ChainSafe
Filecoin.ChainGetNode
The method is not trivial to implement given lack of IPLD tooling at hand, and it is also have no clear use case according to https://docs.google.com/spreadsheets/d/1fFkQuEjvFAd2s1dGX5zGmhxsEMLMUZ4uQFnIXgSZA5w/. Let's do it once it's prioritized higher.
@elmattic Thanks.
Also can you give some pointers about this comment made by @LesnyRumcajs: https://github.com/ChainSafe/forest/issues/4711#issuecomment-2326601066 (about lack of IPLD tooling)
@akaladarshi I don't remember precisely what it was, but looking at Lotus implementation, we'd have to implement some IPLD helpers (IPFS path resolving?) that are not available in the IPLD crate we are using at the moment (but might be available elsewhere). That said, I might have missed something, so please double-check. The relevant Lotus code is here.
@LesnyRumcajs @elmattic I wanted to confirm one small thing, in forest:
- we don't require
DAGandBlockServicesservices like in lotus?. - can blockStore directly fetch the
block or nodeif we have the CID?.
we don't require DAG and BlockServices services like in lotus?.
I'm not sure. This needs more investigation.
can blockStore directly fetch the block or node if we have the CID?
It's a KV store, so whatever is put there with a CID can be later retrieved with it.
we don't require DAG and BlockServices services like in lotus?.
I'm not sure. This needs more investigation.
can blockStore directly fetch the block or node if we have the CID?
It's a KV store, so whatever is put there with a CID can be later retrieved with it.
The Lotus uses DAG and BlockService services both are from the ipfs/boxo and it is a wrapper over the blockstore.
I don't think we have a Rust crate for it, so we are directly using the DB in forest (at least for now).
So we might not require a path to retrieve the node since there is no DAG, but I will try to investigate what is happening and report it.
@LesnyRumcajs I haven't been able to comeback to the investigation after my last comment, It's in my backlog will pick it up again soon.
@akaladarshi I think there's no need to keep it in the backlog. I don't see any demand for it; if someone requests it, we'll revisit the subject.