Re: did:webs vs did:tdw

>
> From what I’ve gathered so far, the main difference is that did:webs
> relies on KERI while did:tdw does not. Is that the only notable difference?
>

The reason that did:tdw does not have a dependency on KERI is because it
avoids the witness and watcher features that a full KERI implementation
would need. This is best understood as a choice to forego certain
redundancy, duplicity detection, and error recovery options in order to
avoid some complexity and ease adoption. I don't want to argue the merits
here, but I do want to make sure that the choice is not framed as the
choice between two identical approaches that differ only in that one has a
dependency and the other doesn't. I think evaluators would be well served
to understand the theory of witnesses and watchers as they evaluate.

This is coming from someone who worked on did:webs, but who also admires
all of Stephen's amazing contributions to the community, so take that with
whatever political baggage seems appropriate. :-)