[POLICY] Make sending to future native witness outputs standard #15846

sipa · 2019-04-18T19:58:16Z

As discussed in the April 18 2019 IRC meeting.

This makes sending to future Segwit versions via native outputs (bech32) standard for relay, mempool acceptance, and mining. The reasons are:

This may interfere with smooth adoption of future segwit versions, if they're defined (by the sender wallet/node).
It violates BIP173 ("Version 0 witness addresses are always 42 or 62 characters, but implementations MUST allow the use of any version."), though admittedly this code was written before BIP173.
It doesn't protect much, as P2SH-embedded segwit cannot be filtered in this way.
As a general policy, the sender shouldn't care what the receiver likes his outputs to be.

Note that spending such outputs (including P2SH-embedded ones) remains nonstandard, as that is actually required for softfork safety.

maflcko · 2019-04-18T20:02:17Z

utACK c634b1e

jnewbery · 2019-04-18T20:53:22Z

IRC discussion here: http://www.erisian.com.au/meetbot/bitcoin-core-dev/2019/bitcoin-core-dev.2019-04-18-19.00.log.html#l-77

Concept ACK this change. I agree that the mempool should accept and relay txs with outputs to any segwit version.

It violates BIP173 ("Version 0 witness addresses are always 42 or 62 characters, but implementations MUST allow the use of any version."), though admittedly this code was written before BIP173.

This isn't relevant to the change here. BIP173 describes an address format and how to decode that into a scriptPubKey. The node/mempool has no concept of addresses.

I think your last three points are actually arguments to not stop the wallet from sending to segwit v1+ addresses, rather than arguments to allow the mempool to accept/relay those txs.

I'd prefer to prevent the Bitcoin Core wallet from sending to v1+ addresses, or at least to warn. We can't prevent all classes of sending insecure or unspendable addresses, but I think it makes sense to do so when we can.

I think the main concern that people have with that is that it might slow down segwit v1 adoption. I disagree - bech32 adoption has mostly been slowed down by large exchanges or other services not implementing send-to-bech32. That's something that is outside our control.

luke-jr · 2019-04-18T21:13:05Z

utACK

gmaxwell · 2019-04-18T21:33:44Z

utACK

harding · 2019-04-24T14:37:48Z

Tested ACK c634b1e

Separately tested a bc1p... v1 address with sendrawtransaction and the wallet GUI on mainnet. Transactions were accepted to the local mempool, relayed to a remote peer I controlled, and entered that peer's mempool (both peers running this PR).

(Hint for anyone else testing, if you want to empty your mempool so you can run abandontransaction, you need to restart with -persistmempool=0 -walletbroadcast=0, run the savemempool RPC, abandon your transaction either via the RPC or the GUI, and then restart again. If you don't savemempool, your node will restore the mempool from the run before you set persistmempool=0 and that mempool will contain your wallet transaction.)

gmaxwell · 2019-04-24T18:17:01Z

utACK

sdaftuar · 2019-04-25T14:25:34Z

utACK

instagibbs · 2019-04-25T17:29:51Z

utACK

meshcollider · 2019-04-27T09:47:49Z

utACK c634b1e

…standard c634b1e [POLICY] Make sending to future native witness outputs standard (Pieter Wuille) Pull request description: As discussed in the April 18 2019 IRC meeting. This makes sending to future Segwit versions via native outputs (bech32) standard for relay, mempool acceptance, and mining. The reasons are: * This may interfere with smooth adoption of future segwit versions, if they're defined (by the sender wallet/node). * It violates BIP173 ("Version 0 witness addresses are always 42 or 62 characters, but implementations MUST allow the use of any version."), though admittedly this code was written before BIP173. * It doesn't protect much, as P2SH-embedded segwit cannot be filtered in this way. * As a general policy, the sender shouldn't care what the receiver likes his outputs to be. Note that _spending_ such outputs (including P2SH-embedded ones) remains nonstandard, as that is actually required for softfork safety. ACKs for commit c634b1: MarcoFalke: utACK c634b1e harding: Tested ACK c634b1e meshcollider: utACK c634b1e Tree-SHA512: e37168a1be9f445a04d4280593f0a92bdae33eee00ecd803d5eb16acb5c9cfc0f1f0a1dfbd5a0cc73da2c9928ec11cbdac7911513a78f85b789ae0d00e1b5962

…utputs standard c634b1e [POLICY] Make sending to future native witness outputs standard (Pieter Wuille) Pull request description: As discussed in the April 18 2019 IRC meeting. This makes sending to future Segwit versions via native outputs (bech32) standard for relay, mempool acceptance, and mining. The reasons are: * This may interfere with smooth adoption of future segwit versions, if they're defined (by the sender wallet/node). * It violates BIP173 ("Version 0 witness addresses are always 42 or 62 characters, but implementations MUST allow the use of any version."), though admittedly this code was written before BIP173. * It doesn't protect much, as P2SH-embedded segwit cannot be filtered in this way. * As a general policy, the sender shouldn't care what the receiver likes his outputs to be. Note that _spending_ such outputs (including P2SH-embedded ones) remains nonstandard, as that is actually required for softfork safety. ACKs for commit c634b1: MarcoFalke: utACK c634b1e harding: Tested ACK c634b1e meshcollider: utACK bitcoin@c634b1e Tree-SHA512: e37168a1be9f445a04d4280593f0a92bdae33eee00ecd803d5eb16acb5c9cfc0f1f0a1dfbd5a0cc73da2c9928ec11cbdac7911513a78f85b789ae0d00e1b5962

