WICG
Define Digital Wallet
Closes #???
Introduces a formal definition for "Digital Wallet", clarifying its role and relationship to the "holder" concept. These changes improve clarity, precision, and alignment with the specification's terminology. Consistently use the defined term "[=digital wallet=]" in place of previous informal references like "digital wallet" or "wallet".
This should be part of the existing terminology work, and defined as a type of credential manager.
#352
https://github.com/w3c-fedid/digital-credentials/tree/tc-352-terminology
#352 is an issue, this addresses part of it.
See my comments there. A separate PR for “client platform” would be appreciated as part of the Model.
“Credential Manager” is (or should be) defined by the Credential Management spec, not our spec. We can then link to it.
We need to coordinate across the three working groups to decide who owns each definition.
@msporny, hi! 👋 can you help us here? We would like define “Digital Wallet” formally in a spec and wrap “holder”.
We are happy for VC 2.0 to own the definition if it’s generic enough, but right now what’s defined in VC 2.0 doesn’t exactly match how we define a “digital wallet” (this PR).
let’s coordinate 🤗
An alternative name to a “wallet” could be a “credential provider” (what on Apple’s platform’s is called a “Document Provider” by the Identity Document Framework).
@mohamedamir, what’s the equivalent on Android? Any idea?
Perhaps then “digital wallet” can remain a colloquial term… as apps that hold credentials also do a lot of other things (e.g., the MyGov app from the Australian government has a wallet but also lets me lodge my taxes, has an “inbox”, and access a range of government services).
Here’s a screenshot showing the different things it provides:
Proposing we close this PR in favor of the existing issue and PR (#387).
Agenda+ to discuss closure.
@timcappalli, as I mentioned in #387 (which was created after this PR), we should add one definition at a time. See also my comments in the relevant issue.
@mohamedamir, what’s the equivalent on Android? Any idea?
Credential Provider is the closest IMHO.
FWIW: the API itself is called Holder API.
@msporny, hi! 👋 can you help us here? We would like define “Digital Wallet” formally in a spec
Hi, yes, we have an exported term for that called a credential repository and note in the definition that a "digital wallet" is a type of credential repository. It's a somewhat awkward term that took several months of bike shedding to get to:
https://www.w3.org/TR/vc-data-model-2.0/#dfn-credential-repositories
That said, we have an active VCWG updating the v2.1 specs and are happy to update the terminology to reflect the current state of affairs based on the needs of the DC API spec. I expect we'll want to tweak that definition a bit.
With all my hats off, though, I'll warn against the usage of the term "digital wallet" as the primary term. It's fine as an imperfect synonym but might not be the term we'll want to use in 5-10 years. The reasoning being that there are other sorts of apps (and websites) that might hold digital credentials that are not considered digital wallets. For example, ride hailing apps or retailer apps do not consider themselves "digital wallets" but are likely to hold digital credentials in the future. We might want to shift to say something like "credential management application", where a "digital wallet" is a subset of that class of application and a "credential repository" is the storage portion of that application.
There is a lot of this discussion happening in the Verifiable Credential API for Lifecycle Management incubation work that's been happening for the past several years as well.
I note some disagreement in this PR, so will wait until you guys explicitly ask the VCWG to raise an issue and modify the spec. I expect once we have that decision, we can rapidly make changes to align terminology to what the DC API spec needs.
and wrap “holder”.
What do you mean by "wrap"? We have an exported term for holder here:
https://www.w3.org/TR/vc-data-model-2.0/#dfn-holders
The terminology in the VC spec probably needs some rework at this point -- define digital wallet, define digital credential, specify that mdocs, VCs, etc. are a subclass of digital credential... specify that digital wallet is a subclass of credential management application (or whatever we can find consensus on), etc.
In any case, the VCWG is ready, willing, and able to coordinate on this.
@msporny
We might want to shift to say something like "credential management application", where a "digital wallet" is a subset of that class of application and a "credential repository" is the storage portion of that application.
Agreed, and that's why credential manager has been used across many other documents in this space, and why credential manager is part of the Terminology PR (#387).
There is no official ask from the Fed ID WG to the VC WG at this time. The ask appears to be from an individual without working group consensus.
@msporny wrote:
What do you mean by "wrap"? We have an exported term for holder here:
What we mean is that a "digital wallet" is a "holder", but also does other stuff (per what's defined in this PR). But that it is a holder is the critical bit.
@timcappalli wrote:
The ask appears to be from an individual without working group consensus.
All PRs start mostly without group consensus so that we can have something to reach consensus on (I'm not aware of any group consensus to send #387 - particularly given the feedback provided in https://github.com/w3c-fedid/digital-credentials/issues/352#issuecomment-3209298982).
It is common place to coordinate across groups where there's strong overlapping interests. Cross-collaboration across working groups is generally considered good practice (or at least, being good neighbors), particularly where there is strong shared interests and where we already depend on their work (as we do on the VC group).
It would be a dereliction of duty, as an Editor of this specification, to not reach out to the VC folks at this stage for their input.
@msporny wrote: What we mean is that a "digital wallet" is a "holder", but also does other stuff (per what's defined in this PR). But that it is a holder is the critical bit.
Imma be pedantic only because we're talking about definitions that are probably going to stick.
Per the VCWG definitions -- "Holder" is a role that an entity plays -- that role has the responsibility of holding on to digital credentials. They do so via an application (software or hardware)... so I'd say that a "digital wallet" is a type of "credential management application" used by a "holder" for the purposes of receiving, holding, managing, and presenting digital credentials.
+1 to get the VCWG involved -- we love bikeshedding terminology like this (for better or worse).
Pedantic is good! And yes, we totally want longevity with these definitions.
Ok, let me put that into a cocktail mixer and see what comes out. 🍸
appreciate the help @msporny.
Ok, @msporny, tried to integrate your suggestions and clarifications.
@timcappalli, I think this is ready to go. Please take a look.
@TallTed, mind also having a look?
@timcappalli, I wonder if we should do this in two parts? Your suggestions seem to broaden the scope of the privacy section to the more general "credential manager" concept, which MAY be ok - but I think we should be really careful with broadening the terminology,
@johannhof, can you check also if you concur with the changes. I'm weary of changing things in the privacy section with @timcappalli's suggestions unless you also think it's ok to use the broader credential manager definition.
Happy to discuss or we can spin up a separate PR?
I'd personally prefer that we do this in two PRs, even if we are rubber-stamping the changes from [=digital wallet=] -> [=credential manager=] in various places.
@timcappalli, is that ok? As we discussed at TPAC, it would be ideal to keep changes as incremental as possible.
Discussed during 25 November 2025 call. Suggestion was to separate the PRs to be explicit and separate re: the definition of the term, the renaming of the term, and the changes to other sections in the doc.
FWIW this PR is fine with me, regardless of whether we go all the way as Tim suggests or keep a few "wallet" uses as Marcos suggests.
Discussed briefly at https://github.com/w3c-fedid/meetings/blob/main/2025/2025-12-02-DCAPI-A-notes.md#define-digital-wallet-386. Consensus on using "credential manager" in spec text. Need to make sure this is changed here and in separate PRs in other areas of the doc that currently use "Digital Wallet."
@marcoscaceres Does that mean we can close this?
Yes, let's close this and I'll create a new PR.
#412, #413, #414