[bitcoin-dev] Examining ScriptPubkeys in Bitcoin Script

[bitcoin-dev] Examining ScriptPubkeys in Bitcoin Script

[Rusty Russell]

Hi all,

        I've done an exploration of what would be required (given
OP_TX/OP_TXHASH or equivalent way of pushing a scriptPubkey on the
stack) to usefully validate Taproot outputs in Bitcoin Script.  Such
functionality is required for usable vaults, at least.

        https://rusty.ozlabs.org/2023/10/20/examining-scriptpubkey-in-script.html

(If anyone wants to collaborate to produce a prototype, and debug my
surely-wrong script examples, please ping me!)

TL;DR: if we have OP_TXHASH/OP_TX, and add OP_MULTISHA256 (or OP_CAT),
OP_KEYADDTWEAK and OP_LESS (or OP_CONDSWAP), and soft-fork weaken the
OP_SUCCESSx rule (or pop-script-from-stack), we can prove a two-leaf
tapscript tree in about 110 bytes of Script.  This allows useful
spending constraints based on a template approach.

Thanks!
Rusty.

Re: [bitcoin-dev] Examining ScriptPubkeys in Bitcoin Script

[Brandon Black]

On 2023-10-20 (Fri) at 14:10:37 +1030, Rusty Russell via bitcoin-dev wrote:
>         I've done an exploration of what would be required (given
> OP_TX/OP_TXHASH or equivalent way of pushing a scriptPubkey on the
> stack) to usefully validate Taproot outputs in Bitcoin Script.  Such
> functionality is required for usable vaults, at least.

So you're proposing this direction as an alternative to the more
constrained OP_UNVAULT that replaces a specific leaf in the taptree in a
specific way? I think the benefits of OP_UNVAULT are pretty big vs. a
generic construction (e.g. ability to unvault multiple inputs sharing
the same scriptPubkey to the same output).

> TL;DR: if we have OP_TXHASH/OP_TX, and add OP_MULTISHA256 (or OP_CAT),
> OP_KEYADDTWEAK and OP_LESS (or OP_CONDSWAP), and soft-fork weaken the
> OP_SUCCESSx rule (or pop-script-from-stack), we can prove a two-leaf
> tapscript tree in about 110 bytes of Script.  This allows useful
> spending constraints based on a template approach.

I agree that this is what is needed. I started pondering this in
response to some discussion about the benefits of OP_CAT or OP_2SHA256
for BitVM.

Personally I'd use OP_TAGGEDCATHASH that pops a tag (empty tag can be
special cased to plain sha256) and a number (n) of elements to hash,
then tagged-hashes the following 'n' elements from the stack.

Best,

--Brandon

Re: [bitcoin-dev] Examining ScriptPubkeys in Bitcoin Script

[Rusty Russell]

Brandon Black <freedom@reardencode.com> writes:
> On 2023-10-20 (Fri) at 14:10:37 +1030, Rusty Russell via bitcoin-dev wrote:
>>         I've done an exploration of what would be required (given
>> OP_TX/OP_TXHASH or equivalent way of pushing a scriptPubkey on the
>> stack) to usefully validate Taproot outputs in Bitcoin Script.  Such
>> functionality is required for usable vaults, at least.
>
> So you're proposing this direction as an alternative to the more
> constrained OP_UNVAULT that replaces a specific leaf in the taptree in a
> specific way? I think the benefits of OP_UNVAULT are pretty big vs. a
> generic construction (e.g. ability to unvault multiple inputs sharing
> the same scriptPubkey to the same output).

I would have to write the scripts exactly (and I'm already uncomfortable
that I haven't actually tested the ones I wrote so far!) to properly
evaluate.

In general, script makes it hard to do N-input evaluation (having no
iteration).  It would be useful to attempt this though, as it might
enlighted us as to OP_TXHASH input selection: for example, we might want
to have an "all *but* one input" mode for this kind of usage.

Dealing with satsoshi amounts is possible, but really messy (that's my next
post).

>> TL;DR: if we have OP_TXHASH/OP_TX, and add OP_MULTISHA256 (or OP_CAT),
>> OP_KEYADDTWEAK and OP_LESS (or OP_CONDSWAP), and soft-fork weaken the
>> OP_SUCCESSx rule (or pop-script-from-stack), we can prove a two-leaf
>> tapscript tree in about 110 bytes of Script.  This allows useful
>> spending constraints based on a template approach.
>
> I agree that this is what is needed. I started pondering this in
> response to some discussion about the benefits of OP_CAT or OP_2SHA256
> for BitVM.

