PSA: MWEB is not a mixer

I’ve been seeing people talk about MWEB like it’s a way to mix funds. It’s not!!! The transaction graph is still visible you can see it at MWEB explorer (I can’t add links, but you’ll find it in Google). For example let’s take a recent peg-in from ltc1qaudvxr9k69pwsfnl02fhagay7ha5euxz6m9xly in transaction 4bedc5836d85a92b296463614dfcb89b81091d4c789fb34ce16464df9a14be3f. It’s in MWEB block 2,673,475, the only transaction (one kernel = one tx) as you can see in the MWEB explorer.

Down in the list of outputs, click “(Spent in block 2,673,479)”. Again, the only transaction in the block. Two outputs. One spent in block 2,673,481], the other in block 2,673,484. You can see that both 2,673,479 and 2,673,481 acted as inputs in 2,673,484, so it’s still the same coins. No other inputs were used. There’s a peg-out in each of the two latter blocks.

The peg-outs went to MCkJCGXL6CwpLbxjeCjuWAPPnMXKWPHPRR (28 LTC) and ltc1qug9y7sc925n5g6lv7lpxacv3zfqtjhjqk6yf70 (20 LTC). Here we traced the coins going from ltc1qaudvxr9k69pwsfnl02fhagay7ha5euxz6m9xly to those addresses. They either transacted between each other or are the same person. No inputs from other peg-ins were used, meaning this link is 100% deterministic!!!

You can keep going through “spent in block” in block 2,673,484 and see more peg-outs. Whoever this is is probably either an exchange or mixing money! But it doesn’t help!

MWEB is only useful if you keep your money inside it, it doesn’t work as a mixer, because the transaction graph (who sends money to who) is still public even if the amounts are hidden!

1 Like

Thanks for the clarification! It’s important to understand that MWEB doesn’t mix funds; the transaction graph is still visible. Transparency remains key even with hidden amounts

Im trying to understand this.
Can you explain why you think the output 2,673,481 and 2,673,484 belongs to the same person, cannot it be that someone else is spending LTC?
As I understand we can fallow transaction because we have just 1 or 2 input and outputs.

To my understanding if we would have 10 page ins and 10 page outs in one block, then it would be not possible to connect transactions? of course you can assume, because the number is small, but if number increases then it becomes hard task.

For this reason Im waiting for the wallet on the phone with mweb support. My expectation is that number of transactions will increase.

It’s the common input ownership heuristic, same way chain analysis companies work. It could be two people, but if it is, they’re related to each other in that one of them sent money to the other! Similar to how any transaction on the main (public) chain can have several interpretations. My point is, the transaction graph is public!

Well if you’re just looking at the blocks then yes! But if I’m not mistaken, MWEB doesn’t do any transaction aggregation on the wire. So each individual transaction can be seen in the mempool before it’s mined! You’d have to run your own node to do that analysis but once you have one it’s fairly trivial!

Grin for example tries to protect from that by doing cut-through on the node level, before the transaction is sent to the public mempool. But even there by running a node you can uncover 96% of the graph!! Since Litecoin doesn’t do any obfuscation that number would be 100% for MWEB! You can find the famous article called “Breaking Mimblewimble’s Privacy Model” it explains how that works.