/7 will be able to run a node and use p2pool. Especially if there is a 1-click setup.

<k​ayabanerve:matrix.org> Oh, also, of course, ArticMine is fundamentally against "PoS" due to believing nothing is actually at stake as the cost of the stake can _theoretically_ be covered by a short position. While I believe _any_ cost to attack _any_ system can _theoretically_ be covered by a short position, I remain curious if we just rename it "Proof of Coin", if AM would potentially be receptive to a finality lay

<s​yntheticbird:monero.social> regarding qubic, most of their hashrate comes from epyc servers, so I hardly see the local monero node as a limitation

<s​yntheticbird:monero.social> requirement for a local monero node as a limitation*

<o​frnxmr:xmr.mx> Like kayaba said, should be run on anonymity layer, else you end up like zano, with everyone knowing youe stake frequency. I think tor might not he a good choice if were reliant on this layer for finality

<u​nt0ld:matrix.org> patch the gui wallet to mine by default. and not pick slow tx fee by default. simple as.

<k​ayabanerve:matrix.org> My concern was primarily we may see 100% rollover of Qubic miners but not 100% rollover of others. It depends on where their hashpower is coming from/if just one or two people control it.

<o​frnxmr:xmr.mx> it doesnt pick slow fee, it picks auto fee, which is normal fee if blocks are full. Want higher fees? Increase usage

<k​ayabanerve:matrix.org> ofrnxmr: Zano PoS isn't a BFT construction and uses PoS to justify building the next block. I understand your calls for network privacy due to the de-anon posted, I just want to distinguish PoS block building from PoS for BFT consensus.

<b​oog900:monero.social> Is this really worth it over just allowing people to set trusted nodes, with some defaults for the ones who don't set any.

<s​gp_:monero.social> The practical challenge with this is that miners are flagged by antiviruses

<u​nt0ld:matrix.org> and monero wallet is already not?

<n​oname-user0:matrix.org> we could require viewkeys for staking nodes ... under the net network upgrade there will be the ability to audit positions. (im against using this though)

<k​ayabanerve:matrix.org> boog900: Modifying DNS checkpoints to only allow checkpointing 10-blocks deep, with a halt if a conflicting checkpoint exists, isn't unacceptable to me. The advantage to Ethereum, over DNS checkpoints, is it remains decentralized and doesn't require a DNS compromise to halt the network: it requires a re-org of the Ethereum network.

<o​frnxmr:xmr.mx> I dont think were talking about dns checkpointing

<o​frnxmr:xmr.mx> As that is wholly centralized

<k​ayabanerve:matrix.org> sgp_: What if I told you, someone had rewritten all of the Monero wallet software in Rust with almost no mutual dependencies to the C++ codebase?

<k​ayabanerve:matrix.org> ofrnxmr: boog900's recent message on 'trusted nodes' is why I brought it up.

<k​ayabanerve:matrix.org> noname: They can just immediately transfer to a new wallet so their view key has a fresh history.

<b​oog900:monero.social> you can add a new p2p message to propagate messaged signed by the nodes you trust

<o​frnxmr:xmr.mx> i would think boog is refering to p2p monero

<b​oog900:monero.social> you don't need DNS

<s​yntheticbird:monero.social> huh akshually, this archaic technology straight from the overengineered land of the MIT was meant to be decentralized and supports a whole lot of completely (buggy) complete features 🤓

<k​ayabanerve:matrix.org> Fair, we can have everyone independently define trusted nodes and attempt to stay on the chain the majority of nodes believe in. That may chain split and collapse with zero formal analysis

<k​ayabanerve:matrix.org> but it just might work 😎

<k​ayabanerve:matrix.org> /s :P I do not support trying to ad-hoc build what is effectively Ripple/Stellar's consensus mechanism

<b​oog900:monero.social> or we do a fat new upgrade to overhaul our consensus algorithm

<b​oog900:monero.social> which may also have a few issues to work out

<n​oname-user0:matrix.org> arcticmine is old school and also has a huge investment in keeping the network pow at least for coin distribution. so his votes here against pos should be taken with an understanding of his personal bias

