Our dependence on WalletConnect's infrastructure is x100 worse than our dependence on commodity RPC infrastructure like Infura/Alchemy. 😞

A doodler and computerer. I like permissive/permissionless open source, smart contracts, p2p systems, room-scale VR, and NixOS.

shazow.net

Currently: WhatsABI

This is because WalletConnect v1 was sunsetted, V2 requires API keys

1. That proves my point, V2 will be sunset eventually too and everything will break too.

2. This is actually V2, but the gated API key ran out of request quota so the dapp is broken for everyone. This is the degen.tips dapp right now btw.

Software Engineer @ Cadence Design Systems. B4: Stanford EE / Illinois ECE

Actually I'm not certain which endpoint it's using, but it does include a project ID.

Yeah, I prefer to use Injected Web3 wherever I can, but some wallets like Argent only support WC and it drives me up the wall

Yea, agreed we should always use injected RPC when available.

My concern is more that WC's RPCs are not commodity, they provide a bunch of proprietary signaling endpoints that aren't replaceable by other providers (they don't federate, it's centralized infrastructure).

"too many frontends hardcoded URLs" <-- that's exactly the problem, which continues to be perpetuated by design, right? How are we supposed to build immutable frontends in this environment?

I get that "p2p is slow" but *at the very least* it should fallback when the centralized "shortcut" provider is gone.

V1 was gossipsub over Waku and no API keys, the server code is still out there for anyone to run, but too many old front ends hardcoded their URLs. Hard to do p2p bootstrapping in-browser.