restricting to node-rebroadcast because...?

In an MRL meeting, enforcing it by consensus rules was considered to be too strict. "Enforcement" at the wallet level is really no enforcement at all. Enforcement at the node level would be a nice compromise.

See the log: monero-project/meta #616

can always change it later on. If it gets more obvious, then you can whine to wallet authors etc.

stagenet testers wanted: reddit.com/r/Monero/comments/t5nz6y…uncing_rino_stagenet_testers_wanted

Oh dude I love the idea of having the decoy selection enforceable at the protocol level

We need this and we need to ban tx_extra

MOR FUNGIBILITY

Is 800 hashes/sec worth mining? I have 4 other clients mining which are on better hardware (2 Ryzen machines, 2 xeons), but have this one server laying around doing nothing (it's a VM server) which I spun another VM up on to mine. Disregarding power costs since the server is on 24/7 anyways, is it worth it to have mine at such a low rate?

mining in a VM will just give you unnecessary overhead

Sure, I get that. But this is a VM server which had some extra resources available, so I just spun up another VM and threw xmrig on that VM.

Furynation: honestly, 800 H/s is not worth it at all

merope: ok good to know, I'll leave it be then. Thanks!

hello

mooo are u here

23:58 < yanmaani> can always change it later on. If it gets more obvious, then you can whine to wallet authors etc.

Start whining now.

yanmaani: It would be best if people can't retroactively distinguish things because things hadn't been updated yet.

11:23 < merope> Furynation: honestly, 800 H/s is not worth it at all

merope: Does it not contribute to network security? It might be worth it in a "contribute to securing the network" sense but not in a "profit" sense.

. . . unless it's really too slow to even be useful for securing the network.

We are talking about 800 H/s on a 3 GH/s network

I don't know the point at which it becomes useless, so . . . that doesn't tell me how far it is from useful in a nonprofit sense.

How much richer would a billionaire with $300B if you gave them $800?

It's such a small drop in the ocean that it's not even worth the time it takes to set it up

How much richer would a billionaire be if you gave the person 10K packages of $800?

300B, not 1B

Does that mean you don't get my point?

8M / 300B = 0.00266%

It means that you're not getting mine

I get yours, but I wonder how much of a network effect is being inflicted by telling people that low-yield mining isn't worth it.

merope: This much

Okay, if it literally means you effectively don't get anything (even at a loss), I guess that would make it useless.

even in a pool

Oh oops, it's 3 GH/s

So 0.266%

What's the network security effect of pooling that with a bunch of others?

Is it measurable?

It doesn't really change much, strictly speaming

*speaking

I haven't had enough math education to make it worth my while to calculate this stuff myself.

I mean how much network security effect you'd get from a pool of a bunch of people doing that, not just one of them, to be clear.

The only difference is that by pooling the hashrate, each individual miner gets lower variance in their payouts

But the security contribution to the network is the same, because it's based on raw hashrate

Presumably they'd actually *get* payments, if infinitesimally small, whereas solo mining you probably wouldn't get any payouts ever at that rate.

Assuming that they actually contribute to finding a block

Yep, pretty much

Okay, so it's raw hashrate that matters, and not whether you end up actually producing a mined block, I guess.

wait

16:30 < merope> Assuming that they actually contribute to finding a block

No, finding a block is what ultimately matters

That goes back to my thought on pooling perhaps being useful for low-yield mining.

Yes, that's exactly why pools exist

But if your hashrate is that small, your contribution is basically negligible, even in a pool

In short, it seems like the advice for the VM server approach would be "make sure you pool mine", not "don't bother", then.

Keep in mind that the network is currently composed of 100k-1M cpus, with an average hashrate of 30kH/s - 3kH/s

. . . because network effect can spread the practice of turning unused capacity into hashrate in aggregate of a lot of people's CPU time.

What does "30kH/s - 3kH/s" mean? Is that subtraction? Is that a range from high to low?

It was a range

30 kH/s if it's 100k cpus, or 3 kH/s if it's 1M cpus

ahh, I see

thanks for clarifying

With 800 H/s and 0 costs, you can expect to earn ~10$/year in the current network and market

Yeah, I'm not talking about profit, or even gross earnings.