<a​tomfried:matrix.org> Ship it

<k​ayabanerve:matrix.org> Or a thin new upgrade to vassal to Ethereum, praise be

<s​yntheticbird:monero.social> Thanks random fallacy acount number 552350, We will consider your opinion proportionally to the number of days you have spent in this chat... which is zero, good lord

<n​oname-user0:matrix.org> it actually isn't super horrible as a stopgap

<k​ayabanerve:matrix.org> The sucky thing is I lost rotten 20 minutes ago :( I did all this for him :(

<s​yntheticbird:monero.social> he will look at it dw

<b​oog900:monero.social> we already have checkpoints that can be optionally disabled, just waiting for a miner with enough willpower and deep pockets

<n​oname-user0:matrix.org> are checkpoints are more or less centralized than temporarily anchoring to eth? or is it exactly the same?

<k​ayabanerve:matrix.org> noname: I actually agree if it'd be time-bounded pending another solution, if in an emergency state (actively facing deep re-orgs/censorship of all other miners)

<k​ayabanerve:matrix.org> Much more centralized

<n​oname-user0:matrix.org> doesnt it depend on who issues the checkpoint/ anchor?

<k​ayabanerve:matrix.org> ... considering here we're discussing a single member of the core team? No

<k​ayabanerve:matrix.org> For any conceivable type of checkpoint? Ethereum is a checkpoint

<k​ayabanerve:matrix.org> a finality layer is a checkpoint

<k​ayabanerve:matrix.org> dns is a checkpoint

We should have some emergency plan in case Qubic can pull off 51% and starts orphaning everyone.

<n​oname-user0:matrix.org> i mean if a distributed consensus comes up with the checkpoint

<k​ayabanerve:matrix.org> my DMs about what blockchain to follow are checkpoints

<k​ayabanerve:matrix.org> tevador: I'd unironically say move Monero consensus to Ethereum.

<b​oog900:monero.social> good idea, we hook up a matrix client to all nodes

<k​ayabanerve:matrix.org> We publish blocks onto Ethereum, and use a SC to keep track of all published blocks.

<b​oog900:monero.social> kayaba now DMs nodes to get them on the correct chain

<k​ayabanerve:matrix.org> Monero nodes sync the first published blocks to Ethereum. If it's invalid, move on. Else, cement.

<s​pirobel:kernal.eu> what if rottenwheel and ofrnxmr start fighting

<s​pirobel:kernal.eu> the chain will halt

<k​ayabanerve:matrix.org> It'd require all nodes run a pruned Ethereum node side-by-side. It'd require miners post their blocks onto Ethereum.

<k​ayabanerve:matrix.org> (or at least an Ethereum light client)

<k​ayabanerve:matrix.org> We also would have to post the entire block onto Ethereum. Else, a miner can post the block hash onto Ethereum, but not broadcast the block onto the chain.

Ethereum is a bloated chain.

<k​ayabanerve:matrix.org> 300kB every 2 minutes on Ethereum... may or may not be pricy? We'd presumably pay for Ethereum's blob storage (a dedicated data relay protocol exactly for a scenario like this)

<n​oname-user0:matrix.org> it's pretty funny but it would probably work if you could vote in matrix only after proving you have stake in the network

<k​ayabanerve:matrix.org> tevador: it's the most secure network on the planet except for arguably Bitcoin, and we can't fit Monero blocks on Bitcoin without decreasing the block size by an order of magnitude.

You can fit the block header on Bitcoin.

<k​ayabanerve:matrix.org> That's insufficient

<k​ayabanerve:matrix.org> I mine a block with TX X. I publish the block header BUT NOT the transaction.

<k​ayabanerve:matrix.org> Does your node accept it OR does your node consider the block invalid, and move to the next block posted to Bitcoin?

<k​ayabanerve:matrix.org> Your nodes either permanently halts (as it can't accept it) or moves on.

<k​ayabanerve:matrix.org> Later, someone else syncs. Now, I do publish TX X. They sync successfully.

<k​ayabanerve:matrix.org> Chain split.

