Feature Request: elements/liquid chains

[tiero]

It looks like it's based on electrs, how much of an effort would be make it work for elements chains?

Not that confident yet with Rust, but I could try to make contribution if you think is feasible to use the same codebase. If not, would be better to do it as a fork?

[shesek]

It's actually just the electrum server implementation (src/electrum/server.rs) that's based on electrs, the index and query engines are separate, so it can't benefit much from bs/electrs's support for elements.

It should however be feasible do to this in a manner similar to how bs/electrs does it, with a compile time flag that switches the data types to elements's and adds some special-case behaviors for CT, issued assets and peg ins/outs.

Can you tell me a little more about the use-case? What are you looking to do with this?

[tiero]

For the time being Blockstream's fork of electrs (namely Esplora?) is the only explorer out there for elements. It's optimized for "enterprise" workflows, but as a "personal" indexer use cases a bit less (ie. too much resource hungry).

When I say personal, I mean "hey this is my xpub, please notify myself when new transactions are coming in related to this wallet, don't care about indexing others"

Therefore HTTP2 server-side sent push (or maybe gRPC streaming even more) would be very useful for couple of micro-services I run.

[tiero]

I think then at this point I would go with doing then a PoC that uses zmq notifications of an elements node to get blocks, and a custom block/tx parser that filters per script of a given xpub.

The idea is to keep the same exact RES API as Esplora plus server-side push/streaming, to make sure clients are cross-compatible

[shesek]

BWT provides a real-time streaming API for wallet events (also available via WebHooks) and has an HTTP API that's somewhat similar to Esplora, but also different in some ways.

I can definitely see how it could be useful to make this available for elements. I will look into this, but it probably won't happen soon. Will gladly accept a PR if it doesn't overly complicate the codebase.