BSV’s solution to DSV is part of the OP_PUSH_TX technique family which is limited in a few “minor” ways and one “major” way, particularly related to unique identifiers (in L1). Take a look at the scrypt guy’s “almost-no-backtracking” L1 tokens (cont)
Post by twetch#2782
(cont) https://powping.com/posts/a8323f5a627b1e26ba6d5493b6c61fdb238c0c439017ee265e4fc307d021cfb6
Then look at user @joe’s response in that thread.
What the chain says
- Block
- 657 760
- Time
- 2020-10-20T23:36:03Z
- Signer
- 1MV8wZFSxwqyadJU6KgMsws8NgHibhqddR
- App
- twetch
- Type
- post
- Content type
- text/plain
- Name in tx
- twetch#2782
Fields the transaction did not carry are omitted. Open the payload to see the bytes as stored.
Signed by
1MV8wZFSxwqyadJU6KgMsws8NgHibhqddR VerifiedReplies (9)
All of nchain’s techniques for globally stateful L1 apps make use of compute oracles or as I call them “CSW oracles”, have a look at Run or SuperAsset for case studies.
Btw I didn’t say I thought it will fail, I said it doesn’t create an L1 network effect
As a last note I am open to being proven wrong, in particular I think there could be some way to use a tree structure to encode “backtrack” transactions into the solution provided by xhliu, making it practical at scale even if theoretically constrained.
But to be clear, this is layers of hacks on hacks on hacks, it is absolutely using tricks to work around limitations which could be easily addressed with a very minor architectural improvement with zero drawbacks.
One more side note, if you’re looking at BCH ops then look at GROUP, not DSV. Despite what coingeek has published, OP_PUSH_TX does not enable the same generality, for the reason the commenter @joe responded in that thread by scrypt guy