Given these examples, I think it's clear that OP_MULTISHA256 is almost
as powerful as OP_CAT, without the stack limit problems.  And OP_2SHA256
is not sufficient in general for CScriptNum generation, for example.

> Personally I'd use OP_TAGGEDCATHASH that pops a tag (empty tag can be
> special cased to plain sha256) and a number (n) of elements to hash,
> then tagged-hashes the following 'n' elements from the stack.

That's definitely a premature optimization to save two opcodes.

Cheers,
Rusty.

Re: [bitcoin-dev] Examining ScriptPubkeys in Bitcoin Script

[Anthony Towns]

On Fri, Oct 20, 2023 at 02:10:37PM +1030, Rusty Russell via bitcoin-dev wrote:
>         I've done an exploration of what would be required (given
> OP_TX/OP_TXHASH or equivalent way of pushing a scriptPubkey on the
> stack) to usefully validate Taproot outputs in Bitcoin Script.  Such
> functionality is required for usable vaults, at least.
> 
>         https://rusty.ozlabs.org/2023/10/20/examining-scriptpubkey-in-script.html
> 
> (If anyone wants to collaborate to produce a prototype, and debug my
> surely-wrong script examples, please ping me!)
> 
> TL;DR: if we have OP_TXHASH/OP_TX, and add OP_MULTISHA256 (or OP_CAT),
> OP_KEYADDTWEAK and OP_LESS (or OP_CONDSWAP), and soft-fork weaken the
> OP_SUCCESSx rule (or pop-script-from-stack), we can prove a two-leaf
> tapscript tree in about 110 bytes of Script.  This allows useful
> spending constraints based on a template approach.

I think there's two reasons to think about this approach:

 (a) we want to do vault operations specifically, and this approach is
     a good balance between being:
       - easy to specify and implement correctly, and
       - easy to use correctly.

 (b) we want to make bitcoin more programmable, so that we can do
     contracting experiments directly in wallet software, without needing
     to justify new soft forks for each experiment, and this approach
     provides a good balance amongst:
       - opening up a wide range of interesting experiments,
       - making it easy to understand the scope/consequences of opening up
         those experiments,
       - being easy to specify and implement correctly, and
       - being easy to use correctly.

Hopefully that's a fair summary? Obviously what balance is "good"
is always a matter of opinion -- if you consider it hard to do soft
forks, then it's perhaps better to err heavily towards being easy to
specify/implement, rather than easy to use, for example.

For (a) I'm pretty skeptical about this approach for vault operations
-- it's not terribly easy to specify/implement (needing 5 opcodes, one
of which has a dozen or so flags controlling how it behaves, then also
needs to change the way OP_SUCCESS works), and it seems super complicated
to use.

By comparison, while the bip 345 OP_VAULT proposal also proposes 3 new
opcodes (OP_CTV, OP_VAULT, OP_VAULT_RECOVER) [0], those opcodes can be
implemented fairly directly (without requiring different semantics for
OP_SUCCESS, eg) and can be used much more easily [1].

[0] Or perhaps 4, if OP_REVAULT were to be separated out from OP_VAULT, cf
    https://github.com/bitcoin/bips/pull/1421#discussion_r1357788739

[1] https://github.com/jamesob/opvault-demo/blob/57f3bb6b8717acc7ce1eae9d9d8a2661f6fa54e5/main.py#L125-L133

I'm not sure, but I think the "deferred check" setup might also
provide additional functionality beyond what you get from cross-input
introspection; that is, with it, you can allow multiple inputs to safely
contribute funds to common outputs, without someone being able to combine
multiple inputs into a tx where the output amount is less than the sum
of all the contributions. Without that feature, you can mimic it, but
only so long as all the input scripts follow known templates that you
can exactly match.

So to me, for the vault use case, the
TXHASH/MULTISHA256/KEYADDTWEAK/LESS/CAT/OP_SUCCESS approach just doesn't
really seem very appealing at all in practical terms: lots of complexity,
hard to use, and doesn't really seem like it works very well even after
you put in tonnes of effort to get it to work at all?



