Add P2A BIP

instagibbs commented at 3:40 pm on September 24, 2025: member

For interoperability with other schemes, it behooves me to write a BIP for this output type. It’s already standard to spend in well over half the network.

h/t roasbeef for bothering me about this repeatedly

jonatack added the label New BIP on Sep 24, 2025

in bip-paytoanchor.mediawiki:28 in 490aa62c71

23+The "anchor" output type is a commonly used pattern in Bitcoin layer 2 systems such as Lightning Network and beyond. This pattern allows transactions which are presigned far in advance to leave a "hook" for CPFP fee bumping later, when prevailing feerates for block inclusion are better known.
24+
25+When the wallet user is unconcerned by *who* is able to bump the transaction fees as long as someone does, this output type may confer a couple of advantages:
26+
27+1. Using an output type with minimal output and spending input footprint saves fees on average
28+2. Keyless operation aids in interoperability of services like watchtowers as there is no privelaged key material for bumping

jonatack commented at 9:13 pm on September 25, 2025:

02. Keyless operation aids in interoperability of services like watchtowers, as there is no privileged key material for bumping,

instagibbs commented at 2:48 pm on October 7, 2025:

done

in bip-paytoanchor.mediawiki:50 in 490aa62c71

45+
46+====Related Work====
47+
48+There are a number of related but separate historical concepts:
49+
50+[https://github.com/bitcoin/bitcoin/pull/30239 Ephemeral Dust]: Relay policies allowing a single dust output as long as it is spent in a transaction package. As an implementation detail, dust for P2A is set at 240 satoshis by default, others dust levels remain unchanged, and the dust-having transaction must be zero fee.

jonatack commented at 9:21 pm on September 25, 2025:

0[https://github.com/bitcoin/bitcoin/pull/30239 Ephemeral Dust]: Relay policies allowing a single dust output as long as it is spent in a transaction package. As an implementation detail, dust for P2A is set at 240 satoshis by default, other dust levels remain unchanged, and the dust-having transaction must be zero fee.

Roasbeef commented at 11:30 pm on September 25, 2025:

I think it’s also worth specifying above that P2A outputs can be zero value as well.

Roasbeef commented at 11:31 pm on September 25, 2025:

Have you considered making this BIP slightly more generalized, and also preicelyf defining the acceptance rules for ephemerl dust?

Eg: one question that’s popped into my mind several times is: can ephemeral dust only be used with v3 transactions?

Roasbeef commented at 11:36 pm on September 25, 2025:

Dust threshold instead of just dust?

As in, values before 240 for P2A are considered dust.

instagibbs commented at 2:48 pm on October 7, 2025:

done

instagibbs commented at 2:49 pm on October 7, 2025:

done

instagibbs commented at 2:50 pm on October 7, 2025:

Ephemeral dust ended up being very implementation-specific, and likely to change in some future.

Wondering if there is a more BOLT-like living document suitable for that.

instagibbs commented at 2:50 pm on October 7, 2025:

fixed

in bip-paytoanchor.mediawiki:36 in 490aa62c71

31+
32+===Specification===
33+
34+A P2A output is defined as one with the exact witness output script
35+
36+<code>OP_1 <0x4e73></code>

jonatack commented at 9:24 pm on September 25, 2025:

Perhaps add a footnote describing “what is the 0x4e73 for?” per your answer in https://github.com/bitcoin/bitcoin/pull/30352#issuecomment-2215089061

instagibbs commented at 2:48 pm on October 7, 2025:

done

in bip-paytoanchor.mediawiki:73 in 490aa62c71 outdated

68+Gloria Zhao,
69+Pieter Wuille,
70+Jeremy Rubin,
71+and Bastien Teinturier.
72+
73+==References and Rationale==

jonatack commented at 9:30 pm on September 25, 2025:

TBD, or remove this header?

Roasbeef commented at 11:40 pm on September 25, 2025:

Should link to the bitcoind PR that implemented this.

Roasbeef commented at 11:43 pm on September 25, 2025:

https://github.com/bitcoin/bitcoin/pull/30352

instagibbs commented at 2:48 pm on October 7, 2025:

one is added, leaving

instagibbs commented at 2:50 pm on October 7, 2025:

Was in “Implementation” section

in bip-paytoanchor.mediawiki:58 in 490aa62c71 outdated

53+
54+[bip-0431.mediawiki TRUC transactions]: TRUC transactions bound the topology including child transaction size, which aids in anti-pinning for the P2A output, as anyone can attach a child transaction including an oversized, low feerate one. There is no ineherent dependency on this concept and P2A.
55+
56+====Backward Compatibility====
57+
58+Creation of P2A outputs has been considered standard since the deployment of Segwit softfork on the network.

jonatack commented at 9:33 pm on September 25, 2025:

Maybe link here to the SegWit BIP section making them standard.

instagibbs commented at 2:48 pm on October 7, 2025:

the BIPs aren’t standardness BIPs, so I’m not sure that helps.

jonatack commented at 9:38 pm on September 25, 2025: member

Quick first pass.

in bip-paytoanchor.mediawiki:30 in 490aa62c71

25+When the wallet user is unconcerned by *who* is able to bump the transaction fees as long as someone does, this output type may confer a couple of advantages:
26+
27+1. Using an output type with minimal output and spending input footprint saves fees on average
28+2. Keyless operation aids in interoperability of services like watchtowers as there is no privelaged key material for bumping
29+
30+with the drawback that selection of this output type could result in griefing if anti-pinning measures are not taken.

Roasbeef commented at 11:28 pm on September 25, 2025:

Recommend that if unsure, then this new type is best used w/ v3 transactions are they have stricter replacement rules?

instagibbs commented at 2:49 pm on October 7, 2025:

done

in bip-paytoanchor.mediawiki:38 in 490aa62c71

33+
34+A P2A output is defined as one with the exact witness output script
35+
36+<code>OP_1 <0x4e73></code>
37+
38+corresponding to the address <code>bc1pfeessrawgf</code> on Bitcoin mainnet.

Roasbeef commented at 11:35 pm on September 25, 2025:

Should also specify the encoding for: regtest, signet, and testnet. Would make for good test vectors.

instagibbs commented at 2:50 pm on October 7, 2025:

done

Roasbeef commented at 11:37 pm on September 25, 2025: contributor

Thanks for making this. Nearly done with an implementation for btcsuite.

Add P2A BIP 9ed387c703

instagibbs force-pushed on Oct 7, 2025

Add P2A BIP #1982