Utilizing taproot to mask Bisq on-chain transactions

[viperperidot]

I am creating this as a placemark for further discussion on this topic. From what I understand, taproot will enable the ability for multi-sig transactions to appear as normal transactions. Currently the on-chain trade protocol for bisq is quite obvious to any third party observation. I am not a developer so I do not know what it would take to implement this, but I do think that it should be a high priority to improve privacy for the bisq trade protocol. If a bisq trade appeared like any other transaction on the blockchain, I think that would be an enormous win and remove the possibility for outside observers trying and link bisq users onions IDs or someone who starts trading on bisq with KYC coins, it would give the possibility to interpret the on-chain activity in many more ways rather than being obvious that someone it purchasing or selling on bisq. Although I do not know exactly how it will look on-chain with the taproot enhancements, from what I have heard it seems to be a great improvement and some developers i'm sure could speak more about this. I also don't know if this would be as large an undertaking as the segwit integration either. Any thoughts or insights on this would be appreciated, this is meant to be a discussion until an issue gets created if this idea is considered to be a good one.

[MaxHillebrand]

Yes, I think this is a priority requirement as soon as taproot is activated, the fee savings and privacy improvements are too numerous to ignore.

Just to get the conversation going, here a two basic ways taproot can be used to improve things:

• MuSig, replace script-based multisignatures with Schnorr public key and signature aggregation.
• Taptree, replace "single-line script" with a tree structure of spending condition, in the cooperative case use keyspend, in the uncooperative case reveal only one out of X spending conditions.

There are numerous more fundamental improvements to the whole Bisq trade architecture that are possible now [consider Lightning and DLC oracles too].

[viperperidot]

Yes this is a great point that I overlooked, there will be fee savings through taproot as well! I know we are going through a big transition in development right now with starting to work on Bisq 2.0 (for fee improvements) and there are lot's of different ideas floating around on that. Some ideas more complicated than others, maybe the fee savings with taproot would be our most actionable approach right now until the work on rebuilding the base layer protocol starts (which is probably more complicated?). It sounds like taproot can be activated as early as November from my understanding, so if the prelim work for adding it to bisq could be started now then potentially we could have it ready for then? That would likely be sooner than the other major overhaul ideas so at least we could get some wins on fee savings sooner. Also from what I have seen for Bisq 2.0, using bitcoin as the base layer will still be an option so all the taproot work will still be applicable moving forward. As far as all of the other more complex features that taproot allows, those sound very exciting from what Max has pointed out here. Not sure if any thought or research has be given to if the new coding abilities provided by taproot.. those might be enough to solve the majority of our issues without moving off the base layer. It would be cool if bisq was one of the first projects to put the full capabilities of taproot to the test. Maybe you can get down to a 1 transaction protocol with taproot by leveraging some of the smart contract functionality? Anyway I have lots of questions myself just as a user but i'm looking forward to see what all can be done with this, very exciting. :)

[Conza88]

It sounds like taproot can be activated as early as November from my understanding, so if the prelim work for adding it to bisq could be started now then potentially we could have it ready for then?

Mid-November, and a 1000* yes! Bisq with one of first Taproot, get lots of interest and signals we're 'on it' imo.

[MaxHillebrand]

If fee savings are the priority, then yes, tweaking the current protocol to use Taproot will already bring in a decent fee savings. I forget the exact percentage of tx vbyte decrease of 2-out-of-2 multisig. Probably roughly 50% reduction in script and witness size, which might take total fee savings to roughly 10%. [again, numbers are my rough guesses]

I think much more substantial fee savings can be accomplished when re-architecting the trade protocol to have less on-chain transactions in general. Cut it from 4 to 2 transactions, and you have a 50% fee-savings on the whole trade.

Here one project worth noting: @RubenSomsen's Succinct Atomic Swap idea utilizes pre-signed transactions and private key sharing, to reduce cross-chain swaps from 4 to 2 transactions. This comes with an online requirement for a certain period of time after the trade for one party. Since Bisq is somewhat designed to be [almost] always online, this might be a reasonable trade-off.

I added a minor contribution with externally funded SAS, where one further transaction can be saved by the actual swap participants.

[Conza88]

Mid-November Taproot activation, has been locked in...

[LeadManPL]

I just tried to send some change from BISQ and it says "The address is incorrect".
Its one year after taproot activation.We really need taproot support IMHO