Questions about Avail DA construction

Hi all, I have a question regarding Avail’s DA construction after reading through the docs. If I understand correctly, Avail use polynomial commitment (KZG) to commit erasure-coded data so it does not require fraud proof to prove misconstruction of erasure-coded data. But in the paragraphs here it says that fraud proof is used to prove misconstruction and block producer creates a Coded Merkle Tree commitment.

Is the doc perhaps outdated or am I missing something?


As far as i know there is dual mechanism in Avail. One is Data Availability Sampling (DAS) to verify block data integrity and other one is Kate commitments to network-wide verification or creating fraud proofs.
1- DAS method eliminates the need to recreate KZG commitments, don’t need to rely on fraud proof. Avail’s high security and data integrity standards maintained by decentralized verification.
2- Kate commitments create fraud proofs to challenge any discrepancies in the data.

This dual mechanism of light clients and full nodes working in tandem also strengthens the overall security and reliability of the network.

Hey, it looks like that is an error from our end. We do not use CMTs, like you said initially we use KZG commitments which ensure that data is erasure-encoded correctly, hence no fraud proof required. Thank you for pointing this out. The change has been made to reflect the docs.

1 Like

No problem. Thanks for the swift response and update!

1 Like

I thought both CMT and KZG commitments are used, thanks for the information :pray:

1 Like