a live nonce distribution visualization would be dope

hrm.... i might be able to do that maybe

Sounds cool ;)

oh fuck yeah. it already exists: apollo.open-resource.org/flight-control/noncewatch

ah damnit its old

is that for monello?

I thought we were over 2M blocks now

"live"

aye. and they're all open this and that but no link to code

but yeah. a running monerod. some database thing to store nonce for a given blockheight. and then one of those javascript chart display thingies

would also be cool to run a live cluster analysis / PCA to detect when new patterns emerge

kl_, hrmm, that looks like its just recent blocks?

hrm, there is a little band there on the bottom right

<ＨａｓｈＶａｕｌｔ> Yeah, last 100k blocks.

indeed, this is something new: i.imgur.com/twOdsl0.png (lower right corner)

it almost looks like someone has high byte of the nonce set to 0x01

The red line is Nov30/Dec1 2019?

red lines are fork heights

it started around block 2144150 (July 17th, noon)

there was a 130 MH/s jump at minexmr at that time

doesn't xmrig-proxy use the higher part of the nonce?

it does, but it's a random byte

actually jump at minexmr doesn't align with that time, 20 MH/s at f2pool seems more likely for July 17th noon

Most people use xmrig, right ? Maybe someone who is not me could code a mode where it mines at a random pool in a list, with probability inversely proportional to that pool's share of the network hash rate. Then encourage pool ops to do "reconciliation", where if Alice has sent 1000 hashes to pool0 and 200 hashes to pool1, and Bob sent 500 hashes to pool0 and 500 hashes to pool1, reconciliation

moves to 1200 for alice/pool0, 300 for bob/pool0, 700 for bob/pool1.

Pools already run on trust, and this reconciliation would avoid the problem of keeping "dead" pending balances when mining not much on a pool, wich causes miners to not want to do that.

This should create an easy (and default) way for miners to avoid pools with large hash rate while keeping the same payout frequency.

It does add a gatekeeper/censor role to whoever maintains the pool list, but that also has positives I guess.

I'd rather see sxmr and minexmr switch to self-select only mode :D

Changing xmrig doesn't rely on pool ops to do something though, which seems like pissing in the wind.

Ooooh...

Change xmrig as I said, and add a 0.5 weight multiplier for pools that don't use self select :D

several pools could mine to a multisig address (if that's possible)

sech1, is there any reason why changing setting the "high byte of the nonce to 0x01" would increase hashrate?

is there something unique about the programs that are created if that byte is changed?

absolutely no reason, it's probably a bug in the proxy they use

ah

sech1, you have the code hosted somewhere for pulling that from the chain?

githubs etc

no

bummer. i've half a mind to make a noncewatch webpage thinger

it's mostly manual and I have to run a bunch of scripts to get that picture

and edit the first one before running to adjust block height

yeah, i get it. i just figured I could hack what you have into something else

man nanopool's been quiet

it calls get_block_headers_range monerod RPC, then extracts nonce values and then calls gnuplot to draw the picture

yep, i can imagine the bash script i would write to make that all happen

<ＨａｓｈＶａｕｌｔ> Well, if you'd like something more than our api already generate - just let me know.

<ＨａｓｈＶａｕｌｔ> Like date range, height range, higher resolution.

ah, didn't know that was u

this interface was pretty awesome, allows for exploring: apollo.open-resource.org/flight-control/noncewatch