do monero devs have any clue how kripos monero tracing works? is it suspected to be different from a dust attack?

<gan:skhron.org> I thought it's already established that Monero isn't really vulnerable to dusting attacks?

CEO_of_Linux: It could be real, or it could be the feds using this opportunity to create FUD and bluff again

Cindy_ i don't think they can bluff in the court case tho. they need to show how they obtained the evidence

Well there is no court case

At least I haven't found one

I'm betting 99% it's just crappy opsec and timing attacks on exchanges

This isn't the first time they lied out of their ass, they like doing it so that they can create some FUD around technology and discourage people from even trying

seems a bit random for norway to do that

<gan:skhron.org> CEO_of_Linux: You've read the news beyond its headline?

Cindy_ there were arrests made, i presume there will be a court case sooner or later

but it's not amazon

There are other ways of deanonymizing people than just attacking Monero

I wrote my theory above

It's like having a giant fortified gate but your walls are made out of cardboard

I think that they have some kind of limited attack and found a clever way to use it

yanmaani: What's your theory?

I wasn't on here to see it

Does it have to do with the coinbase outputs

coinbase doesn't list monero

Coinbase as in the first transaction where a output is created

Not the company

ah sorry

Cindy_: If you look what they write, they use a very specific phrasing: "track the use of Monero in specific cases. This method ... enabled us to identify individuals who paid ... Kripos has also identified two sellers"

They are clearly stating that they were able to identify buyers

While strongly hinting that the sellers were caught for different reasons

So the sellers got the $5 wrench?

The other thing is, maybe it's just how it got translated but

"Track the use of monero"?

Why not "track monero" or "track monero transactions"

I think it has to do with the site they were using

So my guess would be maybe some kind of side channel, maybe network related, OR

something like, first they spent $10 on a site

then spent $10 elsewhere

Then spend $10 on that site again

Whoever runs that site can now make a very strong guess on which ring member was the real transaction.

Ah yeah true

I'm not saying I'm confident of my theory, but it logically fits the details of their claims

How would a side channel occur?

Cindy_: Well if for example you broadcast transactions on clearnet, you do not use a VPN, and they simply monitor your network traffic

that sort of thing

I mean wouldn't the timing be vastly different from when they made the offer?

Oh

Classic chainlytics attack

Getting the txid and then tracking their IP along with it

there are also the malicious nodes that were discovered which afaik nobody knows yet what kind of attack they were being used for

well they also have more resources; they could theoretically do much worse

That would make sense if they also had the sellers too

like "select * from isp where port = 18081"

It would be plausible really, we're talking about pedophiles

and then just look at who makes outgoing data transactions on that port a few minutes before a transaction of interest enters the mempool

They aren't really the smartest tool in the shed

yanmaani: How would you know if someone made a transaction or was just trying to fetch new blocks?

I'm assuming node traffic is encrypted (it would be dumb if it wasn't)

So the signal would be polluted with block update requests

Cindy_: well, a block download goes in the other direction :)

Yeah I know but I mean users polling a node for updates

Like "what block number are you on?" Over and over again

those requests should be smaller, no?

you could also correlate it with the exact size of the transaction

different package sizes, handshakes, timing, etc.

as a transaction won't really compress well

^

True

there are defenses against it, im not aware of such techniques actually being used anywhere

but norway is a small country, i think ISP market is very concentrated

mullvad has DAITA for that

state owns major telecom

I think TLS has some padding btw

For the connection data

they also do some intense stuff "to protect the children"

this is all assuming their wallet was connected to the clearnet. but they were obviously using tor, at least to access the dark web

I don't know if it's to the extent to the size of an average Monero transaction

CEO_of_Linux: DAITA is just connection padding

yeah, it's just a theory. if you want to get REALLY paranoid you can of course say, well if you look at all these tor users here, and then we look at who was connected when X Y and Z happened ...

i mean it can't be more than few thousand people right?

but this is really too depressing to think of

because even though it';s an admirable goal it implies mass surveillance has gone so far that you can just ...

Yeah

nah Tor has a lot more users. all cybercrime operates on it

On another topic (and I know I probably repeated it): payto:// now supports Monero

I thought the GNU Taler guys would not like it, but uhh

They didn't care much

I'm a bit confused tbh. because it's obvious LEA has some insane capabilities. they were insane in the snowden leaks, they only got better since then. but there seem to be very few busts compared to the horrifying amount of crime

i'm not really sure what's going on. there's some cognitive dissonace in all of this

i remember a few months ago there were two guys who threw a grenade inside a cafe in Paris then ran away. they live streamed the whole thing. they never got caught

how is that possible

Incompetence

people shoot each otehr while livestreaming with their face and get away with it. and then there was the actually competent guy who killed the health insurance ceo

gotta use the good stuff to protect the ceos

I hate how the EU gov completely nerfed GNU Taler

It used to be privacy for the buyer, but now it has KYC and age verification

just reading on taler now. it seems like it provides privacy only in so far as you trust the taler exchange to implement the blind signature protocol correctly?

any entity that interfaces with the fiat banking system is going to have to do KYC

Not like it really matters because of the KYC

Taler is just an extension of the fiat banking system, not a cryptocurrency

A payment network/system or something

Yeah really curious to hear more about that because every time I heard about "tracing Monero" (like in the Chainalysis leak) there was a "gotcha".

