layout |
title |
summary |
tags |
author |
post |
Logs for the Monero Research Lab Meeting Held on 2018-11-19 |
Surae work, Sarang work, and miscellaneous |
dev diaries |
crypto |
research |
|
el00ruobuob / surae |
Logs
<sarang> heyo
<sgp_> hello
<serhack> Hi surae!
<suraeNoether> good timing :D
<vtnerd> hi
<suraeNoether> i decided to not look at screens this weekend, because i got like 5 migraines last week, and i made it from the hours of 8am saturday to 1pm saturday
<suraeNoether> but i still felt like a totally new person :D
<suraeNoether> anyone have any questions before we begin?
<sarang> Welcome suraeNewther
<suraeNoether> nice
<suraeNoether> \nick suraEyeOfNewt
<suraeNoether> damn
<sarang> s/\\/\//
<sarang> Shall we begin?
<suraeNoether> omg that should be on a t-shirt with MRL's logo
<suraeNoether> yes
<gingeropolous> how many rings can a ring signature ring if a ring signature can ring rings?
<suraeNoether> let's begin
<suraeNoether> that's actually the subtitle of nick van saberhagen's autobiography
<sarang> Who shall start with updates?
<suraeNoether> Allright, we've gotten greetings out of the way. Sarang: how is that timing code going? still on iteration 18?
<sarang> Nope, 19
<suraeNoether> Sarang is currently running some timing experiments for me on matching bipartite graphs
<suraeNoether> okay, for it to complete 20, i estimate it will be done in mid december, so i say we don't fall prey to the sunk cost fallacy
<suraeNoether> i'm going to re-rig the code and run it over a smaller sample of the parameter space and get some results
<sarang> Want to give your update suraeNoether first?
<sarang> cool
<sarang> Getting some hard numbers with scaling information will be useful
<suraeNoether> Sure: Right now my time is being spent describing this bipartite matching graph problem that traceability problems in monero boil down to
<sarang> very useful
<sarang> having a complete description in terms of graph theory is marvelous to have
<suraeNoether> the idea is simple. if someone were to try to trace the monero blockchain, they would at least have to perform the following matching task. that sort of idea
<suraeNoether> and we have known bounds on the complexity for that matching task
<sarang> Exactly, and now some code to get an idea of scaling
<sarang> at least order of magnitude
<suraeNoether> this is all in the pursuit of a formal model of de-anonymizing a "mixing with an anonymity set" style anonymity system like that of Monero.
<suraeNoether> i've begun writing how we can use this approach + zcash turnstile to place bounds on the difficulty of de-anonymizing zcash as well
<suraeNoether> i should have a draft of the monero portion of this paper done by the end of November
<sarang> sexxy
<suraeNoether> sarang has already seen it and (if i do say so myself) it's a pretty fun paper actually
<sarang> Very much so
<suraeNoether> if it wasn't for the fact that's *criticizing* monero just like MRL-0001, i'd be very excited about writing it
<sarang> Have you informed Zcash of the possible applicability to their project, so they can draft a blog post about it? =p
<suraeNoether> i'm confident that this will make monero better, however, by formalizing some of our concerns that were previously only qualitative
<suraeNoether> sarang not formally, i may have brought it up with zooko in person before
<sgp_> any way I can get a draft copy?
<sarang> We should be good neighbors and let them know
<sarang> sgp_: it's all super early
<suraeNoether> sgp_: yes, i actually want your thoughts becauset his is relevant to pools and mining and i want your recommendations, too, sgp
<sgp_> cool
<suraeNoether> so, that's my MRL update. Sarang, what have you been up to?
