#sxt $SXT I Don't Fully Trust Oracles, and Neither Should You
We built contracts that can't be tampered with, then plugged them into data pipes we basically take on faith. Price feed, indexer, some API somebody runs. It works until it doesn't, and when it doesn't, the loss is real.
@SpaceandTime is trying to fix that part. Proof of SQL lets you run a query and get back a proof that it was computed correctly on data nobody edited. Contract checks the proof. No "trust us."
Why not just throw everything into a zkVM? Proving big data queries that way is slow and costs too much. SXT went narrow instead, one job, done fast. Their benchmarks claim sub-second proving on GPUs. Benchmarks are benchmarks, so I'd wait for real workloads before getting excited.
Chainlink gives you consensus-based feeds. The Graph indexes, but nothing's proven. Brevis and Axiom prove historical chain state well. SXT covers verified analytics across onchain and offchain data. Real category or packaged niche? Can't tell yet. Microsoft backing is nice, but it doesn't create demand.
$SXT only matters if people actually pay for queries. I'd rather see usage and fees creep up than another chart pump. And devs take months to integrate a new data layer.
Not calling it the future of anything. I'm watching whether anyone uses it.
@SpaceandTime
We built contracts that can't be tampered with, then plugged them into data pipes we basically take on faith. Price feed, indexer, some API somebody runs. It works until it doesn't, and when it doesn't, the loss is real.
@SpaceandTime is trying to fix that part. Proof of SQL lets you run a query and get back a proof that it was computed correctly on data nobody edited. Contract checks the proof. No "trust us."
Why not just throw everything into a zkVM? Proving big data queries that way is slow and costs too much. SXT went narrow instead, one job, done fast. Their benchmarks claim sub-second proving on GPUs. Benchmarks are benchmarks, so I'd wait for real workloads before getting excited.
Chainlink gives you consensus-based feeds. The Graph indexes, but nothing's proven. Brevis and Axiom prove historical chain state well. SXT covers verified analytics across onchain and offchain data. Real category or packaged niche? Can't tell yet. Microsoft backing is nice, but it doesn't create demand.
$SXT only matters if people actually pay for queries. I'd rather see usage and fees creep up than another chart pump. And devs take months to integrate a new data layer.
Not calling it the future of anything. I'm watching whether anyone uses it.
@SpaceandTime