BIP173 permits witness versions up to 16. Sending to future native witness outputs will be standard in the 0.19 Bitcoin Core release: bitcoin/bitcoin#15846

fa89bad test: Require standard txs in regtest (MarcoFalke) fa9b419 test: Add test that mainnet requires standard txs (MarcoFalke) fa613ca chainparams: Remove unused fMineBlocksOnDemand (MarcoFalke) Pull request description: I don't see a reason why regtest should allow non-standard txs, as it makes testing mainnet behaviour such as bitcoin#15846 unnecessarily hard and unintuitive. Of course, testnet policy remains unchanged to allow propagation of non-standard txs. ACKs for top commit: ajtowns: ACK fa89bad Tree-SHA512: c4c675affb054868850bd2683aa07f4c741a448cbacb2ea8334191e105f426b0790fe6a468be61e9c5880d24154f7bf1c7075051697172dce92180c1bc3a1c90

In bitcoin 0.19.0, standardness rules are going to be relaxed to allow future witness versions. Once this is widely deployed, it will be safe to accept them, smoothing use of future segwit versions. See: bitcoin/bitcoin#15846 Signed-off-by: Rusty Russell <[email protected]>

fa89bad test: Require standard txs in regtest (MarcoFalke) fa9b419 test: Add test that mainnet requires standard txs (MarcoFalke) fa613ca chainparams: Remove unused fMineBlocksOnDemand (MarcoFalke) Pull request description: I don't see a reason why regtest should allow non-standard txs, as it makes testing mainnet behaviour such as bitcoin#15846 unnecessarily hard and unintuitive. Of course, testnet policy remains unchanged to allow propagation of non-standard txs. ACKs for top commit: ajtowns: ACK fa89bad Tree-SHA512: c4c675affb054868850bd2683aa07f4c741a448cbacb2ea8334191e105f426b0790fe6a468be61e9c5880d24154f7bf1c7075051697172dce92180c1bc3a1c90

[POLICY] Make sending to future native witness outputs standard

c634b1e

maflcko added this to the 0.19.0 milestone Apr 18, 2019

DrahtBot added Tests TX fees and policy labels Apr 18, 2019

maflcko mentioned this pull request Apr 25, 2019

test: Require standard txs in regtest by default #15891

Merged

meshcollider merged commit c634b1e into bitcoin:master Apr 27, 2019

jnewbery mentioned this pull request May 9, 2019

Allow all mempool txs to be replaced after a configurable timeout (default 6h) #10823

Closed

Sjors added a commit to Sjors/libwally-core that referenced this pull request Jun 11, 2019

Allow SegWit versions 0-16

6969493

BIP173 permits witness versions up to 16. Sending to future native witness outputs will be standard in the 0.19 Bitcoin Core release: bitcoin/bitcoin#15846

Sjors mentioned this pull request Jun 11, 2019

Allow SegWit versions 0-16 ElementsProject/libwally-core#114

Closed

rustyrussell mentioned this pull request Sep 18, 2019

BOLT 2: option_shutdown_anysegwit (Feature 26/27) lightning/bolts#672

Merged

jnewbery mentioned this pull request Jul 27, 2020

Newsletters: add 108 (2020-07-29) bitcoinops/bitcoinops.github.io#442

Merged

5 tasks

This was referenced Aug 13, 2020

Address getHash requires data length of 20 or 32 only bcoin-org/bcoin#989

Closed

[POLICY] address: fix isUnknown() to allow nullData and version 0 addresses handshake-org/hsd#498

Merged

jonatack mentioned this pull request Mar 2, 2021

Newsletters: add 138 (2021-03-03) bitcoinops/bitcoinops.github.io#523

Merged

6 tasks

ariard mentioned this pull request Aug 24, 2021

policy: correct (lower) the dust threshold for Taproot outputs #22779

Closed

maflcko mentioned this pull request Jan 20, 2022

policy: treat P2TR outputs with invalid x-only pubkey as non-standard #24106

Closed

maflcko removed the Tests label Jan 20, 2022

bitcoin locked and limited conversation to collaborators Jan 20, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[POLICY] Make sending to future native witness outputs standard #15846

[POLICY] Make sending to future native witness outputs standard #15846

Uh oh!

sipa commented Apr 18, 2019

Uh oh!

maflcko commented Apr 18, 2019

Uh oh!

jnewbery commented Apr 18, 2019

Uh oh!

luke-jr commented Apr 18, 2019

Uh oh!

gmaxwell commented Apr 18, 2019

Uh oh!

harding commented Apr 24, 2019

Uh oh!

gmaxwell commented Apr 24, 2019

Uh oh!

sdaftuar commented Apr 25, 2019

Uh oh!

instagibbs commented Apr 25, 2019

Uh oh!

meshcollider commented Apr 27, 2019

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

10 participants

[POLICY] Make sending to future native witness outputs standard #15846

[POLICY] Make sending to future native witness outputs standard #15846

Uh oh!

Conversation

sipa commented Apr 18, 2019

Uh oh!

maflcko commented Apr 18, 2019

Uh oh!

jnewbery commented Apr 18, 2019

Uh oh!

luke-jr commented Apr 18, 2019

Uh oh!

gmaxwell commented Apr 18, 2019

Uh oh!

harding commented Apr 24, 2019

Uh oh!

gmaxwell commented Apr 24, 2019

Uh oh!

sdaftuar commented Apr 25, 2019

Uh oh!

instagibbs commented Apr 25, 2019

Uh oh!

meshcollider commented Apr 27, 2019

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

10 participants