Monerod reserves a ton of ram while syncing if available but doesn't actually use most of it right?

This is data.mdb mapped into memory

meeting 2hr

fast ECIP (see 12:00) youtube.com/watch?v=Zv-3TBgb_KI

for use in curve trees

meeting time monero-project/meta #829

1. greetings

hello

hi

Hello

Hi

2. updates, what's everyone working on?

me: working on 'implementing seraphis' companion paper for the main paper, and need to prep for monerotopia

I released my analysis of the privacy impact of Mordinals:

reddit.com/r/Monero/comments/12kv5m…acy_impact_of_mordinals_monero_nfts

Things have been quiet on the Mordinals front. Fewer than 50 Mordinals minted in the past week. I am now tracking Mordinal transfer transactions: gist.github.com/Rucknium/67cc9efdf7e43a40c52417611b322d43

According to my count, there have been 126 Mordinal transfer transactions

Considering only outputs from Mordinal minting transactions as black marbles, mean effective ring size is 15.88 as of yesterday.

That number will rise toward 16 asymptotically as time goes on unless many more Mordinals are minted.

plowsof: we are past Liam Eagen's self-imposed deadline, any news?

Ive mainly focused on lws and serialization, but I am _finally_ shifting back to some bp++. going to be a bit I think

koe what was the self-emposed deadline?

was he working on the implementation instead?

vtnerd: he's updating the paper

Hi

Could be that BP++ doesn't work.

<vtnerd> "koe what was the self-emposed..." <- The 14th

at least as late as feb 23rd he was employed by blockstream to work on it

3. discussion

Rucknium @rucknium:monero.social: did some analysis of the distribution of number of outputs in each decoy

I think very telling in for conversation of standardizing outputs.

2 outs are in every decoy (though it seems there were 100 rings where there was only one 2 out decoy). 16 outs have the next highest prevalance and 3-15 are relatively rare

Here's the table: gist.github.com/Rucknium/0737163a980a07cf9c837700771d0dea

There are row >16 there I think because some outputs are selected from very old outputs without the 16 output limit.

is the incidence rate different from what you'd expect?

personally it was what i expected based on how people spend

UkoeHB: was that a question for me? If so, incidence rate of what?

Rucknium[m]: does the distribution in decoys match the expected distribution?

I expect it to roughly match the actual output-per-tx distribution on the chain. Since I don't know what that is (yet), it vacuously matches my expectations :)

I can produce the tables to see if it is close

(not technical answer) looks to me like it doesnt descriminate. Ie, looks to be in lin with real spending behavior, but with unlucky/lucky outliers (100 rings with only one 2 out, for example)

ah

are there any other topics we should discuss today?

I think ofrnxmr believes that outputs from 16-out txs are not credible decoys (i.e. are black marbles)for rings that spend from 2-out txs. I am not convinced, but I will keep an open mind.

oh no

actually - id more think of them as.. grey.

they arent standout transactions, but they are obvious not-real-spends if say walmart is accepting xmr in person

If someone was paid an output in a batch tx from an exchange or mining pool and then spent that output to the merchant, it's the real spend. IMHO, these are credible decoys

BP++: no updates since March 28th where April 14th was promised. no reply to an email sent on the 12th yet

If the tx the merchant recieves has fifteen 16 outs, and one 2 out, the real soend is clear

Yes, but only Mordinals do that. Deliberately.

The 100 on your chart are morbinals?

Mordinal transfer txs

I'm pretty sure. I said it in this channel a few days ago.

Yeah, but those 100 from this chart are confirmed to be morbinals?

ok. I thought it was an outlier

What is the typical tx size when mordinals are involved?

Mordinal minting? I can calculate the mean tx size. One moment

370687101 / 43096 = 8601.427 bytes

Mordinal transfer txs:

281030 / 126 = 2230.397 bytes

There is at least an opportunity for pricing if not looking at tx size limits

This is an issue l am looking at in a general sense

Tx size pricing and limits

The new version of the Mordinals software has a self-imposed fee policy. tx fee rises exponential with ex_extra size

Of course, people can just choose to run the old software and avoid the fee

That is actually good

Yes, they even credit you, ArticMine[m], for the inspiration of the curve :)

... but we can use node relay for this

I think we can end the meeting here, thanks for attending everyone

Appreciate it koe.

UkoeHB: BP++ update: we have the conference version that was submitted to ieee s&p. It is not final / some small things (due to the page constraints) but the author considers it "good for review"

^ the conference version of the BP++ paper

Liam says a full version without page limits will be posted to eprint once done

thanks xmrack, Liam wishes Monero a happy birthday too xD my untrained eye sees its a very different paper, so cypherstack will come back with their quote on this one as the author says its good for review.

nice