Now consider the same on the Monero P2P network. I send you a block header, but not the TX that's in it.

<s​pirobel:kernal.eu> according to this paper its more expensive to attack ethereum than bitcoin papers.ssrn.com/sol3/papers.cfm?abstract_id=4727999

<k​ayabanerve:matrix.org> Yes, but PoW sorts that out.

<k​ayabanerve:matrix.org> *PoW as the best chain rule sorts that out

<k​ayabanerve:matrix.org> The issue here is we're defining 'first published to Bitcoin' as the new best chain rule.

<k​ayabanerve:matrix.org> And that depends on the validity of the block, which depends on access to the data in the block.

We can require the Bitcoin references to reference the previous reference TX.

<s​pirobel:kernal.eu> it is a bad idea and a waste of time to involve any other chain in our consensus

<k​ayabanerve:matrix.org> tevador: The issue is the Monero TX though.

<k​ayabanerve:matrix.org> We need all Monero TXs, and the block header, available via the Bitcoin network.

<s​pirobel:kernal.eu> just build a monero pos system. time better spent

Anyways, I would not endorse any such solution.

<k​ayabanerve:matrix.org> Some nodes will consider the first valid block published to be B, without Monero TX X. Some nodes will consider the first valid block published to be A, with Monero TX X, which was only published after the fact. These two nodes define their best chain as the chain first published to Bitcoin (and therefore finalized). These two nodes will never reconcile.

<o​frnxmr:xmr.mx> Why not tx hashes

<k​ayabanerve:matrix.org> If TX X is published onto Bitcoin with the block publication, they'd both agree the TX is available via Bitcoin and move on.

<k​ayabanerve:matrix.org> ofrnxmr: That doesn't solve the problem.

<b​oog900:monero.social> can't the rule just be that you only follow the blocks on the bitcoin chain if you can see the full block on monero?

<k​ayabanerve:matrix.org> boog900: But then you can trigger a re-org.

<k​ayabanerve:matrix.org> Have nodes see block A, reject as not present.

<k​ayabanerve:matrix.org> Have nodes see block B, accept as present.

<k​ayabanerve:matrix.org> Publish TXs for A onto Monero.

<s​pirobel:kernal.eu> the question we should answer: how much of a percentage of xmr do we need to stake to make an attack more expensive than it is now?

<b​oog900:monero.social> ok yeah

<k​ayabanerve:matrix.org> Nodes MUST re-org to A because it was first published to Monero and is now verifiable.

<k​ayabanerve:matrix.org> So we lose finality, or risk a net split, if we don't publish the Monero TXs onto Bitcoin.

<k​ayabanerve:matrix.org> spirobel: We can piggy-back off Ethereum's PoS in a few weeks. We can do our own PoS finality layer in 1.5 years.

<u​ser2570:unredacted.org> Dear experts,

<u​ser2570:unredacted.org> for a long-term solution, please make sure, whether a blockDAG POW consensus (Dero, Spectre, ..) is more secure against reorgs or not

<r​ottenwheel:unredacted.org> yikes, yeah. I'm good with this. thanks for the insights though. carry on.

<k​ayabanerve:matrix.org> tevador: If a malicious mining pool was actively censoring all other mining pools, would you support using Ethereum as a finality layer to restore a decentralized mining process?

<r​ottenwheel:unredacted.org> nioc where are we going after XMR? WOW?

<e​longated:matrix.org> 51% of all staked and 51% hashrate

<s​pirobel:kernal.eu> there is no immediate danger to monero. qubic cant do a double spend because it would be illegal. i am not joking.

<s​pirobel:kernal.eu> so we should focus on the long term

<k​ayabanerve:matrix.org> Wait, rottenwheel, you are fine with a decentralized finality layer as described OR are saying "yikes, that still sucks" and would still leave the community?

<k​ayabanerve:matrix.org> Sorry, I'm just unclear

rottenwheel monero saved my life so I am emotionally attached to it

<k​ayabanerve:matrix.org> (assuming it was as described and the 10y development process doesn't yield "we just trust our one node to be correct")

<b​oog900:monero.social> they can reorg a lot of blocks though, invalidating a lot of txs if they go beyond 10

<s​pirobel:kernal.eu> pow defenders. let this sync in. the government is currently protecting monero. do you want to keep it like this?

