Delivery-date: Tue, 18 Feb 2025 19:42:38 -0800
Sender: bitcoindev@googlegroups.com
Received-SPF: pass (google.com: domain of dave@dtrt.org designates 2607:fe70:0:3::d as permitted sender) client-ip=2607:fe70:0:3::d;
MIME-Version: 1.0
Date: Tue, 18 Feb 2025 16:29:22 -1000
From: "David A. Harding" <dave@dtrt.org>
To: Jose Storopoli <jose@storopoli.io>
Cc: Bitcoin Development Mailing List <bitcoindev@googlegroups.com>
Subject: Re: [bitcoindev] Announcing Bitcoin BOSD: Standardized Withdrawal
 Output Specification for L2s
In-Reply-To: <94ac8f66-1627-4526-bb34-095361a30fb4n@googlegroups.com>
References: <94ac8f66-1627-4526-bb34-095361a30fb4n@googlegroups.com>
Message-ID: <3d4f98e0886f03337b44f0e4152d1f33@dtrt.org>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Precedence: list
Mailing-list: list bitcoindev@googlegroups.com; contact bitcoindev+owners@googlegroups.com

On 2025-02-15 01:37, Jose Storopoli wrote:
> Dear Bitcoin Dev Community,
> 
> I am excited to introduce Bitcoin BOSD (Bitcoin Output Script
> Descriptor), a new specification and open source Rust implementation
> to simplify relay-safe (standardness-guarantee) on-chain withdrawals
> for Bitcoin Layer 2 solutions.
> 
> Key Features:
> - Standardness-by-construction for withdrawal outputs
> - Eliminates ad-hoc standardness rule implementations in L2s
> - Compact representation

It's unclear to me why you don't simply implement your selected 
standardness rules on top of a BIP385 raw() output script descriptor 
(with the normal BIP380 checksum)?  That gives you an efficient 
representation, compatibility with other software, a compact checksum, 
and local policy.

For the later point, policy being local, there exists software[1] that I 
believe has a moderate userbase (including possibly some miner support) 
that implements less restrictive standard transaction rules than Bitcoin 
Core, specifically allowing OP_RETURN outputs of more than 83 bytes 
cumulative.[2]  If a user of your software runs that node and believes 
his transactions will relay to a compatible miner, what options does he 
have of circumventing BOSD type 0's limit of 80 push bytes if one of his 
users wants to send him such a data carrier output?  He can propose a 
type 5, but you'll probably reject that because >80 bytes will be 
non-standard for other users.  You can create an extension numeric 
range, but then you'll have to centrally coordinate number assignment 
for every random proposal.  Instead, I think it makes more sense to not 
use versioning at all and just perform local verification on raw() 
descriptors; that way users of your software can modify one line of code 
on their side to accept >80 bytes and everyone else can use the default 
tests.

In addition to the above technical criticism, I find the BOSD name 
extremely conflationary with the longstanding use of "output script 
descriptors", or "descriptors" for short, to refer to the language that 
originated in 2018 from Bitcoin Core developers for describing output 
scripts.[3]  I think it would be helpful to everyone for your proposal 
to use a more distinct name.

Thanks,

-Dave

[1] https://github.com/petertodd/bitcoin/tree/libre-relay-v28.1
[2] 
https://github.com/bitcoin/bitcoin/commit/d67f4634e5395fbdf4383d7adcdfc92c0cca7fc9
[3] https://bitcoinops.org/en/topics/output-script-descriptors/

-- 
You received this message because you are subscribed to the Google Groups "Bitcoin Development Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bitcoindev+unsubscribe@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/bitcoindev/3d4f98e0886f03337b44f0e4152d1f33%40dtrt.org.