Blog
Updates, guides, and insights from the Lodestar team.
How to Build and Deploy a Data Service on The Graph's Horizon Framework
A complete technical walkthrough of building a production Horizon data service — from Solidity contract to off-chain payment collection — drawn from two real implementations: Dispatch and SubstreamsDataService.
Dispatch vs Alchemy vs Infura: Real RPC Pricing Numbers
Concrete, code-derived cost comparisons between Dispatch and the major centralised RPC providers — plus a step-by-step guide to getting an RPC URL from the Dispatch network today.
How to Run a Local Amp Node, Part 2: What You'll See After a Day
Part 1 covered installing ampd and deploying a dataset. This post is about what happens next — what Amp actually writes to disk, what it costs in RAM, and the operational details the setup docs don't mention.
Dispatch: An Experimental Community JSON-RPC Service on Horizon
Every dApp on Earth relies on Alchemy or Infura for JSON-RPC. Dispatch is a community-built experiment exploring what a decentralised JSON-RPC data service on The Graph's Horizon protocol might look like — the first provider is live and serving real traffic on Arbitrum One.
Graphite: Write The Graph Subgraphs in Rust
AssemblyScript is the only language graph-node accepts — or so everyone assumed. Graphite compiles Rust to WASM that graph-node cannot tell apart from AssemblyScript. No patches. No forks. Already live on Arbitrum One.
GRT at the Floor: Illiquidity, Inflation, and the Long Game
The Graph's token is near historical lows. Here's the honest story of how it got there — and what, if anything, changes next.
How Much RAM Does It Actually Take to Index 1,000 Subgraphs?
The Graph's 184 GB 'Large' tier was written four years ago for a few hundred subgraphs. Here's what the memory math actually looks like for 1,000 — and what it would take to index all 15,500.
How to Run a Local Amp Node
Edge & Node's Amp is available as a hosted service — but you can run it yourself. Here's how to self-host a local ampd node, with no waitlists and no dependency on E&N infrastructure.
Arbitrum Indexer Infrastructure: Smaller Archives & Fixing Sync Lag
Two problems, one post: the new PathDB/PebbleDB archive node that cuts disk from 38TB to 4TB, and the two graph-node settings that stop your subgraphs sitting 10–30 blocks behind chain head forever.
Intro to Amp: The Graph's Blockchain-Native Database
Edge & Node built a blockchain-native database that turns smart contract events into SQL tables automatically. Here's what Amp is, how it works, and what it means for The Graph ecosystem.
Intro to Tycho and the Tycho Data Service
Tycho indexes DEX liquidity in real time so solvers, market makers, and searchers don't have to. Here's what it is, how it works, and why it's coming to The Graph Network.
The 3 Rules of Graph-Node Stack Architecture
There are many ways to set up your graph-node stack, but the constraints boil down to 3 rules. A practical guide to node roles, sharding, tiered indexing, and the pitfalls that catch everyone.