Payjoin sends #608

DanGould · 2023-10-19T19:17:27Z

supersede #572
fix ci errors
awaiting mutiny-node merge as a dependency for this

Relies on MutinyWallet/mutiny-node#647 and MutinyWallet/bitcoin-waila#8

Test with @DanGould's payjoin receiver
visit https://mutiny.payjoin.org:4433/bip21?amount=0.00066666 to get a bip21

waila part

node part

-- @benalleng

benalleng

In my testing I got it to work twice!
https://mutinynet.com/tx/3c5436f1edf7d4c32a5ccf2448c1e963f52bb8a0fb6f8688d7e78a14e1cbe80b
https://mutinynet.com/tx/ca2bca759e84730e8bd995b98edbcce0df0bea26965c39d7293cdae2bae65206

After those 2 however, I continued getting a 400 bad request error
XHRPOST https://mutiny.payjoin.org:4433/?v=1&additionalfeeoutputindex=1&maxadditionalfeecontribution=59&minfeerate=1.008 [HTTP/1.1 400 Bad Request 1532ms] { "errorCode": "original-psbt-rejected", "message": "The receiver rejected the original PSBT." }

~~probably has to do with my hacky fixes to the types in the node where I made ConfirmationTarget::OnChainSweep? not 100% sure~~

DanGould · 2023-11-01T15:45:24Z

In my testing I got it to work twice!
https://mutinynet.com/tx/3c5436f1edf7d4c32a5ccf2448c1e963f52bb8a0fb6f8688d7e78a14e1cbe80b
https://mutinynet.com/tx/ca2bca759e84730e8bd995b98edbcce0df0bea26965c39d7293cdae2bae65206

After those 2 however, I continued getting a 400 bad request error
XHRPOST https://mutiny.payjoin.org:4433/?v=1&additionalfeeoutputindex=1&maxadditionalfeecontribution=59&minfeerate=1.008 [HTTP/1.1 400 Bad Request 1532ms] { "errorCode": "original-psbt-rejected", "message": "The receiver rejected the original PSBT." }

@benalleng

I don't think this failed because of your hack, but because for whatever reason requests after your last success reused the same input 3c5436f1edf7d4c32a5ccf2448c1e963f52bb8a0fb6f8688d7e78a14e1cbe80b:0. That input was banned to prevent probing of the receiver's UTXOs. Does mutiny employ a mechanism to lock UTXOs from being added to subsequent PSBT constructions? If so, perhaps the input in question was unlocked after a wallet restart, then once it tried to spend again and failed it couldn't get past the ban.

One way to mitigate this would be to add a switch that defaults to payjoin but allows you to spend in a naive way by switching off after pj failure. The receiver could also include a more detailed "message" to post to the debug log that says "input {txid:vout} seen before, original PSBT rejected to prevent a probing attack."

benalleng · 2023-11-01T16:24:11Z

src/routes/Send.tsx

+                    sentDetails.txid = txid;
+                    sentDetails.fee_estimate = feeEstimate() ?? 0;
+                } else if (payjoinEnabled()) {
+                    const txid = await state.mutiny_wallet?.send_payjoin(


should we add a try-catch block to enable that fallback to normal send?

try { const txid = await state.mutiny_wallet?.send_payjoin( destination()!.original, amountSats(), tags ); } catch { const txid = await state.mutiny_wallet?.send_to_address( address()!, amountSats(), tags ); }

though this brings up the comment you made in the mutiny-node about where the waila should parse the receive string as some parsing would be needed between destination.original() and address()

easy enough to just regex away the unimportant stuff, but still worth considering

I'm not sure we should assume someone wants to spend as naive tx if they can avoid it. some automated payment processors will automatically spend the naive tx if payjoin fails but senders may want to try another utxo or choose something else like ln instead of immediately sending w/o a privacy preserving measure

right, that makes sense

benalleng · 2023-11-01T16:26:57Z

Was it a user error? Did I send too rapidly? what do you think was the flow that caused the ban, if this is a mutiny-wallet bug wonder how we can get around it to make sure any user doesn't get banned as manual UTXO selection by the user is avoided

benalleng · 2023-11-01T16:53:08Z

reproduced the error
Got several successes before failure though!
https://mutinynet.com/tx/cd38e01043445e086f4289239909f8404bf809c1ce45ea25bce6281b6fb92de8
https://mutinynet.com/tx/828d97d3a2532121fbd8942d97e6225eb0529b5ba1c547f006b0e9c4541f4543
https://mutinynet.com/tx/494b5144b69cb67cec315080bc4b7e9fc88fcd42c7e96e2f7679c37b31096d53
https://mutinynet.com/tx/d1ebaf0e09a841f1cd61ab67729a94d096d8d6e0bc482e5e4982a07104e91db0

attempt 5:

got a frontend only error
Failed too sign unsure what caused it

attempt 6:

got the error i have seen before likely indicating I am banned, perhaps there is some relation to the error in attempt 5?

