Does BIP 110 cause problems for existing miniscript-based wallets?Does BIP 110 cause problems for existing miniscript-based wallets?

Kevin Loaec, founder of Liana Wallet, recently posted this:

For context, for the 110 fork people:
Miniscript uses things called fragments. Some of these fragments use op_if, that your fork disables for some reason.

A user defines what their wallet should be, for example "a 2-of-5 multisig". Miniscript will output a Bitcoin Script (a wallet basically) that is the best and safest way currently known to do this policy in Bitcoin.

Disabling an opcode that is very much a building brick for safe Bitcoin wallets is dumb, but also has a big engineering cost on:

• the engineers who build the Miniscript engine
• the devs that implemented it, in Core and in multiple libraries like Rust-Miniscript
• wallet devs
• wallet users who may unknowingly have such a policy.

This cost cannot just be denied or ignored by "there is a 2 weeks window" it'll take a lot more than 2 weeks to do (took 3 years to get Miniscript anywhere), and users may be impacted without knowing so.
Bip 110 is all about virtue signaling and ignoring real costs on the ecosystem, including potential loss of coins for users who did nothing wrong, just used a good wallet software.

There has been a debate about whether BIP 110 actually interferes with the scripts produced by miniscript wallets like Liana, Nunchuk, and Bitcoin Keeper, particularly because BIP 110 proposes making OP_IF invalid in Taproot scripts. @dathon_ohm maintains that

BIP-110 does not affect UTXOs that are already confirmed before the softfork activates, so there is zero chance of Liana users being affected, even if Liana devs are irresponsible and let them create Tapleaves with OP_IF.

As well as saying this in the BIP:

Specifically regarding Liana Wallet and OP_IF, Loaec says:

I believe we don't have "if" currently in any Liana taproot scripts. We could in the future either through change of our rules (enabling more complex policy) or changes of the Miniscript version itself (outside of our control). Tagging @giacomozucco here.
AFAIK our current paths are segregated in different leaves, not using IF (in tap-miniscript).

This is about Liana, I don't know what other wallets do.

I do not support the disabling of useful opcodes though.

Nunchuk wallet weighed-in saying:

It is possible to generate OP_IF tapscripts using custom miniscripts and the standard Miniscript compiler. @mononautical went into an example a few months ago. Nunchuk supports these types of scripts.

And here is the mononautical post (part of a larger thread) that Nunchuk references:

this taproot OP_IF example is much more worrying for the RDTS proposal than I first realized.

it's actually a miniscript wallet policy:
thresh(3,pk(A),pk(B),pk(C),pk(D),older(T1),older(T2))

meaning "a 3-of-4 multisig, decaying to 2-of-4 after time T1, and 1-of-4 after time T2"

Users of miniscript wallets are not de facto irresponsibleUsers of miniscript wallets are not de facto irresponsible

BIP 110 says this about breaking user space:

Yes, this proposal intentionally breaks user space, specifically the data storage user space. This is necessary in order to communicate that data storage is not supported. Pains have been taken to avoid breaking monetary use cases, and it is unlikely that any such use cases have been affected, but in theory they might be. See the Tradeoffs section for more details.

One of the things that has been most worrying to me about BIP 110 is how its supporters have continually downplayed the disruption caused by BIP 110 changing what Bitcoin users were previously able to do, even as I quoted above, calling devs irresponsible if they don't rewrite miniscript and all its various libraries.

@giacomozucco expressed this well in a response to dathon's tweet:

Using miniscript would not have been "irresponsible", and some user may have wanted to use it after next September. A Bitcoin that allows that is objectively better money. Not the case for Liana through, as clarified.

Loaec also did a nice job explaining how breaking user space in this manner is highly disruptive:

People use whatever they use without knowing what script is actually in place.
For example, do you know what exact Bitcoin script your wallet uses? You probably don't, and shouldn't have to anyways.
You wouldn't know if you are using op_if unless you check on chain.

Bitcoin users should be able to have confidence that a utxo they create today will be spendable in 10 years. If a BIP risks this property, it needs to be for a very clear and near-universally supported reason.Bitcoin users should be able to have confidence that a utxo they create today will be spendable in 10 years. If a BIP risks this property, it needs to be for a very clear and near-universally supported reason.

Here is dathon's response to Loaec's post from above:

And here is Luke Dashjr's response: