BbsSignatureProof2020 is broken

[OR13]

The current libraries for handling the proof of knowledge are sensitive to disclosed message order, in ways that make certain derived credentials (ones over multiple blank nodes) not verifiable.

This is not a break in the crypto, it is an assumption about revealed statement order which needs to be accounted for.

It's not the job of this suite to define this, but we can't call whatever the solution is BbsSignatureProof2020.... so we will need a new name.

For now, folks interested in this should head over to https://github.com/decentralized-identity/bbs-signature to help.

I suggest we simply rev the year and keep the name the same, so the new proof type will be:

BbsSignatureProof2021, and it won't be sensitive to revealed statement order.

[OR13]

See also: mattrglobal/jsonld-signatures-bbs#158

[OR13]

This issue is still unresolved and the BBS+ work at DIF has chosen to address other issues instead of this issue.

[OR13]

I'm calling this resolved, since it has been removed from the latest draft.