PROPOSAL: Universal values for a *null DID* as well as an *empty DID*

[mwherman2000]

These concepts are very common place in the world of software development today.

e.g. null string = 0; empty string = "";
e.g. null array = 0; empty array = [];

PROPOSAL: Universal values for a null DID as well as an empty DID (universal => DID Method independent):

null DID = "did:0";
empty DID = "did:";

including pre-defined functions like: bool nullempty = Did.IsNullOrEmpty(someDIDvalue);

Any other ideas/proposals?

Michael Herman
Web 7.0™ / TDW AgenticOS™

[davidlehn]

There's a spec and specific implementation for the did:null method:

• https://digitalbazaar.github.io/did-method-null-spec/
• https://github.com/digitalbazaar/did-method-null

The entire concept was rejected from the registry:

• Add did:null method. did-extensions#431
• Add did:null method. did-extensions#555

Maybe I'll refresh and try again on April 1. :-)

did:0 and did: are not valid based on the DID spec. I saw no reason to challenge that and picked did:null:null for naming as it seemed to best align with web platform terminology and DID spec expectations. I figured character counts were not really important. Mostly because the idea was either as a joke or as something real that is simple and static that does nothing but could be used for a minimal base case for test suites or checking validators or something.

[mwherman2000]

@davidlehn Thank you for engaging. Your comment suggest an alternative. This is no joke.

did:constant:null
did:constant:empty

[mwherman2000]

@davidlehn Extending the list...

did:constant:null
did:constant:empty
did:constant:true
did:constant:false
did:constant:yes
did:constant:no
did:constant:na
did:constant:a
did:constant:b
did:constant:c
...
did:constant:z
did:constant:sunny
...
did:constant:clear
did:constant:happy
...
did:constant:wink

[wip-abramson]

Thanks for the proposal @mwherman2000.

Unfortunately, there are a number of reasons why this proposal is unlikely to get support from the DIDWG.

First, the DID specification is in maintenance mode at the moment meaning only up to class three changes can be made to the spec. No new features are allowed. This would be a new feature. So at best, it would have to wait until the DIDWG is rechartered to work on a version 2 of the DID spec.

Second, I am not clear on the use case justifying these universal values you are proposing. For any feature to be seriously considered for addition to the DID specification we would need to see clear demand from multiple stakeholders as well as justification for why it should be included within the core DID spec.

The entire concept was rejected from the registry:

As @davidlehn highlights, the DIDWG previously has not shown much enthusiasm for such a method. Neither have implementers, to my knowledge, been asking for such a DID method independent feature to be standardized.

That said, I assume you have a need / usecases for did:constant:null or similar. There is nothing stopping you from pursuing these within your own context. Even add another DID method to the W3C list if you wish. Then if others also recognize the need for this, they can adopt your approach.

I am failing to see why this belongs in the DID specification, even if we were chartered to make such a change.

[w3cbot]

This was discussed during the #did meeting on 05 February 2026.

View the transcript

w3c/did#914

manu: did:null was submitted as a proposal as an April Fools Joke, but kinda serious. Signals could be worked on at the Method level. From experience, we have never needed a null/empty DID, so not clear there is a need. Could, and Methods may have a way to express.

Wip: Did look at it, consensus was you don't need this. Suggest that issue submitter propose a DID Method.

<Zakim> JoeAndrieu, you wanted to follow URLs/URIs

smccown: Agree let him submit. Have not seen a reason -- other than Comp Sci often needed. Not clear how serious given the other examples he provides. Perhaps just going for completeness, but I don't see the point.

JoeAndrieu: Make own did Method -- that said, there is no such thing as a "null" URL and a DID is a URL.

dmitriz: re: did:sunny, etc. -- URN is for that -- point him to that.

---

[jandrieu]

I agree. This doesn't belong in DID Core.

First, there's an easy way to build this out by creating a DID method. That's probably your easiest route to proving the viability of your approach @mwherman2000.

Second, DIDs are URLs, which have survived 30+ years of use without needing some formal way to represent a "null" URL. So, I'm not sure its worth the effort to standardize a NULL value in this way.

That said, the URN is the current open, internet standard for defining namespaces of fixed values. You could create a new URN namespace and put those constants there, e.g., urn:did:null can be established to mean a "null" DID, whatever that means, whether that is taken to return a specific DID document from resolution or to be understood to be treated as a NULL.

In any case, I don't see how it fits into the DID Core without breaking the boundary that DID Core does not specify DID methods.

Also, I was confused by definitions like "did:constant:yes". What is returned as the DID document when resolving this DID?

IMO, if the DID can't resolve to a DID Document, it's not really a DID.

[ottomorac]

Based on group consensus from the WG meeting yesterday:

• This does not belong or is impeded by the DID Core spec
• There is an alternative option that the issue submitter propose a DID Method if he thinks it relevant

The issue is therefore closed.

[mwherman2000]

Second, DIDs are URLs, which have survived 30+ years of use without needing some formal way to represent a "null" URL. So, I'm not sure its worth the effort to standardize a NULL value in this way.

Completely elevant comment. Domain names were never used to name Every Little Thing (ELT) on the planet.

For the record: an estimated 98-99% of URLs are domain names (350–370 million registered domain names in total) vs. trillions upon trillions of ELTs.

Some Trivia: there are at least 6-8 different subtypes of myopia: https://chatgpt.com/s/t_69874f5b6b188191bfbbf5829dcbddc7

[mwherman2000]

We can add them to the DID7 spec (https://datatracker.ietf.org/doc/html/draft-herman-did7-identifier-00):

did7://web7/constant_1_0:null
did7://web7/constant_1_0:empty
did7://web7/constant_1_0:nullid
...
did7://web7/constant_1_0:true
did7://web7/constant_1_0:false
did7://web7/constant_1_0:yes
did7://web7/constant_1_0:no
...
did7://web7/constant_1_0:weather_1-0:sunny
did7://web7/constant_1_0:weather_1-0:partly-cloudy
did7://web7/constant_1_0:weather_1-0:rain
did7://web7/constant_1_0:weather_1-0:snow
...
did7://web7/constant_1_0:emoji_1-0:happy-face
did7://web7/constant_1_0:emoji_1-0:angry-face
did7://web7/constant_1_0:emoji_1-0:teary-face
...
did7://web7/constant_1_0:actions_1-0:approve
did7://web7/constant_1_0:actions_1-0:reject 
...
did7://web7/workflow_1-0:states_1_0:new
did7://web7/workflow_1-0:states_1_0:ready
did7://web7/workflow_1-0:states_1_0:in-process
did7://web7/workflow_1-0:states_1_0:blocked-waiting
did7://web7/workflow_1-0:states_1_0:completed
did7://web7/workflow_1-0:states_1_0:failed
did7://web7/workflow_1-0:states_1_0:canceled 
did7://web7/workflow_1-0:states_1_0:timed-out

Web 7.0 supports the authority-based version of the following DID model:

Image: Copyright © 2026 Michael Herman (Bindloss, Alberta, Canada) – Creative Commons Attribution-ShareAlike 4.0 International Public License

[mwherman2000]

Addressed/solved here: https://www.ietf.org/archive/id/draft-herman-did-w3c-drn-00.html