<sarang> A few things
<sarang> First, housekeeping for monero-site updates to migrate our papers over, and add the new ones
<sarang> Second, more lit review on graph theory approaches to anonymity
<sarang> Third, reviewing some other papers relating to ring sigs and zk proving
<sarang> Fourth, fixing and writing up a cross-curve discrete log equality proving system
<sarang> there's also working toy code for that, using ed25519 and ed448
<sarang> This allows you to use "discrete log preimages" across curves or groups arbitrarily
<sarang> provided the spaces are big enough (they are)
<sarang> Here's the writeup, for those curious: https://v2.overleaf.com/read/jcyscybzhzmy
<sarang> Note that I did not invent this, but this is the first correct writeup I've seen and I wanted one for completeness
<sarang> That's about it for me
<suraeNoether> sarang has been on fire btw
<sarang> As in, running around with my hair on fire, sure!
<suraeNoether> if you guys haven't noticed, he's implemented several toy implementations of various crypto schemes in the past two months alone
<suraeNoether> you have hair?
<sarang> Not anymore, it burned off
<suraeNoether> nice
<sarang> So going forward, this paper (and others) will be on getmonero.org
<sarang> A few newer papers are in the PR pipeline
<sarang> Questions on any of my stuff?
<suraeNoether> the only other topics on my mind are only half research-related: 1) research: the post-thanksgiving Monero face-to-face being hosted by Tari bringing sarang and endogenic and i together in Nashville again... I heard a rumor someone else important was coming :D 2) not really research: the coin center privacy workshop in December I'm considering attending, 3) research: the Monero Konferenco, and...
<suraeNoether> well, the last bit is related to my nonprofit whihc is a selfish thing to bring up so I'll leave that alone
<endogenic> too l8
<endogenic> bring it up anyway
<suraeNoether> heh
<suraeNoether> okay
<suraeNoether> so, for 1) as folks may know, Tari has paid the expenses required to get sarang, endogenic, and i face to face for a meeting before... and they are doing it again, and it is looking like it may become a quarterly thing
<sarang> I suspect funding may be stalled a bit while belts tighten :/
<endogenic> oh btw i drive in
<suraeNoether> this is largely a research powwow over a few days, at the last meeting this whole bipartite matching thing was initially estimated and it kicked off my current research paper
<suraeNoether> and i should say: it's not just Tari, it's also MyMonero
<endogenic> one of these next ones i'm going to try to get surae to stay at my house ^_^
<suraeNoether> We aren't burdening the community financially with these meetings, but we also want the community to learn of our financials in this way, to prevent accusations of opacity
<suraeNoether> For 2) I really want to go to this thing on some levels, but i am concerned coin center is going to look to me as a voice of the Monero community. I'm not sure if they want me to come if I'm coming as a private individual not as an individual representing Monero
<suraeNoether> Isthmus was already nice enough to offer a place to stay for me in SFO so the only financial cost would be a plane ticket
<suraeNoether> it seemed to me like folks were lukewarm on the idea of me attending last time I brought it up
<suraeNoether> trying to sense the temperature now that folks have had a week or two to sit on the idea
<sarang> In terms of funding, it seems to have more value for the space as a whole, rather than just for the Monero community
<sarang> I suspect you're right about the "voice of Monero" thing, but I don't know how bad of a thing that is
<suraeNoether> yeah, and I would go on my own dime
<sarang> I remain disappointed that this is a "be in person or don't have a say" thing
<sarang> but that's neither here nor there
<suraeNoether> for 3), the Monero Konferenco: we are sitting almost at 20% funded. https://forum.getmonero.org/8/funding-required/90909/surae-noether-first-denver-monero-konferenco-spring-2019
<sarang> that's impressive
<suraeNoether> i know, right?
<suraeNoether> that'll be enough to put a deposit down on a location
<sarang> I assume you'll wait until funding is closer to guaranteed before deposits?