<r​ottenwheel:unredacted.org> [@kayabanerve:matrix.org](https://matrix.to/#/@kayabanerve:matrix.org) don't be sorry, maybe I'm not being super clear typing today... been a busy summer for me. yes, I mean I would totally leave. 100% no doubt.

<k​ayabanerve:matrix.org> What's your remaining issues if you don't mind me asking?

What would stop Qubic from posting their malicious chain on Ethereum?

<k​ayabanerve:matrix.org> I get I may not be able to convince you. I just don't even understand you at this time, but I'd like to

<f​iatmoneysucks:matrix.org> Community rule number 1: Mess with Monero and you're poking a hornet's nest.

<v​enture:monero.social> the thing about finality is a hunch that it *will consolidate to power brokers. either via rapid price movements causing coins to change hands or any other way. Pow remains completely egalitarian

<r​ottenwheel:unredacted.org> sorry to be so autistic. there is no possible way you can convince me to accept PoS in XMR. no rational or irrational for that matter.

<b​oog900:monero.social> tevador they would have to do it first allowing us to see their chain

<k​ayabanerve:matrix.org> tevador: The fact they'd have to be first to publish, and a miner with 1% of the hash rate who finds the next block 1% of the time can publish onto Ethereum before Qubic can mine their own next block

<r​ottenwheel:unredacted.org> I don't care if it is centralized or decentralized.

<v​enture:monero.social> *PoS-Finality

<r​ottenwheel:unredacted.org> no PoS. full stop.

They will be the first to publish. They will just ignore any blocks that are not theirs. They can do that because they have 51%.

<k​ayabanerve:matrix.org> they won't be the first to publish every block. they can't ignore other blocks.

<v​enture:monero.social> i agree, but that's my reasoning for no PoS

kayabanerve: if using ETH, would I really need to run an ETH node in addition to my monero one? If so what are the hardware requirements to do so?

<k​ayabanerve:matrix.org> So, Qubic mines block A and posts to Ethereum.

<k​ayabanerve:matrix.org> The 1% miner mines block B and posts to Ethereum.

<k​ayabanerve:matrix.org> Monero nodes see A, see B, and see they're valid. Monero nodes now only sync blocks descending from A, B.

<k​ayabanerve:matrix.org> Qubic mines block B2, descending from A, and posts to Ethereum.

<k​ayabanerve:matrix.org> Monero nodes see B2 and ignore it as B came first.

<b​oog900:monero.social> guys I still like my trusted node idea, very simple, no need to run eth nodes

<k​ayabanerve:matrix.org> nioc: If we use Ethereum as a finality layer, you'd need an Ethereum (pruned) node or light client.

<v​enture:monero.social> I like my idea with the hashing 😇

I am not familiar with eth at all

<k​ayabanerve:matrix.org> tevador: Using Ethereum as a finality layer would prevent any re-organizations from occurring. While Qubit may have 51% in this theoretical scenario, they'd be unable to reorganize out honestly submitted blocks. It'd create mostly fair mining.

<n​oname-user0:matrix.org> we need to test the final proposals to see if they can be gamed or not

<r​ottenwheel:unredacted.org> that's admirable! I don't take your insistence on this the wrong way lol. as I said facetiously, I'm sure this has more to do with my irrevocable despise against PoS than any rational logic explaining why it'd be useful and safe for our use case. while I may use other PoS chains, as I hinted earlier, just reading the discussion of staking natively in XMR through that final layer t<clipped

<r​ottenwheel:unredacted.org> hing made me feel so bad, I was like yeah, no, thanks, I'm out. carry on. that's it. all cool. chill vibes. 😁

<k​ayabanerve:matrix.org> (Ethereum blocks are every 12s, so there is a 12s window where two blocks mined at roughly the same time will be considered mined at the same time, despite not actually being IRL)

<u​ser2570:unredacted.org> Doesn't PoS finality add a network layer attack vector with mandatory Tor connection for validator privacy?