I think in the context of (b), ie enabling experimentation more generally,
it's much more interesting. eg, CAT alone would allow for various
interesting constraints on signatures ("you must sign this tx with the
given R value -- so attempting to double spend, eg via a feebump, will
reveal the corresponding private key"), and adding CSFS would allow you
to include authenticated data in a script, eg market data sourced from
a trusted oracle.

But even then, it still seems fairly crippled -- script is a very
limited programming language, and it just isn't really very helpful
if you want to do things that are novel. It doesn't allow you to (eg)
loop over the inputs and select just the ones you're interested in, you
need the opcode to do the looping for you, and that has to be hardcoded
as a matter of consensus (eg, Steven Roose's TXHASH [2] proposal allows
you to select the first-n inputs/outputs, but not the last-n).

[2] https://github.com/bitcoin/bips/pull/1500

I've said previously [3] that I think using a lisp variant would
be a promising solution here: you can replace script's "two stacks
of byte-strings" with "(recursive) lists of byte-strings", and go
from a fairly limited language, to a fairly complete one. I've been
experimenting with this on and off since then [4], and so far I haven't
seen anything much to dissuade me from that view. I think you can get
a pretty effective language with perhaps 43 opcodes [5] (compared to
script's ~60 active opcodes), and I don't think you need to do anything
too fancy to implement it.

[3] https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2022-March/020036.html
[4] https://github.com/ajtowns/lisp-play/
[5] https://github.com/ajtowns/lisp-play/blob/5975870423f9dace902ef42208b965f9d8a0f005/btclisp.py#L738

Here's an example. I've included a "CSFS" equivalent opcode, namely
"(bip340_verify pk msg sig)" that validates a signature per BIP340,
and also a "(bip342_txmsg)" opcode that generates a "msg" corresponding
to the BIP 342 "Signature Validation" spec (just calling the bitcoind
core test framework code), which then allows me to verify existing
signatures on existing transactions via lisp code, rather than executing
the actual script.

But what if we wanted to experiment with a new SIGHASH mode? For that,
I've added an OP_TX like opcode, '(tx N)' that allows you to select
various information about the tx by choosing N -- '(tx 1)' gives you the
locktime, '(tx 10)' gives you your input's nSequence, '(tx (10 . 3))'
gives you the nSequence of the 4th input, eg. With that, it's possible
to select whichever bits of the transaction you like, in whatever order
you like, and pass the results through the '(sha256)' opcode, then pass
that into the signature check.

Unlike the OP_TXHASH proposals and the like, it's possible (though perhaps
not *easy*) to exactly mimic existing hash constructs, eg "(bip342_txmsg)"
(for SIGHASH_ALL) can be constructed manually via:

ENV=(a (i 14 '(a 8 8 12 (+ 10 '1) (- 14 '1) (cat 3 (a 12 10))) '3))

  ^-- (basically a for loop, so that "(a 1 1 'X '0 K)" will
       invoke "X" with values [0, K), and cat the results
       together; used with K=(tx '2) to do inputs, and (tx '3) to
       dou outputs)

PROGRAM=(a '(sha256 4 4 '0x00 6 3) (sha256 '\"TapSighash\") (cat '0x00 (tx '0) (tx '1) (sha256 (a 1 1 '(cat (tx (c '11 1)) (tx (c '12 1))) '0 (tx '2) 'nil)) (sha256 (a 1 1 '(tx (c '15 1)) '0 (tx '2) 'nil)) (sha256 (a 1 1 '(a '(cat (strlen 1) 1) (tx (c '16 '0))) '0 (tx '2) 'nil)) (sha256 (a 1 1 '(tx (c '10 1)) '0 (tx '2) 'nil)) (sha256 (a 1 1 '(cat (tx (c '20 1)) (a '(cat (strlen 1) 1) (tx (c '21 1)))) '0 (tx '3) 'nil)) (i (tx '7) '0x03 '0x01) (substr (cat (tx '4) '0x00000000) 'nil '4) (i (tx '7) (sha256 (a '(cat (strlen 1) 1) (tx '7))) 'nil)) (cat (tx '6) '0x00 '0xffffffff))

  ^-- (sha256's the sha256 of TapSighash twice, then the epoch, then
       the sigmsg, then the extension; with the SIGHASH_ALL logic
       being hardcoded)

That's obviously not easy to read, but it's also essentially programming
in assembler, and would be much improved by having a higher-level
macro-enabled lisp variation that allows you to define your own
symbols/variable names, and translate that down to the raw code. (Or
even just having a parser that allows you to add comments, I guess)

What I've implemented is essentially an eager interpretor with some tail
call optimisations to allow memory to be freed up a bit earlier. I think
it would be better to do it as a properly lazy iinterpretor though --
that way you can actually have the same memory efficiency as streaming
sha256 operators provide, even with the additional flexibility provided
by iteration/recursion/function calls.