<sarang> to the extent possible, that is
<suraeNoether> i'm concerned about waiting until it's totally funded for stuff like that, and I can't think of an easy quick solution. maybe rehrar has some thoughts
<nioc> suraeNoether: I see that there are no milestones in the FFS for payout
<suraeNoether> nioc yeah, we should consider how to structure that asap
<suraeNoether> because a milestone like putting a deposit down on a location requires the money before the milestone occurs
<sarang> I think nioc's point is good particularly because of the natural payouts that are needed for this
<sarang> yeah
<suraeNoether> so perhaps we invert the milestones
<sarang> suraeNoether: what's the downside to waiting, besides the risk of losing venue?
<suraeNoether> i could make a milestone post to request funding to complete a milestone
<sarang> I think that's fine
<suraeNoether> sarang: volatility in price over the long term makes the actual funding receive much more variable
<sarang> There's no independent way for donors to verify the milestones happened anyway
<sarang> unless you posted receipts and such
<suraeNoether> we'll be posting contracts signed with venues and receipts, etc, all of which are very easily faked, unfortunately
<sarang> Yeah, that's unavoidable
<sarang> I think people understand this
<suraeNoether> curious parties could always call the literal venues themselves and check, I suppose
<sarang> Goal should be to maximize transparency and accountability within the limits of the unique circumstances
<suraeNoether> so, how shall we go about doing this? Should I edit the funding request post to include all this information? I feel like that's changing the terms of the request after we received donations already, which isn't necessarily fair to the previous donors
<sarang> As long as you're doing the same things with the money, updating for more clear scheduling seems entirely reasonable
<nioc> sarang: I believe it's up to core not the community to verify receipts
<nioc> since they release the funds
<sarang> nioc: I only mean this in the sense that most funding requests have a tangible, publicly-verifiable work output
<suraeNoether> maybe fluffypony luigi1111 binaryFate or ArticMine could weigh in on this.
<sarang> Whereas this is a bit different
<nioc> not that you couldn't make it public
<suraeNoether> thanks for that observation, sarang
<suraeNoether> okay, i'm going to edit the current funding request post to include a handful of milestones and a description of how we are going to invert the milestone process for this event
<sarang> I think a payout with a clear understanding of what happens with it (e.g. venue deposit) and some kind of immediate transparency for a modicum of verification (e.g. invoices and core team verifies somehow) makes sense
<sarang> Donors likely already implicitly assumed something along these lines
<sarang> The "work output" is a conference next year :)
<luigi1111> I would look for community agreement if there seemed to be anything "shady"
<sarang> Plus suraeNoether already has to have evidence of this for corporate tax purposes anyway
<sarang> cool cool
<suraeNoether> one milestone can be a deposit on a venue, AV stuff, caterer and (if appropriate) a deal at a hotel so attendees can get a discount. another milestone can be purchase of flights and hotels for speakers. a final milestone can be for the remainder of the cost of the event to pay for things like media, publicity, printing pamphlets and posters, assembling shwag bags, etc
<suraeNoether> thanks for the input luigi1111 i believe you are 100% correct
<parasew[m]> we will be testing a few things at the 35C3 (Monero at the Chaos Communication Congress 27.-30.dec) which could be valuable for the conferenco: submission management, streaming, etc. will let everyone know when there is something to see!
<suraeNoether> we mentioned having defcon-style badges from the hardware team, but i think we are going to hold off on those until the second year. this will keep our costs down and allow the HW team to focus on the wallet, etc
<sarang> ^ good idea, on both counts
<suraeNoether> parasew[m]: regarding the 35c3 conference, if sarang sgp and myself all want to come (I do!) we need to make our post for travel funds this afternoon
<suraeNoether> i'm holding off on renewing my passport until after it so i don't have to worry about not getting my passport back in time
<suraeNoether> rehrar sarang and sgp_ are you guys still interested in going to 35c3
<suraeNoether> ?
