02:31:45 <gingeropolous> proof of space has been around for a while
03:56:37 <xmrpow> hyc: What is your opinion on the premise that it would reduce the energy consumption?
10:32:36 <hyc> yes, proof of space has been around for a while, but all of a sudden Chia is popular.
10:32:56 <hyc> so popular even StorageReview is talking about choosing the best HDDs and SSDs for it
10:34:03 <hyc> I don't believe it's reducing energy in any meaningful way, since energy is still being locked up in producing the drives themselves and keeping them online
10:34:42 <sech1> Plotting in Chia is also compute-intensive and you need to keep plotting constantly to stay competitive
10:34:49 <Quarky93[m]> With earlier methods you could cheat with more compute power. Chia enforce the use of a VDF which takes care of that
10:38:16 <hyc> with all that storage space devoted just to mining, how much space does their ledger take?
10:38:50 <hyc> 1 plot is 100GB
10:40:38 <hyc> none of these systems make any sense in the long run. RandomX 1cpu=1vote would become meaningless if it were to take over the world.
10:40:56 <hyc> imagine 10 billion people all using Monero and RandomX. 1 vote out of 10 billion is just noise.
10:42:02 <hyc> Real voting only makes sense because it has hierarchical scale. Your vote in local elections has a large impact on your life. Vote on broader elections, steate, federal has decreasing weight.
10:44:12 <hyc> mining should be geographically scoped. maybe by ping time.
10:46:35 <Quarky93[m]> <hyc "mining should be geographically "> Hmmm interesting take actually. Something like sharding essentially
10:47:07 <Quarky93[m]> * > <@freenode_hyc:matrix.org> mining should be geographically scoped. maybe by ping time.
10:47:07 <Quarky93[m]> Hmmm interesting take actually. Something like sharding essentially?
10:48:05 <hyc> I suppose, yeah
11:14:53 <gingeropolous> <hyc> imagine 10 billion people all using Monero and RandomX. 1 vote out of 10 billion is just noise. >>> well right, but there are 10 billion votes. this doesn't make sense to me
11:16:12 <hyc> the PoW is there to secure the network. at 10 billion miners, +/- any 1 particular miner makes no difference
11:16:24 <hyc> so that 1 extra miner is just wasted resources
11:16:55 <hyc> tx validation, mining beyond a certain scale is meaningless
11:17:22 <hyc> there's no reason for 10 billion people to validate my coffee purchase
11:24:47 <sech1> oh there is a reason. Miner rewards.
11:30:56 <gingeropolous> i agree re: transaction validation, but i don't think there's a reasonable way for a node to say "well, this tx has been seen and validated by 10 billion ppl, so I don't need to do it"
11:32:03 <gingeropolous> that would require tiers of nodes with different levels of authority, and then you get into that quagmire.
11:32:45 <gingeropolous> decentralization etc. is inefficient, but its emergent property is priceless.
11:33:07 <gingeropolous> and grand sweeping generalizations are great
11:38:19 <hyc> sech1 at 10 billion users I think the odds become worse than playing the lotto
11:40:00 <xmrpow> It s very impressive what media attention Chia got over the last weeks. Even german traditional media mentioned it, which is quite strange compared to that their horizon does normally not scale beyond bitcoin and ethereum price attention.
11:41:03 <hyc> yeah, to me it seemed like they came out of nowhere and all of a sudden HDD prices ar ethru the roof
11:44:49 <xmrpow> https://blog.trailofbits.com/2018/10/12/introduction-to-verifiable-delay-functions-vdfs/ At least chia seems to be a thing since 2018.
11:45:42 <sech1> Chia only enabled transactions on their blockchain on May 3rd, this is why all the hype
11:48:27 <hyc> ah ok
11:49:17 <moneromooo> A 3 year chain with no txes ?
11:49:33 <xmrpow> Took them long :)
11:50:33 <hyc> maybe they had a 3 year testnet
11:50:37 <sech1> As weird as it sounds
11:51:06 <moneromooo> Can't have tested much then.
11:52:37 <xmrpow> It would be interesting if they solved the open problems mentioned by the upper article. Does anybody know?
11:53:39 <hyc> haven't followed it at all
12:50:27 <Quarky93[m]> <xmrpow "It would be interesting if they "> One main thing is ASIC acceleration of VDFs. The faster a hypothetical ASIC, the wider the tolerance that has to be set for the VDF (such that the ASIC advantage is still too slow to front run), but that comes with consequences for the network, I.e speed.
12:51:25 <Quarky93[m]> Ethereum foundation is looking to manufacture ASICs and give them out for free
12:52:18 <Quarky93[m]> One nice property of that particular VDF scheme is that very few VDF nodes need to be good actors for the system to work
12:52:38 <Quarky93[m]> * One nice property of that particular VDF scheme is that very few VDF nodes need to be good actors (and fast enough) for the system to work
12:55:11 <Quarky93[m]> One could also set the VDF tolerance with FPGAs in mind. Good actors with FPGAs can help defend against unknown ASICs, allowing the tolerance to be set lower than with just CPU
13:18:16 <gingeropolous> <Quarky93[m]> Ethereum foundation is looking to manufacture ASICs and give them out for free >>> i thought they were going PoS
13:25:41 <gingeropolous> was doin ok until this "finite abelian "
15:30:30 <Quarky93[m]> <gingeropolous "<Quarky93[m]> Ethereum foundatio"> VDF can be used for proof of stake too. As a part of distributed random number generator
17:49:34 <yanmaani> hyc: the system is self-regulating; if randomX takes over the world, as long as you allow block sizes to go up, mining profitability will plummet, meaning only the lowest  cost producers will bother
17:50:24 <yanmaani> btw, had an idea on how to fix p2pool but noboduy responded in #monero
17:50:44 <yanmaani> say you'd run p2pool, a sharechain with lower difficulty than the mainchain
17:51:00 <yanmaani> then you'd run another sharechain on top of that, with say 100 other miners, with even lower difficulty
17:51:36 <yanmaani> and then you'd do PPLNS on the lower sharechain too, so you'd have two levels
17:52:23 <yanmaani> that would allow you to get an acceptable difficulty - one global monero chain, then one global p2pool chain, the one, say, Asian chain, and then one for region XXX, and then ...
17:55:15 <yanmaani> you would need to put in a lot of effort to develop the bootstrapping, but shouldn't be impossible
18:51:36 <gingeropolous> yanmaani, cool.
19:29:29 <yanmaani> I'm not sure if it'd actually work or if it just makes sense in my head :P
19:30:32 <yanmaani> not really sure how you'd distribute it geographically, for starters. The P2P work would be non-trivial
19:37:41 <sech1> I can't even... https://www.randomxfidgets.com/
19:56:48 <selsta> bot generated shop? lol
22:30:55 <xmrpow> Has anybody of you already played with ryzen threadripper pro concerning mining? Are the extra memory channels worth it?