Hey everybody, just catching up. Sorry I had a conflicting meeting on Wednesday. Unfortunately I’ve been swamped with work and life, so my IRC time has been very limited :- (

Interesting discussion about the word ‘heuristic’ - seems like it gets used differently in every field. In the case of statistical heuristics, I usually use it to mean any of viewing the data that provides some sort of probabilistic insight.

e.g. “transactions with ringsize 41 were typically generated by MyMonero users” is something I would call a ‘heuristic’ because it allows me to glean some likely information. It has neither positive nor negative connotations.

RE volume anomaly -

Seems like they were using standard fees. Also given that most of the outputs were spent within 15 blocks, we know that they weren’t sitting in the mempool too long or generated in advance.

Mempool analysis would be interesting, we might be able to do some of that from archival project data sets.

RE contributor to help with organization -

I think that in terms of keeping up with literature we are not quite on top of things. Through 2019 I believe that MRL was doing an excellent job of tracking relevant literature and studying new methods as they were published. WHen a new paper hit arxiv we’d have a TL;DR and ELI5 of implications within a day or two.

Our coverage has not been as complete lately, and frankly (for the safety of Monero users) Monero -research and -dev can’t afford to be behind the curve. I think that it would be great if somebody could take a few months to comb through literature and combine a list of relevant papers, then write up an overall threat assessment.

Having MRL be up to date with Monero analysis best practices is especially important as we try to assess the threat posed by Ruck’s work regarding ring signature deanonymization.

To ascertain if or how alarmed we should be, it’ll need to be evaluated in the context of other advances such as implementation within a Hopcroft–Karp algorithm compatible framework seeded with probable edges from the other analysis methods developed over the last few years.

Because we haven’t been keeping up with literature as closely, it’s hard for me to tell whether a competent adversary would be able to combine these approaches to deanonymize 0.2% of ring signatures or 2% of ring signatures or 20% of ring signatures.

I prefer to be reserved and measured in public statements. Unfortunately some of my private comments have been plastered around IRC lately. While not how I would have chosen to communicate my concerns to the community, it is true that my perspective is that we should avoid complacency and have a few qualified people very seriously evaluate how all of the analyses in the public domain or developed internally can be

combined.

First step of this is getting 100% up to date on the state of research through 2021. To avoid wasting time around what constitutes centralization or endorsement (imho just a distraction after a certain point), maybe we could just get somebody to be “a director” (instead of “the director”) of MRL, lol.

I think that a survey of the landscape would be a top priority and first deliverable. Another possible early deliverable could be an organization of Monero’s ideas list, wish list, etc. It would be amazing if they produced something like this: github.com/ZcashFoundation/zomg/blo…/ZOMG_whitepaper_v1.0_published.pdf

When somebody wants to start contributing to a privacy coin, it’s VERY clear how to get involved with Zcash, and hazy how to get involved with Monero. Telling an eager new contributor ‘go read the github issues’ is less helpful because it doesn’t provide context around what is urgent, what is a moonshot, what is contentious, what will actually make it to CCS and become funded.

RE terminology for decoys -

I believe that the best terminology is the least ambiguous terminology. For that, I would very much motion to retire the term “mixins” because it gives the impression that Monero is a mixnet which is fundamentally wrong.

My vote is for “decoys” because it is absolutely unambiguous

While “fake_outs” is in the name in the codebase I’m not a huge fan of using it as the main term in discussions because I’m not sure that people who haven’t been around IRC would intuit that a dummy out and a fake out are different things (dummy output = 0-value output generated by a transaction, fake outs = unknown-value decoy input to a transaction)

Anyways, as we’re going to be producing more information and onboarding new people we should pick something and try to adopt it throughout discussions. As long as it’s not “mixin” lol.

ithmus I am talking with a few profs about a director

fyi

decoys is used widely

I've been on board with decoy for a long time so fine by me

thx for stopping by isthmus.

I have been using decoy since ~2017 and I like it

To be clear, the reason I have used "mixins" is that it is the terminology in Moser et al. (2018), and generally it's nice to be consistent with papers you are citing. But yes, everyone has given good reasons for changing to "decoys". I can edit my CCS to reflect that.

it's nice to be consistent with the practitioners of the technology. Moser was not, and mixin is not the accepted term

mixin is associated with coinjoin and other unrelated mechanisms; using it here is misleading.

hyc: That makes sense. I wasn't aware of the history of the different terms. Going forward, I will refer to it as the "decoy selection algorithm"

excellent

i think i'm "ok" at tapping into zeitgeisty things, and this is what I think...

lots of folks, especially those who are influenced by identity politics, are swayed by Monero's use of electricity. however there are also good arguments that the current system of maintaining fiat currency is enormously hard on our planet. therefore debunking this is one important thing, i think.

the other one...

when folks buy bitcoin they think that when they use it that it will function like monero. basically, people need to be aware that monero is what most folks want when they buy BTC.

finallly...

i don't know much about the monero user or dev community, but i DO think it's important that people are respectful even if another person's gender, etc, might be puzzling to them. it's important to create a welcoming atmosphere for ALL people, no matter how weird their political, sexual, or general orientations may be.

i'm NOT suggesting to buy into identity politics though.

that's my main 3.