XHRPOST
https://mutiny.payjoin.org:4433/?v=1&additionalfeeoutputindex=0&maxadditionalfeecontribution=59&minfeerate=1.008
[HTTP/1.1 400 Bad Request 1695ms]

1

{ "errorCode": "original-psbt-rejected", "message": "The receiver rejected the original PSBT." }

attempt 7:

same error as attempt 6

DanGould · 2023-11-10T19:33:41Z

If a payjoin_psbt response failed to sign and a request was replayed, that utxo would be banned so this would be expected behavior. Do we have any more logs / detailed on why Failed too sign cropped up?

benalleng · 2023-11-13T20:51:37Z

Couldn't recreate the Failed to sign error this time and only got one success before failure, unsure of why its happening

DanGould · 2023-11-13T23:52:40Z

Does this issue require a playwright e2e test to proceed? Have you any logs from those failed runs?

DanGould · 2023-12-04T17:13:47Z

I believe this is now ready for review. @benalleng Do we know why OpenSSL failed on iOS here? I don't think it's related to this PR, as it adds no changes to OpenSSL crypto

benalleng · 2023-12-04T20:30:39Z

Yeah I brought it up to everybody on the matrix dev server, the consensus is that it is related to some secret that non-branch PR's / non-contributors don't have access to

benalleng · 2023-12-04T20:31:38Z

As for this PR I think it's ready to go, I know at the moment we are going to hold off on adding the optionality for the sender, i.e. the toggle to switch to naive send so we don't need to add that here

DanGould · 2023-12-05T18:54:53Z

I agree the toggle can come later. The same function can be had for now by pasting in the plain address instead of a &pj= containing uri.

futurepaul · 2023-12-12T17:31:07Z

I'm getting "Failed to create payjoin request" running this locally. It doesn't look like the request to https://mutiny.payjoin.org:4433/?v=1&additionalfeeoutputindex=0&maxadditionalfeecontribution=59&minfeerate=1.008 is getting a response?

DanGould · 2023-12-13T00:27:45Z

Oops, the public ip of that server was changed. I reset the DNS A record but it may take some hours to propagate. mutiny.payjoin.org should be 110.169.157.242

futurepaul · 2023-12-13T17:56:13Z

Hm still not working for me, same timeout for https://110.169.157.242:4433/?v=1&additionalfeeoutputindex=1&maxadditionalfeecontribution=59&minfeerate=1.008

DanGould · 2023-12-14T02:48:35Z

Thanks for your patience on this, my server was moved to one with a dynamic ip. It has been changed to a static one at 157.245.59.46 which mutiny.payjoin.org:4433 should point to. The https cert is not right yet, but I don't think mutiny-node's reqwest is configured without tls because default-features = false so it could still work.

However, ignoring TLS certificates is very dangerous and should be fixed immediately https://github.com/MutinyWallet/mutiny-node/blob/2c332b43269ade626b8c195fc0a3a7870c640b94/mutiny-core/Cargo.toml#L37 reqwest uses browser TLS in a wasm environment.

DanGould · 2023-12-14T18:44:05Z

https://mutiny.payjoin.org:4433 is running and secured by TLS now. Please excuse the wait 🙏

benalleng · 2023-12-14T21:32:01Z

getting an error when trying to test now: Can't broadcast. PSBT rejected by mempool. has there been a new change to the mutinynet mempool policy?

Edit: full error looks like it might be happening in payjoin server
400 error at https://mutiny.payjoin.org:4433/?v=1&additionalfeeoutputindex=0&maxadditionalfeecontribution=59&minfeerate=1.008
request: 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
reponse: { "errorCode": "original-psbt-rejected", "message": "Can't broadcast. PSBT rejected by mempool." }

benthecarman · 2023-12-14T21:38:30Z

getting an error when trying to test now: Can't broadcast. PSBT rejected by mempool. has there been a new change to the mutinynet mempool policy?