There are various other tricks that aren't done in my python testbed,
eg encoding/decoding lists as a byte stream rather than a parenthesised
string; working out whether string comparison should be normal or reversed
(so that you can comapre proof-of-work) or both, providing other crypto
ops like ecdsa, doing bignum maths rather than just uint64, keeping track
of allocations when an exception occurs, providing an easy way to tell
how much computation will be required to evaluate an input script and
inflate the tx's weight correspondingly if necessary, etc.

I've also only done fairly toy-level problems: factorial and fibonacci
calculations, reimplementing an existing sighash, etc. I think doing
TLUV or VAULT or graftroot should be feasible (at least given opcodes
to provide secp256k1 tweaks and deferred-checks), but haven't actually
done it.

Anyway, this seems to me to be a much more promising approach for
experimentation than trying to fit everything into script's square hole
[6], and perhaps also more promising than Simplicity for the reasons
discussed at the end of [3]. Once you have the nicer structure that a
lisp-like language provides, compared to script, I think OP_TX, OP_CAT,
OP_CSFS etc all end up working pretty great.

[6] https://twitter.com/TiredActor/status/1609641593836822530

Cheers,
aj

Re: [bitcoin-dev] Examining ScriptPubkeys in Bitcoin Script

[Rusty Russell]

Anthony Towns <aj@erisian.com.au> writes:
> On Fri, Oct 20, 2023 at 02:10:37PM +1030, Rusty Russell via bitcoin-dev wrote:
>>         I've done an exploration of what would be required (given
>> OP_TX/OP_TXHASH or equivalent way of pushing a scriptPubkey on the
>> stack) to usefully validate Taproot outputs in Bitcoin Script.  Such
>> functionality is required for usable vaults, at least.
>> 
>>         https://rusty.ozlabs.org/2023/10/20/examining-scriptpubkey-in-script.html
>> 
>> (If anyone wants to collaborate to produce a prototype, and debug my
>> surely-wrong script examples, please ping me!)
>> 
>> TL;DR: if we have OP_TXHASH/OP_TX, and add OP_MULTISHA256 (or OP_CAT),
>> OP_KEYADDTWEAK and OP_LESS (or OP_CONDSWAP), and soft-fork weaken the
>> OP_SUCCESSx rule (or pop-script-from-stack), we can prove a two-leaf
>> tapscript tree in about 110 bytes of Script.  This allows useful
>> spending constraints based on a template approach.
>
> I think there's two reasons to think about this approach:
>
>  (a) we want to do vault operations specifically, and this approach is
>      a good balance between being:
>        - easy to specify and implement correctly, and
>        - easy to use correctly.
>
>  (b) we want to make bitcoin more programmable, so that we can do
>      contracting experiments directly in wallet software, without needing
>      to justify new soft forks for each experiment, and this approach
>      provides a good balance amongst:
>        - opening up a wide range of interesting experiments,
>        - making it easy to understand the scope/consequences of opening up
>          those experiments,
>        - being easy to specify and implement correctly, and
>        - being easy to use correctly.
>
> Hopefully that's a fair summary? Obviously what balance is "good"
> is always a matter of opinion -- if you consider it hard to do soft
> forks, then it's perhaps better to err heavily towards being easy to
> specify/implement, rather than easy to use, for example.
>
> For (a) I'm pretty skeptical about this approach for vault operations
> -- it's not terribly easy to specify/implement (needing 5 opcodes, one
> of which has a dozen or so flags controlling how it behaves, then also
> needs to change the way OP_SUCCESS works), and it seems super complicated
> to use.

But AFAICT there are multiple perfectly reasonable variants of vaults,
too.  One would be:

1. master key can do anything
2. OR normal key can send back to vault addr without delay
3. OR normal key can do anything else after a delay.

Another would be:
1. normal key can send to P2WPKH(master)
2. OR normal key can send to P2WPKH(normal key) after a delay.