<k​ayabanerve:matrix.org> Unless Qubic witholds the next block until a competitor tries to publish, sniping them as they publish, but that's why private mempools exist? To prevent such sniping attacks?

<k​ayabanerve:matrix.org> rottenwheel: I can understand and accept it's irrational. Thanks for letting me in. It just sucks that potentially doing better in this way is off the table for you :/

<k​ayabanerve:matrix.org> Because this wasn't my first proposal and wouldn't have been. It's just the best proposal in front of us. I believe any/all PoW tweaks will be much too subtle.

<r​ottenwheel:unredacted.org> I was born and raised in a monero community where the news were that we were on AlphaBay, massive in all big dnm, botnet farms... what I'm hinting at... pardon my directness here... I am used to Monero, the currency of the outlaws.

<n​oname-user0:matrix.org> cant we just publish after a lag? where we finalize 10 blocks out with no reorg?

<o​frnxmr:xmr.mx> Ok, and for immwdiate mitigations to prevent 34/51% attacks?

<r​ottenwheel:unredacted.org> when you tell me that currency will become your next let's keep growing it by staking it...

<r​ottenwheel:unredacted.org> I feel...

<r​ottenwheel:unredacted.org> offended. let's leave it at that.

<k​ayabanerve:matrix.org> If publishing multiple blocks at once to Ethereum, noname, then a 51% attacker would always win. You have to immediately publish.

I don't like the idea of Monero being just a payload on Ethereum. Wy bother with PoW at all then?

<k​ayabanerve:matrix.org> ofrnxmr: Vassalize to Ethereum. DNS checkpoints. Replicate Ripple/Stellar on Monero with some ad-hoc 'trusted node' code.

<k​ayabanerve:matrix.org> tevador: For rights to publish the next block. Else, anyone could just make blocks and spam the contract with them (empty, no transactions, bloating the chain)

<o​frnxmr:xmr.mx> So the latter 2, can do start on that?

<n​oname-user0:matrix.org> i guess you cant prevent reorg or orpans anyway for the first 10 blocks

<r​ottenwheel:unredacted.org> or the monero where sech1 would break ASICs...

<k​ayabanerve:matrix.org> Plus, if we maintain PoW, we can remove Ethereum and revert to PoW, and potentially enact a soft fork

<s​pirobel:kernal.eu> sorry but this idea is pointless and does more damage than good. if we go PoS lets just have our own

<o​frnxmr:xmr.mx> Dns checkpoints are done, maybe can be improved or toggled on by default

<n​oname-user0:matrix.org> so you arent against stake. you are against staking rewards.

<r​ottenwheel:unredacted.org> /ignore spirobel*

<r​ottenwheel:unredacted.org> easy.

<k​ayabanerve:matrix.org> They'd need to be actively updated by a team member, and never conflict, and we should have code to detect if they conflicted.

<s​pirobel:kernal.eu> rottenwheels feelings are in the room with us

<b​oog900:monero.social> do DNS checkpoints cause reorgs, I thought it was just a warning?

<k​ayabanerve:matrix.org> The fact they exist doesn't mean they have an appropriate amount of infrastructure.

<s​pirobel:kernal.eu> treat carefully

<r​ottenwheel:unredacted.org> I guess... I'm against anything stake. but yes. the rewards part is disgusting by itself.

<o​frnxmr:xmr.mx> reorgs if enabled

<r​ottenwheel:unredacted.org> we're talking about a change of incentives in the network. this has a chain effect. the whole culture around monero may change a bit, if not fully.

<o​frnxmr:xmr.mx> Bt default its juat a warning, bur --enforce-dns-checkpoints (sp?) Causes reorgs

<r​ottenwheel:unredacted.org> fuck all that.

<s​pirobel:kernal.eu> i am also against staking rewards. nominal inflation is pointless

<k​ayabanerve:matrix.org> I'm not going to spend my next few hours chatting in circles in this room, like I have for the past bit. Best of luck to y'all. I have work to do.

