Henry Tsai

Making the failure explicit is better than the current approach of inferring failure based on transaction number, it will also be more consistent with the pattern used downstream.

This will remove 99% of the storage usage used by the Bitcoin Core client.

Currently in reference implementation, a Sidetree request gets queued up and waits for the Batch Writer to write it to the network. This can take minutes and even hours depending...

Should respond that the DID is invalid/unresolvable as per v1 sidetree spec.

Add precise language on DID unique suffix generation and test vector

12

Example test vectors can be found at: https://tools.ietf.org/html/rfc3394 Suffix = ASCII(Base64URL(Multihash(SHA256, UTF8(Suffix JSON data))))

Align implementation with resolution metadata portion of the spec

5

The implementation deviated from the latest changes to the resolution metadata portion of the spec. e.g. `deactivated` property is newly added in the spec.

Add client guidelines in main spec

1

Replicate/move content see in https://github.com/decentralized-identity/sidetree/blob/master/docs/design.md#sidetree-client-guidelines in main spec.

Add compressed file size check in Batch Writer

1

The Batch Writer should ensure that the **compressed** CAS files it writes does not exceed protocol limits so that the operation batch is not rejected by the Observer. Also consider...