Rune balances for spent outputs

[koirikivi]

I noticed that /output/<txid>:<vout> doesn't show the Rune balances for spent outputs. It also shows those outputs as indexed: false.

I'm working on a service that accepts Rune deposits from users (and will eventually spend the outputs associated with the users' deposits). The way I'm have currently implemented this is using the listsinceblock Bitcoin RPC api to list the all received transactions, and then for each txid/vout pair I query /output/<txid>:<vout> from the ord api to see the Rune amounts transferred as a part of that transaction. However, this doesn't work for outputs that have been spent before scanning the received transactions.

It would be very helpful to have access to the historical rune amounts transferred as a part of some output.
Maybe we could have a flag like --index-spent-runes, in congruence with --index-spent-sats?

[gmart7t2]

Spending the outputs before knowing their runic contents will spend their runic contents willy nilly. Wouldn't you want to know their contents before spending them?

[koirikivi]

Spending the outputs before knowing their runic contents will spend their runic contents willy nilly. Wouldn't you want to know their contents before spending them?

In my use case, the UTXOs are held in an n-of-m P2WSH wallet that's being scanned by multiple backend instances. It's possible that one instance is down and the other instances agree to spend the outputs. After that, it's not possible for the other instance to scan them properly.

Even if we ignore my use case, it would be a useful feature to be able to construct the Rune transaction history of a wallet. I don't think it's currently possible, at least easily.

[casey]

I think this is a pretty common thing to need. Just saw this tweet today which mentions the same thing.

[raphjaph]

We could revive our transaction index and adapt it to also store runic information. It's more overhead to index and store but exchanges and @koirikivi probably wouldn't mind the extra overhead.

[gmart7t2]

It probably makes sense to have an option to index spent runic outputs without that also indexing every spent output. Runic outputs are presumably still only a small percentage of all the outputs that have ever existed