> By comparison, while the bip 345 OP_VAULT proposal also proposes 3 new
> opcodes (OP_CTV, OP_VAULT, OP_VAULT_RECOVER) [0], those opcodes can be
> implemented fairly directly (without requiring different semantics for
> OP_SUCCESS, eg) and can be used much more easily [1].

I'm interested in vaults because they're a concrete example I can get my
head around.  Not because I think they'll be widely used!  So I feel
that anyone who has the ability to protect two distinct keys, and make
two transactions per transfer is not a great candidate for optimization
or convenience.

> I'm not sure, but I think the "deferred check" setup might also
> provide additional functionality beyond what you get from cross-input
> introspection; that is, with it, you can allow multiple inputs to safely
> contribute funds to common outputs, without someone being able to combine
> multiple inputs into a tx where the output amount is less than the sum
> of all the contributions. Without that feature, you can mimic it, but
> only so long as all the input scripts follow known templates that you
> can exactly match.

Agreed, I don't think you would implement anything but 1:1 unvaulting in
bitcoin script, except as a party trick.

> So to me, for the vault use case, the
> TXHASH/MULTISHA256/KEYADDTWEAK/LESS/CAT/OP_SUCCESS approach just doesn't
> really seem very appealing at all in practical terms: lots of complexity,
> hard to use, and doesn't really seem like it works very well even after
> you put in tonnes of effort to get it to work at all?

Well, I found the vault BIP really hard to understand.  I think it wants
to be a new address format, not script opcodes.

I don't think spelling it out in script is actually that much more
complex to use, either.  "Use these templates".  And modulo
consolidation, I think it works as well.

> I think in the context of (b), ie enabling experimentation more generally,
> it's much more interesting. eg, CAT alone would allow for various
> interesting constraints on signatures ("you must sign this tx with the
> given R value -- so attempting to double spend, eg via a feebump, will
> reveal the corresponding private key"), and adding CSFS would allow you
> to include authenticated data in a script, eg market data sourced from
> a trusted oracle.

Oh, oracles like this are the first CSFS use case I've heard of that
doesn't seem like abusing signatures to do hashing; nice!

(Seems like there should be a way to do this without CSFS, but I can't
see it...)

> But even then, it still seems fairly crippled -- script is a very
> limited programming language, and it just isn't really very helpful
> if you want to do things that are novel. It doesn't allow you to (eg)
> loop over the inputs and select just the ones you're interested in, you
> need the opcode to do the looping for you, and that has to be hardcoded
> as a matter of consensus (eg, Steven Roose's TXHASH [2] proposal allows
> you to select the first-n inputs/outputs, but not the last-n).

Indeed, but I still think there's much room for improvement before a
replacement.  It's hard to compare the hobbled script we have today with
an alternative, since most interesting cases are impossible.

Cheers,
Rusty.

Re: [bitcoin-dev] Examining ScriptPubkeys in Bitcoin Script

[James O'Beirne]

[-- Attachment #1: Type: text/plain, Size: 1288 bytes --]

On Sat, Oct 28, 2023 at 12:51 AM Rusty Russell via bitcoin-dev <
bitcoin-dev@lists.linuxfoundation.org> wrote:

> But AFAICT there are multiple perfectly reasonable variants of vaults,
> too.  One would be:
>
> 1. master key can do anything
> 2. OR normal key can send back to vault addr without delay
> 3. OR normal key can do anything else after a delay.
>
> Another would be:
> 1. normal key can send to P2WPKH(master)
> 2. OR normal key can send to P2WPKH(normal key) after a delay.
>

I'm confused by what you mean here. I'm pretty sure that BIP-345 VAULT
handles the cases that you're outlining, though I don't understand your
terminology -- "master" vs. "normal", and why we are caring about P2WPKH
vs. anything else. Using the OP_VAULT* codes can be done in an arbitrary
arrangement of tapleaves, facilitating any number of vaultish spending
conditions, alongside other non-VAULT leaves.

Well, I found the vault BIP really hard to understand.  I think it wants
> to be a new address format, not script opcodes.
>

Again confused here. This is like saying "CHECKLOCKTIMEVERIFY wants to be a
new address format, not a script opcode."

That said, I'm sure some VAULT patterns could be abstracted into the
miniscript/descriptor layer to good effect.