<s​yntheticbird:monero.social> I'm surprised you had the patience to continue

<f​iatmoneysucks:matrix.org> Sorry if this is a dumb question, but what if we use several different PoW blockchains for consensus? BTC, BCH, ETC, etc. That way, we wouldn't be a vassal state, and it would still protect Monero.

<s​yntheticbird:monero.social> good night kayaba

I'd find DNS checkpoints less controversial than the Ethereum proposal.

<f​iatmoneysucks:matrix.org> Sorry if this is a dumb question, but what if we use several different PoW blockchains for consensus? BTC, BCH, ETC and others. That way, we wouldn't be a vassal state, and it would still protect Monero.

<b​oog900:monero.social> you just increased not requirements 100x

<s​yntheticbird:monero.social> this one is spicy but funny to imagine

<b​oog900:monero.social> node*

<r​ottenwheel:unredacted.org> +1 tevador

<k​ayabanerve:matrix.org> Agreed less controversial, but they're practically much more centralized and problematic.

<k​ayabanerve:matrix.org> 3.14stache: you can publish differing TXs to each of those, and then you need a protocol to decide between them. Does not work.

<n​oname-user0:matrix.org> this doesn't make sense. you can use big chains to anchor the monero consensus or you can adopt their consensus mechanic

<n​oname-user0:matrix.org> yes lets keep them for software updates as intended

<f​iatmoneysucks:matrix.org> I forgot to point out that users could choose to run XMR + BTC, XMR + BCH nodes and so on.

<r​ottenwheel:unredacted.org> we're talking about a change of incentives in the network. this has a chain effect. the whole culture around monero may change a bit, if not fully. <<>> thanks for being here

<o​frnxmr:xmr.mx> Dns checkpoints literallt already exist

<s​yntheticbird:monero.social> nioc you are a well deserved light in this time of darkness

<s​pirobel:kernal.eu> noic: i thought the culture was to be pragmatic and make changes not like the bitcoiners stuck in their ways

Ethereum already has "ask a friend" subjectivity, which chain is the canonical one. DNS checkpoint = asking a friend, which chain is canonical.

<s​pirobel:kernal.eu> not making this change is being like bitcoin and going down with the sinking ship of proof of work

<s​yntheticbird:monero.social> ironic coming from the guy that show no pragmatism by constantly deflecting counterarguments

<b​oog900:monero.social> I still would rather it be by the p2p network rather than DNS

<b​arthman132:matrix.org> A change of culture? Man you’re being very dramatic. The fact is both POW and POS can be bought out. It’s just can be bought out in different ways

<b​oog900:monero.social> On handshakes nodes share which people (pubkeys) they trust, then the pass along any message signed by a pubkey both trust

spirobel, I understand and appreciate rotten is all I am saying

<o​frnxmr:xmr.mx> Can. 1. Dns operator can use `--reorg-notify` -> run script that checks hash of new vs old chain, checkpoints the old one, and updates dns checkpoints 2. use seed nodes as trusted nodes if not specified on the command line (?)

we understand you but you are now perseverating

I guess many are :D

<b​oog900:monero.social> the problem with just using seeds is the seeds can be taken offline

<n​oname-user0:matrix.org> this is ok. we dont like the incentive to 51% attack the network now which is why we are discussing changes.

<s​yntheticbird:monero.social> skid using remote crash zero day earlier this year as an example

<b​oog900:monero.social> using pubkeys and passing along messages you can run a node on Tor that passes messages to the whole network

<n​oname-user0:matrix.org> how do we setup this trusted layer and prevent a trusted node from doing something bad in the future

<s​yntheticbird:monero.social> honestly I like this idea

<n​oname-user0:matrix.org> basically eth pos slashing was their answer to this

<n​oname-user0:matrix.org> trusted nodes are ones that had x stake

<n​oname-user0:matrix.org> and if they did something bad they lost their stake

<b​oog900:monero.social> I don't know what you mean by how, but we have to rely on enough nodes being honest

<b​oog900:monero.social> it wouldn't ever be 1 going bad can cause it problem

<b​oog900:monero.social> and you would be able to change the trusted nodes

<n​oname-user0:matrix.org> it can be sybil attacked

<b​oog900:monero.social> although most will use defaults

<s​yntheticbird:monero.social> yeah

<b​oog900:monero.social> like the current network ...