No, haven't in months

DanGould · 2023-12-14T22:07:29Z

Weird. When I run testmempoolaccept on the tx you submitted I get non-final error.

bitcoin-cli testmempoolaccept '["01000000000101ed05c0a2016487c310e082b6bfd4c0f4018c14de9b528d501d91a207b398dc980100000000fdffffff0225140d00000000002251207f809c1307a11ed63b852947c499789be8b1cbe3d50dc0614d92056ab0d2171f1027000000000000225120c8df14b2a973139855a61323a6fd9d2889a272fea87e2a274afa6ec84cf822930140fb6be4528387d682adc3e16bf7621b98883ded27af60b228916de6c417ff3f64b941bf88794c2a26a956257dddc8ac23b93fbd62c6a501fb7c36b7c0e0f1abe6ac250a00"]'
[
  {
    "txid": "875ed948c456d763225354658b12241af6ebf9df45bb11596b270d6d3f143c60",
    "wtxid": "b1b5bf107baa8155bda3f54bbdf90cdaefa1ad1616b14f9512890e1bb7a0cc48",
    "allowed": false,
    "reject-reason": "non-final"
  }
]

DanGould · 2023-12-15T02:25:16Z

I just sent a successful payjoin from this branch. The signet server there was not synced before after maintenence. It should be fixed now

futurepaul · 2023-12-18T23:38:28Z

Hate to be a continual bearer of bad news but I tried to send to your sample address and got another error:

benalleng · 2023-12-19T00:03:23Z

Can you see the network request response to see if it was an error communicating with the payjoin server or something else?

futurepaul · 2023-12-19T18:08:05Z

it's a 400 bad request:
{ "errorCode": "original-psbt-rejected", "message": "The receiver rejected the original PSBT." }

DanGould · 2023-12-19T18:17:43Z

The prior request that failed still included an input which was recorded as seen, so this subsequent request was rejected. I've deleted the seen_inputs database and restarted the server. Thanks for the persistence.

DanGould · 2023-12-22T17:45:13Z

src/routes/Send.tsx

+                    sentDetails.destination = address();
+                    sentDetails.txid = txid;
+                    sentDetails.fee_estimate = feeEstimate() ?? 0;
+                } else if (payjoinEnabled()) {


I'm not sure a sweep should necessarily preclude payjoin on the isMax condition. A receiver offered a sweep could still take the opportunity to contribute an input or use transaction cut-through, and otherwise just broadcast the original psbt as a basic sweep. It doesn't break this PR but should be considered for incremental change later.

futurepaul · 2023-12-27T18:35:26Z

Error processing payjoin response: couldn't decode PSBT: error in PSBT base64 encoding
{errorCode: "original-psbt-rejected", message: "Missing payment."} (400 bad request error)

request:
https://mutiny.payjoin.org:4433/?v=1&additionalfeeoutputindex=1&maxadditionalfeecontribution=59&minfeerate=1.008
payload:
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

DanGould · 2024-01-03T18:34:12Z

@futurepaul It looks like the original bip21 from the description belonged to an old wallet, so the output did not pay the receiver you were calling. I updated the description to let you fetch a bip21 from the live server which I'll repeat here:

https://mutiny.payjoin.org:4433/bip21?amount=0.00066666, which I have just tested on my local machine to success

DanGould · 2024-01-08T19:51:51Z

mutiny.payjoin.org is down for maintenance and should be up tomorrow

DanGould · 2024-01-09T20:07:31Z

mutiny.payjoin.org is up and synced

DanGould · 2024-01-11T20:32:40Z

rebased on master w/ new ActivityEditor change

benthecarman · 2024-01-22T21:20:48Z

Tested and it worked for me. The payjoin notice is a bit big compared to everything else, not sure what a better way would be though

DanGould · 2024-01-22T21:44:52Z

@benthecarman hell yeah brother

Happy to follow up with a toggle that fits in the table like bitcoin design recommends

benthecarman · 2024-01-22T21:46:05Z

Yeah that might be cleaner

DanGould · 2024-01-24T18:20:08Z

as of now this is the same element / size as other warning InfoBoxes

We could also spruce up the notice in VStack like this but I do believe that's out of the scope of this PR

benthecarman

lets just get this in

futurepaul

works for me! added one line so the balance shows up when an onchain request has an amount set (that was my bad)