I recall one case where the culprit was exchanging exact amoungs on a cooperating swapper

<jeffro256> Regarding the USA, the Roman Sterlingov case demonstrated that BS companies can hand wave away the rigorousness of their techniques and just tell the jury "trust me bro, the clusters are speaking to me". Also IDK about standards of evidence in Norway, where Kripos is HQ'ed. > <CEO_of_Linux> Cindy_ i don't think they can bluff in the court case tho. they need to show how they obtained the evidence

<jeffro256> As far as it has been applied so far in the US, BS companies actually can sort of bluff.

Cindy_ | It used to be privacy for the buyer, but now it has KYC and age verification <- I recall watching a presentation on Taler and tbh I still don't entirely understand it. So what point is there now?

<jeffro256> So far BS on transparent chains like BTC fails the Daubert Standard, but I guess that US courts don't care about that anymore

the BS theory would work better for an individual caught using parallel construction. but this was a large operation that conssitently gave them multiple results

On one hand, I REALLY want for something less resource-intensive than crypto to fill the "non-KYC payment" niche, but at least this implementation is not it. I wonder if there's any way forward with the core technology?

there are cryptos which aren't resource intensive. but they're not as robust i guess

Like?

the ones that use proof of stake

Ah, that. Yeah, that's kinda concerning. Also, is the resource intensivity that much lower? Either way, that seems so easily controllable so idk if there even is any point in that direction...

it's much lower yeah

Like, I still hope that we figure something less resource-intense eventually.

<jeffro256> Alledegely... > <CEO_of_Linux> the BS theory would work better for an individual caught using parallel construction. but this was a large operation that conssitently gave them multiple results

if i have a wallet seed and no key images or knowledge of whether any funds were ever present on it, would putting it in the gui show me all historical incoming transactions to it or would i have to manually force the gui to calculate all possible addresses in all possible accounts ((2^32)^2 i think?) for it to scan the blockchain for old incoming tx's to any of them?

<rbrunner7> chiselfuse: If it's the seed of a wallet that was used "normally" it's enough to just restore and let the GUI wallet app do its work; just make sure to give it the correct restore height (before any transactions happened) and it will find transactions to the main "4" address and also transactions to subaddresses in any accounts that were used.

rbrunner7: really? how does that work?

<ofrnxmr> The lookahead generates/scans the first 50 account and 200 subaddresses of each of them

<rbrunner7> Not sure what you mean with "really". If nothing really special happened with the wallet, accounts and subaddresses with low numbers were used, and it will catch those

<ofrnxmr> If funds are found on any if those, the lookahead is offset by the last address or account with funds on it

ofrnxmr: and if funds aren't found it stops after 50x200 accounts?

<ofrnxmr> yes

ofrnxmr: but isn't that unreliable? what if someone had used account 51 and none of the preceeding ones

<ofrnxmr> Then you change the lookahead before scanning the block those txs were in, or when you create/restore the wallet

is there an option for that in the gui or do i use the cli

<ofrnxmr> Use the cli

<ofrnxmr> Easiest way is to use the commandline flag when restoring the wallet

ofrnxmr: which commandline flag are you talking about

<ofrnxmr:xmr.mx> Are you on linux?

<ofrnxmr:xmr.mx> ./monero-wallet-cli --help | grep subaddress

<rbrunner7> "but isn't that unreliable?" In theory, yes. As a prank, or to scam you, somebody can send funds to subaddress #1,000,000. You will never find those funds unless you get to know somehow which index was used.

<rbrunner7> But under normal circumstances this all does not really apply and you could as well worry that in the next second an asteroid will hit your house, and that's the end of the story ...

ofrnxmr: oh, i thought you were talking of a cli flag to the gui

rbrunner7: yea i guess. i still wonder how expensive it would be to scan (2^32)^2 addresses just to be exhaustive

afaik the account is int32 and so is the address index

<rbrunner7> I am pretty sure that with the way it's currently implemented your RAM would not be enough to hold the necessary lookup tables, and even if, preparing those tables might take days.

<rbrunner7> Seraphis (which won't get implemented) had a way to solve this problem, interestingly, but the necessary data was one of the reasons Seraphis addresses would have been very long.