Hi everyone,
I wanted to share an idea I’ve been working on while building some Zcash projects like SocialMask. One of the biggest issues I’ve seen—both for new and experienced developers—is how often apps break when a lightwalletd endpoint goes down.
Most apps or wallets connect to one node, and when that node stops responding or falls behind the chain, the whole app can stop working. If users are sending funds or checking balances, this can even cause stuck transactions or temporary loss of access—something that’s already happened a few times in the ecosystem.
That’s why I’m proposing a client-side Auto-Failover SDK designed for lightwalletd.
It’s simple: the SDK lets apps automatically switch to another healthy endpoint without needing to restart, run full nodes, or depend on a central proxy.
Main points
- Multi-language support – works in JS/TS, Python, and PHP (plus a Rust helper). Not everyone builds in Rust, so this opens reliability to more teams.
- Smart failover – checks node latency, sync height, and health in real time, then switches automatically if one fails.
- Zcash-specific safety – idempotent retries, transaction deduplication, and optional TLS pinning.
- Lightweight – runs 100 % client-side; no coordinator, no new single point of failure.
- Better UX & lower costs – fewer stuck wallets, fewer support tickets, and smaller infra bills.
I’d really like to know if you think this could be useful for the ecosystem — especially for wallet teams or anyone using lightwalletd regularly.
If you want to read more or leave feedback, here’s the link:
Forum discussion
Thanks for reading, and I’d love to hear your thoughts!
— Emilio (@Emilio983)