handshakealliance / nomenclate Goto Github PK

Small feature, but when a new full instance of Nomenclate is synced, show the time it took to do that.

Right now getting an error on the address endpoint of get mempool, it's returning: Error object is not a valid address.

txids, and name hashes.

This will be a meta issue for tracking all remaining features that need to be implemented in Nomenclate. These are the remaining features that need to be implemented:

Http:

• Block -> header
• Block -> headers
• Estimate Fee
• Headers -> suscribe
• Relay Fee
• Address -> get mempool
• Address -> list unspent
• Address -> Subscribe
• Address -> Utxos
• Tx -> Broadcast
• Tx -> Get
• Tx -> Get Merkle
• Tx -> ID from Pos
• Mempool -> Changes
• Mempool -> Fee Histogram
• Server -> Add Peer
• Server -> Banner
• Server -> Donation_address
• Server -> Features
• Server -> Ping
• Server -> Version
• Peers -> Subscribe

rpc:

• Block -> header
• Block -> headers
• Estimate Fee
• Headers -> suscribe
• Relay Fee
• Address -> get mempool
• Address -> list unspent
• Address -> Subscribe
• Address -> Utxos
• Tx -> Broadcast
• Tx -> Get
• Tx -> Get Merkle
• Tx -> ID from Pos
• Mempool -> Changes
• Mempool -> Fee Histogram
• Server -> Add Peer
• Server -> Banner
• Server -> Donation_address
• Server -> Features
• Server -> Ping
• Server -> Version
• Peers -> Subscribe

GRPC:

• Block -> header
• Block -> headers
• Estimate Fee
• Headers -> suscribe
• Relay Fee
• Address -> get mempool
• Address -> list unspent
• Address -> Subscribe
• Address -> Utxos
• Tx -> Broadcast
• Tx -> Get
• Tx -> Get Merkle
• Tx -> ID from Pos
• Mempool -> Changes
• Mempool -> Fee Histogram
• Server -> Add Peer
• Server -> Banner
• Server -> Donation_address
• Server -> Features
• Server -> Ping
• Server -> Version
• Peers -> Subscribe

Would be great to replicate the functionality in https://github.com/romanz/electrs which allows indexing without the full node needing to index txs or addresses.

This will rely on a few P.Rs that need to be created and merged for HSD.

We have a statement in http options that should disable apikey requirements if we are listening on localhost.

This was failing when no-auth was not passed in through options, so let's double check that, that logic gets fixed.

From my understanding, standard Electrum servers run on port 8081 for http, and 8082 for https.

This will be a slight change due to the fact that we already run on port 8080, but I think it would be a change in the right direction if we want to make Nomenclate as compatible as Electrum.

There is also a case to be made that since we assume some people might run both instances at the same time, we should run at port 8083, and 8084. Either way, a switch from 8080 is likely required.

Currently Nomenclate only runs as a plugin to hsd. Not all use cases will want to run Nomenclate along with a full handshake node, and thus we should build in an option to run Nomenclate as a standalone server - and connect to a running node, or even connect to other Nomenclate servers.

Current I think if someone searches their address, the claims may not show very well.

Let's ensure that claims and airdrops are being indexed correctly

On startup Nomenclate will sync all the way up to the block just before tip, and then stop. Once we see a new block and get an event the syncing will resolve itself, but we definitely should fix this so that we are fully sync at startup.

We had a reorg on testnet3 block 6090, and this has essentially broken all address lookups.

We should have a downloadable insomnia file that allows people to test out the API. We should also see if it's possible to include a postman doc as well.

Currently with a few calls, we allow the client to send limit and offset variables to tell the server how many items to send back. Currently we sort the items by the highest height first. We should accept a new variable perhaps: direction that could be "ascending or descending"

Right now we don't accurately reindex blocks and txs during a reorg.

I would like to make this an additional feature which allows the client to index airdrops if they so choose.

This has been happening before -> I believe it may be due to reorgs, so we need to double check that addresses are handled correctly in the case of a reorg.

That being said, we probably also should get a test written for a reorg to test the history of addresses to see if this exactly what is causing the issue. If not, more investigation will be needed.

I think we can speed up our interface by utilizing gRPC

Can we make the interface even more efficient by implementing the QUIC Protocol? I'm curious to see the kind of progress we can make on this.

I'm seeing some weird time differences on HNScan between the time of the block and the index time of the auction transaction.

Let's do some testing to ensure that these times are the same.

For some reason while implementing the Rust bindings for Nomenclate, I discovered that some txs are being saved without heights.

Not sure what's causing this, but investigating now.

Right now Nomenclate only syncs from HSD nodes. It would be a useful feature to allow syncing from other Nomenclate servers.

This will need to be well planned out as I'm not sure what kind of security issues this could pose. We likely would still want to check anything that we sync from other servers, to ensure they aren't feeding us bad data.

Relevant to the conversation in the Handshake channel today, I'm wondering if there is any way for us to build a more decentralized whois server.

This could be incorporated into Nomenclate so that someone pinging the server would get a random distribution of nodes to (slightly) reduce the chance of someone recording all the searches for a specific name.

We should potentially buy and own the name whois/ to host this there.

Index name claims.