<parasew[m]> suraeNoether: sure! yes! (my planning got heavily delayed but the stage and everything got confirmed yesterday)
<sarang> I was just checking my schedule yesterday, and it simply will not work for me due to family commitments
<sarang> (the timing of the event is awful)
<parasew[m]> timing indeed is a problem
<suraeNoether> it really is. flights are super expensive on the 26th and 25th in general
<sarang> I had hoped that I could work around the family stuff, but it's not possible
<sarang> Plus my brother, sarangbro, is expecting a kiddo during that week
<sarang> very exciting
<luigi1111> nice name
<sarang> ikr
<luigi1111> we welcome sarangbrokid
<sarang> in olden tymes, they'd have invented a new last name, like sarangson
<parasew[m]> best wishes to sarangbro+sarangbro_junior! :)
<suraeNoether> yes :D
<sarang> BTW, IACR has been chock full of interesting relevant papers lately
<suraeNoether> yes. yes it ihas.
<sarang> I have a long list for this week
<sarang> I try to hit up lit review weekly but some weeks it gets just bonkers there
<suraeNoether> btw
<suraeNoether> everyone, i really think sarang needs a vacation
<suraeNoether> like five days of no computer screens and some sun or something, and he barely takes weekends off
<suraeNoether> sarang is a very driven person
<sarang> psh, do any of us?
<suraeNoether> ikr
<sarang> It's like Newman, who once opined that the mail never stops
<suraeNoether> thing is, you see Ethereum people on twitter bragging about working 18 hours a day and you know they are producing some straight up crap in those conditions.
<sarang> I suspect many of us will be effectively taking time off around Christmahannukwanzaka
<sarang> for one reason or another
<suraeNoether> OH there is one FFS i would like to direct everyone's attention to
<sarang> ?
<suraeNoether> TheCharlatan is proposing development of reproducible builds here: https://forum.getmonero.org/6/ideas/91098/funding-for-development-of-reproducible-builds
<sarang> outreach?
<suraeNoether> I strongly support this FFS project, and I think it would be a nice security gain for Monero
<sarang> These have been desired for a while
<sarang> So the whole request is for 6 XMR?
<suraeNoether> yep, and i don't think it's out of line with endogenic's recent efforts to encourage reworking/refactoring Monero (i'm almost certainly mischaracterizing Endo's goals)
<suraeNoether> it appears he's only asking us to pay for his VPS
<sarang> I don't know the details of repro build complexity, but that seems like a great deal
<sarang> unless we can get VPS support for free, as some others had indicated
<suraeNoether> ^ he works at Shift, the hardware wallet developer, and he started gridcoin, and admittedly wildly insecure but super fun cryptocurrency experiment
<suraeNoether> or at least, when i met him, he was doing work with Shift
<suraeNoether> anyway
<suraeNoether> I have no further information to report
<suraeNoether> although I'm always interested in getting community feedback in general
<sarang> cool
<sarang> Any other fun news to report
<sarang> ?
<suraeNoether> aha, so MAGIC received its first non-board member donation today, so I'm totally energized to encourage folks interested in contributing to MAGIC to check out what we are about at https://www.magicgrants.org ... we are an educational and scientific non-profit focused on scholarships and research grants in cryptocurrencies.
<suraeNoether> our scholarship program will be open starting in January and we are currently fundraising for next year
<suraeNoether> if we were in a bubble I'd say "hey, come reduce your tax burden possibly" but I'm fully aware of the current state of the market. :P
<sarang> lol
<parasew[m]> this is amazing, congrats for magic!
<sarang> I must take off shortly to meet up with someone
<suraeNoether> thank you! hopefully we can reduce the financial burden of studying cryptocurrencies for students while also incentivizing universities to make cryptocurrency curricula
<suraeNoether> and eventually? build primary schools, libraries, and computer labs in the developing world
<suraeNoether> can't have a crypto infrastructure without comptuers (technically a false statement, but theory and practice disagree here :P)
<suraeNoether> but since that's not research related and is a coin-agnostic project, it may be one of the last times I bring it up during an MRL research meeting
<suraeNoether> okay
<suraeNoether> i believe we are good to go on today's meeting
<suraeNoether> EVERYONE. you must know this: i love you