<b​oog900:monero.social> and any p2p network

<s​yntheticbird:monero.social> as all distributed system with decentralized consensus can

<n​oname-user0:matrix.org> yes it changes the problem but doesnt solve it

<b​awdyanarchist:matrix.org> I still would like to keep open the idea of pure PoW with time-adjusted blockweight. Individual nodes calculate the time-adjust weight of each block, based on the UTC time of when they personally saw the block. Maybe that would be an extra 5% every 2 minutes, for a maximum of +100% time-weighting for blocks older than 40 minutes.

<b​awdyanarchist:matrix.org> This would make it so that a would-be attacker needs an increasing amount of hashrate (beyond 51% of the network) to do a long chain reorg

<b​oog900:monero.social> it solves deep reorgs for the nodes not sybil

<b​oog900:monero.social> the eth proposal fails if you are sybiled too

<s​yntheticbird:monero.social> Bawdy I think for it to be thoroughly discussed before next MRL meeting you should maybe take this up github

<n​oname-user0:matrix.org> not against this but really dont see the upgrade path to switch over that keeps everything decentralized in that switchover

<v​enture:monero.social> I'm not sure how this will work for syncing / IBD. it can get too messy.

<b​oog900:monero.social> you are adding a local element to consensus which could lead to chain splits for new nodes

<b​awdyanarchist:matrix.org> I'm .."proposing".. it (more like floating it), because there does seem to be alot of consternation with a finality layer, and it's something that could actually be implemented quickly

<r​ucknium:monero.social> BawdyAnarchist: Doesn't this have similar pros and cons to a rolling N block checkpoint?

<v​enture:monero.social> what do you think about my hashing idea (stolen from an old paper), it effectively limits re-orgs to ppl with >51% hash power only, no edge for selfish mining IMHO

<v​enture:monero.social> It's described in the other MRL room

<s​yntheticbird:monero.social> don't mind linking it?

<b​awdyanarchist:matrix.org> That's one reason why there'd be a cap on how much a block could be weighted. For blocks where the header is older than say, 1 day, you can effectively just count that time-adjusted blockweight as "at the max"

<b​awdyanarchist:matrix.org> I think so, but it does still leave open the possibility of reorgs

<v​enture:monero.social> matrix.to/#/!toFcRZtpaiwiyapgVO:mat…a=monero.social&via=cypherstack.com

<r​ucknium:monero.social> I am not for or against a rolling N block checkpoint, like BCH has.

<s​yntheticbird:monero.social> thx

<s​yntheticbird:monero.social> > One way to break a tie without helping selfish miners,

<s​yntheticbird:monero.social> is to increase the entropy of the output of the determinis-

<s​yntheticbird:monero.social> tic prioritization function.

<s​yntheticbird:monero.social> interesting

<b​oog900:monero.social> this doesn't stop big reorgs

<b​oog900:monero.social> I think that is the main target IMO

<s​yntheticbird:monero.social> would 10 block max + entropy in prioritization be incompatible?

<b​oog900:monero.social> yes

<s​yntheticbird:monero.social> oh

<b​awdyanarchist:matrix.org> By not having an N block rolling mechanism, but instead a time-adjustment to the blockweight, it allows nodes that might be temporarily at the edge or eclipsed, to still reorg and re-unite with the rest of the chain

<r​ucknium:monero.social> Maybe, if Monero leans on network-level qualifications for consensus (with rolling checkpoint), nodes could also require blocks to be filled in a sane way with current txpool contents. That would fix empty and/or censored block mining. The Trailing Finality Layer doesn't address that problem, AFAIK.

