Why is there a difference between the editor's draft and the latest published version?

[brentzundel]

The DID WG resolved to automatically update the DID Specification Registries via echidna whenever a new PR is merged: https://www.w3.org/2019/did-wg/Meetings/Minutes/2020-06-09-did#resolution1

Since this is the case, why is the latest published version different than the editor's draft?

edited to add the links: editors draft: https://w3c.github.io/did-spec-registries/ latest published version: https://www.w3.org/TR/did-spec-registries/

[brentzundel]

@msporny @iherman do you know why this is?

[msporny]

Since this is the case, why is the latest published version different than the editor's draft?

In what way is it different? (if it is different, I expect it might be a caching issue wrt. the list of DID Methods .json file)

I just checked against the latest commit:

https://github.com/w3c/did-spec-registries/commit/35763b97f1774f6e563cf082676bbf8e06c4f99a

at this URL:

https://w3c.github.io/did-spec-registries/

and I note that document is up to date:

... ooooh, but hold on... ECHIDNA stopped auto-publishing back in Nov 2021:

https://www.w3.org/TR/did-spec-registries/

@iherman, I think this is an ECHIDNA issue... checking the config now.

[msporny]

Huh, looks like we never set up ECHIDNA for DID Spec Registries. :)

[brentzundel]

oops

[iherman]

oops ^ 2 ☹️

The good news is that, administratively, we have things in line:

• resolution to get it done: https://www.w3.org/2019/did-wg/Meetings/Minutes/2020-06-09-did#resolution1
• there is a 'secret token' in the repository, namely W3C_TR_TOKEN, which (I hope) contains the right value

I wonder if there was not already an action script properly set up at some point in the past and it got inadvertently wiped out when a number of action scripts were added...

I think some cut and paste using, e.g., https://github.com/w3c/did-rubric/blob/main/.github/workflows/auto-publish.yml might work.

Unfortunately, I am out until next Tuesday now; unless you guys pre-empt me until then we can get this settle next week.

[iherman]

I just see that there is an echidna script in worflows, namely auto-publish.yml, except that it combines the echidna workflows with some specific actions (and I mostly do these scripts in a cut-and-paste manner, I do not claim to have a clear view of what is happening...). Ie, forget about my comments https://github.com/w3c/did-spec-registries/issues/420#issuecomment-1051037609.

However, I see https://github.com/w3c/did-spec-registries/actions/workflows/auto-publish.yml which says "This workflow was disabled manually." I do not know who did that and why and I prefer not enabling it without knowing it. May it be as simple as you guys having forgotten to re-enable it?

[brentzundel]

I think echidna publishing still isn't working, can anyone confirm whether or not this is the case?

[OR13]

I'm always surprised when echidna works.

[pchampin]

Sorry for the delay in handling that. I confirm that Echidna is nor working, and the reason seems to be that the WG charter has expired.

https://github.com/w3c/did-spec-registries/actions/runs/5381071207/jobs/9765055284#step:7:691

I have requested a charter extension (actually, before I noticed this issue). This should fix it.

[wip-abramson]

I assume this is now resolved, can we close?

[msporny]

Looks like Echidna is still broken. We can't close this until that's fixed. I know we were waiting for the new /TR/ space documents to be published. That happened. I can also confirm that the Github page publications are working. However, the Echidna process is still failing (the Github Actions exist and are enabled, but they're failing to publish the specification):

https://github.com/w3c/did-extensions/actions/runs/12711711250/job/35435623241

Hmm, and it looks like they're failing because the new Github Actions were written such that they skip building the DID Method extensions file. I'll see if I can fix that now.

[wip-abramson]

Ah could it be to do with this issue #589

[msporny]

After a bit of work, we have a published version of DID Methods in TR/ space:

https://www.w3.org/TR/2025/NOTE-did-extensions-methods-20250120/

The issue had to do with:

• Not building DID Methods when trying to publish via Echidna (missing steps in the new Github Action)
• A malformed email field in the new did:webvh registration (multiple email addresses were registered, and we didn't allow that before)

I have fixed both bugs and we now have confirmed Echidna auto-publishing working for /TR/did-extensions/ and /TR/did-extensions-methods/.

This issue is now resolved, closing.