[-- Attachment #2: Type: text/html, Size: 1834 bytes --]

Re: [bitcoin-dev] Examining ScriptPubkeys in Bitcoin Script

[Rusty Russell]

"James O'Beirne" <james.obeirne@gmail.com> writes:
> On Sat, Oct 28, 2023 at 12:51 AM Rusty Russell via bitcoin-dev <
> bitcoin-dev@lists.linuxfoundation.org> wrote:
>
>> But AFAICT there are multiple perfectly reasonable variants of vaults,
>> too.  One would be:
>>
>> 1. master key can do anything
>> 2. OR normal key can send back to vault addr without delay
>> 3. OR normal key can do anything else after a delay.
>>
>> Another would be:
>> 1. normal key can send to P2WPKH(master)
>> 2. OR normal key can send to P2WPKH(normal key) after a delay.
>>
>
> I'm confused by what you mean here. I'm pretty sure that BIP-345 VAULT
> handles the cases that you're outlining, though I don't understand your
> terminology -- "master" vs. "normal", and why we are caring about P2WPKH
> vs. anything else. Using the OP_VAULT* codes can be done in an arbitrary
> arrangement of tapleaves, facilitating any number of vaultish spending
> conditions, alongside other non-VAULT leaves.

I was thinking from a user POV: the "master" key is the one they keep
super safe in case of emergencies, the "normal" is the delayed spend
key.

OP_VAULT certainly can encapsulate this, but I have yet to do the kind
of thorough review that I'd need to evaluate the various design
decisions.

> Well, I found the vault BIP really hard to understand.  I think it wants
>> to be a new address format, not script opcodes.
>>
>
> Again confused here. This is like saying "CHECKLOCKTIMEVERIFY wants to be a
> new address format, not a script opcode."

I mean in an ideal world, Bitcoin Script would be powerful enough to
implement vaults, and once a popular use pattern emerged we'd introduce
a new address type, defined to expand to that template.  Like P2WPK or
P2PKH.

Sadly, we're not in that world!  BIP 345 introduces a number of separate
mechanisms, such as limited script delegation, iteration and amount
arithmetic which are not expressible in Script (ok, amount arithmetic
kind of is, but ick!).

To form a real opinion, I need to consider all these elements, and
whether they should exist inside OP_VAULT, or as separate things.
That's a slow process, sorry :(

> That said, I'm sure some VAULT patterns could be abstracted into the
> miniscript/descriptor layer to good effect.

That would be very interesting, but hard.  Volunteers? :)

Cheers,
Rusty.

Re: [bitcoin-dev] Examining ScriptPubkeys in Bitcoin Script

[Anthony Towns]

On Sat, Oct 28, 2023 at 03:19:30PM +1030, Rusty Russell via bitcoin-dev wrote:

[Quoted text has been reordered]

> > I think there's two reasons to think about this approach:
> >  (a) we want to do vault operations specifically,

> I'm interested in vaults because they're a concrete example I can get my
> head around.  Not because I think they'll be widely used!

I don't think that's likely to make for a very productive discussion: we
shouldn't be changing consensus to support toy examples, after all. If
there's a *useful* thing to do that would be possible with similar
consensus changes, lets discuss that thing; if there's nothing useful
that needs these consensus changes, then lets discuss something useful
that doesn't need consensus changes instead.

To me, it seems pretty likely that if you're designing an API where
you don't expect anyone to actually use it for anything important, then
you're going to end up making pretty bad API -- after all, why put in
the effort to understand the use case and make a good API if you're sure
it will never be useful anyway?

There are some articles on API design that I quite like:

  https://ozlabs.org/~rusty/index.cgi/tech/2008-03-30.html
  https://ozlabs.org/~rusty/index.cgi/tech/2008-04-01.html

I'd rate the "lets have a mass of incomprehensible script that no one
really understands and is incredibly dangerous to modify, and just make it
a template" approach as somewhere between "3. Read the documentation and
you'll get it right" (at best) and "-5 Do it right and it will sometimes
break at runtime" (more likely).

Anyway, for the specific examples:

> But AFAICT there are multiple perfectly reasonable variants of vaults,
> too.  One would be:
>
> 1. master key can do anything
> 2. OR normal key can send back to vault addr without delay
> 3. OR normal key can do anything else after a delay.

I don't think there's any point having (2) here unless you're allowing for
consolidation transactions (two or more vault inputs spending to a single
output that's the same vault), which you've dismissed as a party trick.