DanGould marked this pull request as draft October 19, 2023 19:18

benalleng mentioned this pull request Oct 31, 2023

Payjoin sends #572

Closed

2 tasks

benalleng reviewed Oct 31, 2023

View reviewed changes

benalleng reviewed Nov 1, 2023

View reviewed changes

benalleng force-pushed the payjoin-sends branch 4 times, most recently from 0d13499 to b4cc22f Compare November 22, 2023 22:58

benalleng force-pushed the payjoin-sends branch 5 times, most recently from e47c2c0 to b9a089c Compare December 2, 2023 07:12

benalleng marked this pull request as ready for review December 4, 2023 20:32

DanGould commented Dec 22, 2023

View reviewed changes

DanGould force-pushed the payjoin-sends branch from b9a089c to b57eb4f Compare January 11, 2024 20:32

DanGould force-pushed the payjoin-sends branch from b57eb4f to 216c748 Compare January 16, 2024 22:04

feat: Add payjoin sends

736cf67

benalleng force-pushed the payjoin-sends branch from 216c748 to 736cf67 Compare January 18, 2024 19:34

benalleng mentioned this pull request Jan 23, 2024

Toggle Element for Payjoin send UX #847

Open

benthecarman approved these changes Jan 25, 2024

View reviewed changes

fix on-chain amount not showing

ab9c0c4

futurepaul approved these changes Jan 25, 2024

View reviewed changes

benthecarman approved these changes Jan 25, 2024

View reviewed changes

benthecarman merged commit 0324b49 into MutinyWallet:master Jan 25, 2024
3 of 4 checks passed

DanGould deleted the payjoin-sends branch January 25, 2024 16:16

DanGould mentioned this pull request Feb 8, 2024

Payjoin shows an absolute fee when a variable fee may be consumed #873

Open

Payjoin sends #608

Payjoin sends #608

Conversation

DanGould commented Oct 19, 2023 • edited Loading

benalleng left a comment • edited Loading

Choose a reason for hiding this comment

DanGould commented Nov 1, 2023 • edited Loading

benalleng Nov 1, 2023

Choose a reason for hiding this comment

benalleng Nov 1, 2023

Choose a reason for hiding this comment

DanGould Nov 1, 2023 • edited Loading

Choose a reason for hiding this comment

benalleng Nov 1, 2023

Choose a reason for hiding this comment

benalleng commented Nov 1, 2023 • edited Loading

benalleng commented Nov 1, 2023 • edited Loading

attempt 5:

attempt 6:

attempt 7:

DanGould commented Nov 10, 2023

benalleng commented Nov 13, 2023

DanGould commented Nov 13, 2023 • edited Loading

DanGould commented Dec 4, 2023

benalleng commented Dec 4, 2023 • edited Loading

benalleng commented Dec 4, 2023

DanGould commented Dec 5, 2023

futurepaul commented Dec 12, 2023

DanGould commented Dec 13, 2023

futurepaul commented Dec 13, 2023

DanGould commented Dec 14, 2023 • edited Loading

DanGould commented Dec 14, 2023

benalleng commented Dec 14, 2023 • edited Loading

benthecarman commented Dec 14, 2023

DanGould commented Dec 14, 2023

DanGould commented Dec 15, 2023

futurepaul commented Dec 18, 2023

benalleng commented Dec 19, 2023

futurepaul commented Dec 19, 2023

DanGould commented Dec 19, 2023

DanGould Dec 22, 2023

Choose a reason for hiding this comment

futurepaul commented Dec 27, 2023 • edited Loading

DanGould commented Jan 3, 2024 • edited Loading

DanGould commented Jan 8, 2024

DanGould commented Jan 9, 2024

DanGould commented Jan 11, 2024

benthecarman commented Jan 22, 2024

DanGould commented Jan 22, 2024

benthecarman commented Jan 22, 2024

DanGould commented Jan 24, 2024 • edited Loading

benthecarman left a comment

Choose a reason for hiding this comment

futurepaul left a comment

Choose a reason for hiding this comment

DanGould commented Oct 19, 2023 •

edited

Loading

benalleng left a comment •

edited

Loading

DanGould commented Nov 1, 2023 •

edited

Loading

DanGould Nov 1, 2023 •

edited

Loading

benalleng commented Nov 1, 2023 •

edited

Loading

benalleng commented Nov 1, 2023 •

edited

Loading

DanGould commented Nov 13, 2023 •

edited

Loading

benalleng commented Dec 4, 2023 •

edited

Loading

DanGould commented Dec 14, 2023 •

edited

Loading

benalleng commented Dec 14, 2023 •

edited

Loading

futurepaul commented Dec 27, 2023 •

edited

Loading

DanGould commented Jan 3, 2024 •

edited

Loading

DanGould commented Jan 24, 2024 •

edited

Loading