<b​oog900:monero.social> well no, but it has the same problems as the 10 block rolling checkpoint

<b​awdyanarchist:matrix.org> Partitioning at the edges is one of the failure modes of rolling checkpoints though. A time-weighted approach might prevent that, or at least allow for easier recovery

<o​frnxmr:xmr.mx> Problem woth txpool sanity is we have some very weird txpool behavior

<o​frnxmr:xmr.mx> ive had my 3 txpools diverge significantly on occasion, and take a while to sync up with one another

<r​ucknium:monero.social> BawdyAnarchist: I agree that your idea may have some advantages over rolling checkpoint. On the other hand, if the N in N block rolling checkpoint is 10, then the 10 block lock will never be violated, in theory.

<b​oog900:monero.social> I still don't think it solves the problem of announcing a new block in an alt chain right the reorg boundary when a new main chain block is seen

<b​oog900:monero.social> oh wait your proposal doesn't have a reorg boundary my bad

<b​oog900:monero.social> but yeah then it still allows big reorgs as Ruck said

<b​awdyanarchist:matrix.org> Yeah it definitely doesnt solve all of our problems. But it does give us some additional protection against immediate threats, while still preserving the ability of nodes to always reorg

<b​awdyanarchist:matrix.org> For example, if an attacker now needs 66% of HP to succesfully reorg > 10 blocks, that's a significant protection boost.

<b​awdyanarchist:matrix.org> It's also a change that should be easier to model/understand, and get consensus than either pegging to some other change, or developing a finality layer.

<b​awdyanarchist:matrix.org> *some other chain*, sorry

<s​yntheticbird:monero.social> even a solo miner can reorg > 10 blocks with just incredibly unlikely odd. What's the threshold here

<b​oog900:monero.social> if the adversary publishes their alt chain as they mine they can get around the penalty right?

<r​ucknium:monero.social> We would want more node stability under stress. Maybe #1 on that list is the more efficient tx propagation proposal by boog900 , with a PR by 0xfffc

<r​ucknium:monero.social> Or everyone switch to cuprate ;)

<s​yntheticbird:monero.social> no more wallets for everyone

<b​oog900:monero.social> If I mine in step with mainnet for a day I can overtake after a day and reorg the last day

<u​ser2570:unredacted.org> Dear experts,

<u​ser2570:unredacted.org> for a long-term solution, please make sure, whether a blockDAG POW consensus (Dero, Spectre, ..) is more secure against reorgs or not🙏

<s​yntheticbird:monero.social> kayabanerve already stated that DAG have no security benefits

<v​enture:monero.social> the thing about re-orgs is, they are not harmful, unless they are sustained. you would have to have incredible luck + a double spend attack victim at the same time

<b​oog900:monero.social> with monero any reorg past 10 blocks starts invalidating other txs

<v​enture:monero.social> of course, if we are saying "that doesn't stop one to get 51%, (ie sustained), then this is the case against PoW altogether

<b​oog900:monero.social> hence all the arguments to add a backup system

<v​enture:monero.social> it never happens, when there is no edge / noone having more than 51% or can selfish-mine with less

<v​enture:monero.social> to some extend, it's a community thing to keep an eye on pool share

<b​oog900:monero.social> what if they do get 51%?

<v​enture:monero.social> depends on what they do with it

<s​yntheticbird:monero.social> do people remember the minexmr event?

<s​yntheticbird:monero.social> they had 51% of the network

<s​yntheticbird:monero.social> at some point

<v​enture:monero.social> but that's the nature of PoW, would be to have that threshold than 25% or 33%

<v​enture:monero.social> no. what happened?

<s​yntheticbird:monero.social> they achieved 51% of the hashrate, people panicked. minexmr owner decided to pull the plug on purpose

<b​oog900:monero.social> so your point is that we should just do nothing and hope that we don't suffer a 51% attack?

<v​enture:monero.social> well done i guess

<s​yntheticbird:monero.social> rie or die

<s​yntheticbird:monero.social> ride*

<b​oog900:monero.social> yolo

<v​enture:monero.social> we should at least raise the bar from the current 25/33% selfish-mine attack (not sure which threshold is significant)

<s​yntheticbird:monero.social> yo elongated I'm joking

<v​enture:monero.social> and do nothing afterwards is an option... that whole qubic thing might be a psyop