There are evidently a lot of people intentionally mining at a loss because of poor electricity bill economics just for the sake of helping the network.

. . . and that's cool with me.

Ironically, those people mining at a loss are also lowering the profitability for everybody else, thus increasing the number of people mining at a loss

It's even self-serving, to some degree: doing so for network health can help to encourage others to do the same, thus inspiring many by the actions of one, resulting in a measurable benefit, in theory.

Or pushing out people who can't afford to lose money

har, good point, re: reducing mining profits

I hadn't considered that in this discussion.

No, the whole idea behind pow mining was to introduce a financial incentive to support the stability of the network

Mining at a loss fucks with that incentive for everyone

I, for one, would not stop mining in the hypothetical case of getting down to only breaking even, because even then it's basically a way to get free exchange of fiat for Monero via the electricity utility provider.

In fact, at a slight loss, it'd still be cheaper than buying, much of the time.

That assumes you don't have hardware costs to recover

of course

If you don't have anything else to do with the hardware, though . . .

If you do, breaking even is still a bad deal

Some people need only three computers: a laptop, a node, and a mining rig. If the rig is breaking even, and you'd take a bigger loss trying to sell it as used hardware, you might as well just convert fiat into Monero.

It's all based on one's own local needs.

. . . but your point is well-taken on mining at a loss potentially causing problems for mining incentives more broadly.

(if a lot of people do it "altruistically")

It's often the case that "altruistic" actions end up being worse for everyone, overall.

No "potentially", just look at the number of posts complaining about the lack of profitability

merope: Thanks (Again? Did I already say thanks?) for mention of the incentive effects on the broader network.

I'd still say "potentially" because I don't know for sure that all the complaining translates into ongoing negative incentives for the network as a whole -- but it seems (to me) *probable* that this is an overall negative effect.

Just the other day somebody wrote in the MRL chat that "Devs MUST do something about it! We can't let poor miners lose money like that!"

I'm glad you appreciate the different perspective :)

Disincentivizing extremely large mining operations seems like a good thing, after all. That has to be balanced against incentives smaller mining operations, whether pools or solo, which help to keep the network decentralized more.

s/smaller/for smaller/

Large mining operations are already naturally disincentivized

. . . and I don't have the statistics to make precise judgements on the likely ideal level of anything.

sure, between CPU mining and P2Pool and so on, there are such disincentives

At the same time, though, it wasn't enough (yet) to avoid the 50%+ pool effect recently.

No, the hardware and the pooling software is irrelevant

The incentive agaist large ops comes from the fact that, assuming the network stays constant, the % of the total blocks you find (and thus block rewards you earn) goes up as x/(k + x)

Where x is your hashrate, and k is the (constant) "rest of the network"

For x<<k, this is basically linear

But once you start controlling a significant chunk of the network ( >1%) you grow at a sub-linear rate

(Of course, this assumes a static network, which is not really the case in reality)

The general incentive is for the least profitable people to step out first, while the most profitable miners have the greatest incentive to keep mining

And profitability of an individual miner depends on their hardware's efficiency (H/s/W) and their electricity cost ($/kWh)

The greater your efficiency, and/or the cheaper your electricity, the more profitable you are, the greater the incentive to mine

This is geneally true for all PoW coins, regardless of algos and hardware

The reason why asics dominate other coins like Bitcoin is that they are significantly more efficient (by several orders of magnitude) than common hardware, and they are financially profitable to manufacture and sell for the companies that make them

Monero's RandomX limits the efficiency gains that can be acheived because it would essentially require designing a full cpu, from scratch

But if you can do that, and do it profitably, then you are directly in competition with Intel/AMD/Nvidia/Samsung at their own job, so you might as well forget about mining and just enter the cpu manufacturing business

Hardware *does* factor into it, in that there isn't the same kind of price-tier problem that you get with (for instance) ASIC mined blockchains, where *only* people with a fair bit of money to spend can be profitable.

. . . and P2Pool also factors into disincentives in that big mining operations, including centralized pools, can now have more serious competition from a decentralized pooling mechanism.

. . . but yeah, what you describe is obviously part of the incentive to avoid running big mining operations, to some degree.

It doesn't feel like enough, all by itself, given the way BTC has gone.