From: Luke Dashjr <luke@dashjr.org>
To: bitcoin-dev@lists.linuxfoundation.org
Date: Tue, 18 Jan 2022 21:19:02 +0000
User-Agent: KMail/1.9.10
MIME-Version: 1.0
Content-Type: Text/Plain;
  charset="us-ascii"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <202201182119.02687.luke@dashjr.org>
Subject: [bitcoin-dev] CTV BIP review
Precedence: list

tl;dr: I don't think CTV is ready yet (but probably close), and in any case 
definitely not worth reviving BIP 9 with its known flaws and vulnerability.

My review here is based solely on the BIP, with no outside context (aside from 
current consensus rules, of course). In particular, I have _not_ looked at 
the CTV code proposed for Bitcoin Core yet.

>Covenants are restrictions on how a coin may be spent beyond key ownership. 

nit: Poorly phrased. Even simple scripts can do that already.

>A few examples are described below, which should be the subject of future 
non-consensus standardization efforts.

I would ideally like to see fully implemented BIPs for at least one of these 
(preferably the claimed CoinJoin improvements) before we move toward 
activation.

>Congestion Controlled Transactions

I think this use case hasn't been fully thought through yet. It seems like it 
would be desirable for this purpose, to allow any of the recipients to claim 
their portion of the payment without footing the fee for every other payment 
included in the batch. This is still a covenant-type solution, but one that 
BIP 119 cannot support as-is.

(I realise this may be a known and accepted limitation, but I think it should 
be addressed in the BIP)

>Payment Channels

Why batch mere channel creation? Seems like the spending transaction should 
really be the channel closing.

>CHECKTEMPLATEVERIFY makes it much easier to set up trustless CoinJoins than 
previously because participants agree on a single output which pays all 
participants, which will be lower fee than before.

I don't see how. They still have to agree in advance on the outputs, and the 
total fees will logically be higher than not using CTV...?

>Further Each participant doesn't need to know the totality of the outputs 
committed to by that output, they only have to verify their own sub-tree will 
pay them.

I don't see any way to do this with the provided implementation.

>Deployment could be done via BIP 9 VersionBits deployed through Speedy Trial.

Hard NACK on this. BIP 9 at this point represents developers attempting to 
disregard and impose their will over community consensus, as well as an 
attempt to force a miner veto backdoor/vulnerability on deployment. It should 
never be used again.

Speedy Trial implemented with BIP 8 made sense* as a possible neutral 
compromise between LOT=True and LOT=False (which could be deployed prior to 
or in parallel), but using BIP 9 would destroy this.

As with Taproot, any future deployments should use BIP 8 again, until a better 
solution is developed. Reverting back to a known flawed and vulnerable 
activation method should not be done, and it would be better not to deploy 
CTV at all at such an expense.

The fact that certain developers attempted to deploy a BIP 9 alternative 
activation for Taproot against community consensus, and that even managed to 
get released as "Bitcoin Core", makes it all the more important that the 
community firmly rejects any further action to force this regression.

* it is my opinion a BIP 8 ST would be an okay compromise under those 
circumstances; others do disagree that ST is acceptable at all

> This ensures that for a given known input, the TXIDs can also be known ahead 
of time. Otherwise, CHECKTEMPLATEVERIFY would not be usable for Batched 
Channel Creation constructions as the redemption TXID could be malleated and 
pre-signed transactions invalidated, unless the channels are built using an 
Eltoo-like protocol.

Why is it a problem for them to use an Eltoo-like protocol?

Why not just commit to the txid itself if that's the goal?

>P2SH is incompatible with CHECKTEMPLATEVERIFY 

Maybe the CTV opcode should only be defined/enforced within witness scripts?

>nLockTime should generally be fixed to 0 (in the case of a payment tree, only 
the *first* lock time is needed to prevent fee-sniping the root)

Your "Congestion Controlled Transactions" example would only make sense with 
the spending transaction much later than the "root", and so could benefit 
from fee sniping malleability. (In fact, in that example, it would be better 
not to commit to locktime at all.)

>In the CHECKTEMPLATEVERIFY approach, the covenants are severely restricted to 
simple templates. The structure of CHECKTEMPLATEVERIFY template is such that 
the outputs must be known exactly at the time of construction. Based on a 
destructuring argument, it is only possible to create templates which expand 
in a finite number of steps.

It's not clear to me that this holds if OP_CAT or OP_SHA256STREAM get added.

>For example, a exchange's hot wallet might use an address which can 
automatically be moved to a cold storage address after a relative timeout.

Wouldn't it make more sense to just have a UTXO both cold+hot can spend, then 
throw away the hot key?

>In contrast to previous forks, OP_CHECKTEMPLATEVERIFY will not make scripts 
valid for policy until the new rule is active.

Policy isn't validity, and cannot be dictated by BIPs (or anyone/anything, for 
that matter).

Luke