Re: Call for DID Method implementations

On 5/18/21 6:57 PM, Kim Hamilton wrote:
> I propose:
> 
> * Pushing the deadline out based on input from implementers (like us) *
> Allowing case-by-case extensions
> 
> Please respond to this thread if you'd also like an extension for your
> method, and we can try to figure out whether a global extension or
> case-by-case makes sense.

Hey Kim, thanks for the feedback. I'll take this back to the DID WG to make
sure they consider the request (or feel free to raise an issue on the DID test
suite).

To be clear about the deadline, this is the deadline for implementers to
provide feedback during the W3C Candidate Recommendation phase... all the DID
WG is looking for here is that at least two independent implementers
implemented each feature in the DID Core specification. If we get more
implementations for each feature, great... but it's unnecessary for this stage
of the standardization process.

More importantly:

Any implementer can submit implementations at any point from now until the DID
test suite ceases to exist, which is expected to be around for as long as DIDs
continue to be a thing.

So, don't mistake the CR implementation deadline for some notion of never
being able to have your DID Method implementation added to the test suite...
the option of adding your implementation to the DID test suite will be there
in the future for those that won't have their DID Methods ready by the CR
deadline.

More elaboration below... but folks that have a passing interest can stop
reading now.

---------------------------------------------------------

To put this in perspective, we already have 9+ independent and conforming
implementations of DID Syntax ABNF rules:

https://pr-preview.s3.amazonaws.com/shigeya/did-test-suite/pull/105.html#conformance-by-test-suites-2

... but as you can see just below that link above... zero implementations of
the `hl` DID Parameter, `service` DID Parameter, and `relativeRef` DID
Parameter. It's those latter items that we're currently concerned about... if
we don't get at least two independent implementations of those features, we
will have to consider removing them from the specification before advancing it.

If BTCR (or any DID Method) is going to implement features that are not
implemented at present, then the WG would probably wait on BTCR to implement.
If BTCR is going to implement the same features everyone else has implemented,
then the WG is far less likely to wait on BTCR.

The DID WG's charter ends in September and we need at least two months before
then to wrap things up and move the spec to PR, get an official vote from W3C,
make any W3C Member change requests at that point (possibly moving the item
back into the WG), and then publishing the final REC... and that means we
really should be completely done by the end of June, and that's not even
counting if we have to go through another CR, which will take another month...
which means we should be out of CR by the end of May.

All that to say -- we're cutting it extremely close. I'm not going to be
hyperbolic and say that if the spec isn't done by the time the charter expires
that we won't have a DID spec... it just creates an enormous amount of pain on
the DID WG, W3C Membership, W3C staff and the rest of the folks that have to
coordinate a charter extension. Those sorts of things are painful processes at
W3C on purpose -- they consume time and resources that could be used on groups
that did what they said they would do during the time that was allotted to them.

At this point, the thing that is most likely to push the deadline out is lack
of implementation support for features the DID WG thinks are key to the
specification.

Hope that helps shed some light on what this deadline is, how implementers
should think about it, what is likely to push the deadline out, and the
dangers of doing that.

-- manu

-- 
Manu Sporny - https://www.linkedin.com/in/manusporny/
Founder/CEO - Digital Bazaar, Inc.
blog: Veres One Decentralized Identifier Blockchain Launches
https://tinyurl.com/veres-one-launches

Received on Wednesday, 19 May 2021 00:04:48 UTC