> Another would be:
> 1. normal key can send to P2WPKH(master)
> 2. OR normal key can send to P2WPKH(normal key) after a delay.

I don't think there's any meaningful difference here between (2) here
and (3) above -- after the delay, you post one transaction spending
to p2wpkh(normal) signed by the normal key, then immediately post a
second transaction spending that new output, which is also signed with
the normal key, so you've just found a way of saying "normal key can do
anything else after a delay" that takes up more blockspace.

Both these approaches mirror the model that kanzure posted about in 2018
[0] (which can already be done via presigned transactions) and they share
the same fundamental flaw [1], namely that once someone compromises the
"normal key", all they have to do is wait for the legitimate owner to
trigger a spend, at which point they can steal the funds by racing the
owner, with the owner having no recourse beyond burning most of the
vault's funds to fees.

(I think the above two variants are meaningfully worse than kanzure's
in that someone with the normal key just needs to wait for the vault
utxo to age, at which point they can steal the funds immediately. In
kanzure's model, you first need to broadcast a trigger transaction,
then wait for it to age before funds can be stolen, which isn't the
greatest protection, but it at least adds some difficulty)

[0] https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2019-August/017229.html
[1] https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2019-August/017231.html
    also, https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2022-April/020284.html

If you don't mind that flaw, you can setup a vault along the lines of
kanzure's design with BIP 345 fairly simply:

  A: ipk=master,
     tapscript=<normal> SWAP OVER CHECKSIGVERIFY 1 "CHECKSIGVERIFY 144 CSV" OP_VAULT
       (witness values: <revault-amt> <revault-idx> <spend-idx> <sig>)

  B: ipk=master
     tapscript=<normal> CHECKSIGVERIFY 144 CSV

You put funds into the vault by creating a utxo with address "A", at which
point you can do anything with the funds via the master key, or you can
trigger an unvault via the normal key, moving funds to address "B", which
then has a 144 block delay before it can be spent via the normal key.

That also natively supports vault spends that only include part of the
vault value, or that combine two or more (compatible) vaults into a
single payment.

To avoid the flaw, you need to precommit to the spend that you're
authorising, while still allowing clawback/recovery by the owner. One
way to make that work with BIP 345 is using BIP 119's CTV to force a
precommitment to the spend:

  A: ipk=master,
     tapscript=<masterspkhash> OP_VAULT_RECOVER
     tapscript=<normal> CHECKSIGVERIFY 1 "CTV DROP 144 CSV" OP_VAULT
       (witness values: <revault-amt> <revault-idx> <spend-idx> <spendcommit> <sig>)

  B: ipk=master
     tapscript=<masterspkhash> OP_VAULT_RECOVER
     tapscript=<spendcommit> CTV DROP 144 CSV

Once you have funds in the vault in address A, you can spend them directly
via a key path spend with the master private key, or you can make
them only spendable via the master key via the OP_VAULT_RECOVER path,
or you can do a precommitted spend via the OP_VAULT path by including
"spendcommit", the CTV hash of where you want to send funds to. That
moves funds into address B, which again can be recovered to the master
key via the first two paths, or after a day you can use the CTV path to
complete the vault withdrawal.

Again, that natively supports vault spends that only include part of
the vault value, or that combine two or more (compatible) vaults into
a single payment.

> Oh, oracles like this are the first CSFS use case I've heard of that
> doesn't seem like abusing signatures to do hashing; nice!
>
> (Seems like there should be a way to do this without CSFS, but I can't
> see it...)

It's not really a novel observation: oracles are the third item listed
on the optech topic page for CSFS [2]...

The scriptless way of getting similar functionality is via discreet
log contracts [3], where the oracle with public key P, picks an event,
publishes a unique (hardened) public key R for that event, and when
the outcome of the event (m) is known, publishes 's' such that sG = R +
H(R,P,m)*P, so that (R,s) is a valid BIP 340 schnorr signature for
message m and pubkey P. That can then be used via an adaptor signature
eg to make on-chain contracts [4], eg.

[2] https://bitcoinops.org/en/topics/op_checksigfromstack/

[3] https://www.dlc.wiki/
    https://bitcoinops.org/en/topics/discreet-log-contracts/

[4] https://suredbits.com/category/discreet-log-contracts/
    https://atomic.finance/blog/discreet-log-contracts/

Cheers,
aj