Re: 4-Corner Credential Interoperability Model

I am aware discussing the details of UBL is 
irrelevant to Michael's scenario, but I feel an 
urge to respond to something Steve brought up, in 
order to justify the good work of the committee I 
used to chair and other committees.

At 2023-02-05 18:14 +1100, steve capell wrote:
>Good point about digital document complexity 
>monsters.  I've never understood why UBL / 
>CEFACT / ISO etc feel the need to develop these 
>huge data models of things like a commercial 
>invoice - that has all properties that any use 
>case could ever want.  It's based on the premise 
>that, if you are going digital then you should digitise everything.

My UBL committee's responsibility is to digitize 
everything it thinks the wider sets of communities might need.

It is each community's responsibility to declare 
which subset they are going to use for the 
representation of the semantics they need to interchange. And ignore the rest.

Interoperability is offered when different 
communities choose to support the same semantics, 
thus they use the same standardized 
representations of those semantics, thus the 
committee has achieved its purpose. No longer do 
communities need to dream up labels and 
structures for semantics they all work with every 
day. That is the savings and the benefit brought 
from standardization of comprehensive (not complex) data models.

Never is it expected that any user community 
would implement the entirety of the specification.

I'm quite confident the same is true for the other committees cited.

>I think that's a mistake that adds huge cost and 
>presents a barrier to scalability.  The real 
>question is not what can be digitised but rather 
>what must be digitised for me as a buyer to 
>integrate the invoice into my ERP.  usually 
>that's just seller ID and invoice lines 
>containing productID, quantity, unit price, total.

Amen! And so that is all that a community needs 
to cherry pick from the specification. And two 
communities who cherry pick the same selection 
can interoperate where they have aligned.

I welcome discussion, though, really, it isn't 
relevant at all to Michael's network topology discussions.

.. . . . . . Ken

>   All the rest can stay in an attached PDF in 
> case a human wants to read it.
>
>On Sun, 5 Feb 2023 at 16:11, Anders Rundgren 
><<mailto:anders.rundgren.net@gmail.com>anders.rundgren.net@gmail.com> wrote:
>On 2023-02-05 5:31, Michael Herman (Trusted Digital Web) wrote:
> > 80% of it.  Iā€™ll have a full demo at 
> IIW.  Thereā€™s actually a small flaw in the 
> ā€œversion 0.65ā€ diagram in my previous post 
> because it is difficult to actually transform 
> and re-sign the credential on the right side as 
> well as re-auth-encrypt the DIDComm credential 
> exchange message with the re-signed 
> credential.  ā€œNecessity being the mother of 
> inventionā€ led to the creation of the concept 
> of DIDComm Agent logical composition: If you 
> have N DIDComm Agents connected serially from a 
> message transfer perspective, this scenario can 
> be recoded as a single physical DIDComm Agent 
> with N inbound service endpoints.  Using 
> DIDComm Agent logical composition, the 
> equivalent working version 0.66 solution looks like this
>
>Since I agree with what Steve Capell wrote about 
>message integrity etc., I'm curious about how 
>the revised solution copes with this.
>
>AFAICT, end-2-end security requires the 
>communicating parties agreeing on everything 
>from formats to cryptographic algorithms and 
>trust anchors.  This only works satisfactory for 
>closed-loop systems which is why I began toying 
>with public discovery services.  I have had this 
>running (albeit only in lab-scale...) since 2015.
><https://www.linkedin.com/posts/andersrundgren_many-payment-authorization-systems-build-activity-7002012686137835520-p-2X>https://www.linkedin.com/posts/andersrundgren_many-payment-authorization-systems-build-activity-7002012686137835520-p-2X
>Since discovery service information can be 
>cached, the overhead is quite limited.  Yes, 
>cache refreshes of course add some minor delays every now and then.
>
>IMNSHO, the ISO 20022 folks are on the wrong 
>track; extremely complex messages (with tons of 
>"nice to have" options subject to 
>interpretation), is primarily an asset for consultants :)
>
>Anders
>
>
> >
> > *From:* Wayne Chang 
> <mailto:wayne@spruceid.com>wayne@spruceid.com <mailto:wayne@spruceid.com>
> > *Sent:* Saturday, February 4, 2023 10:08 PM
> > *To:* Michael Herman (Trusted Digital Web) 
> <<mailto:mwherman@parallelspace.net>mwherman@parallelspace.net>
> > *Cc:* Christopher Allen 
> <<mailto:ChristopherA@lifewithalacrity.com>ChristopherA@lifewithalacrity.com>; 
> G. Ken Holman 
> (<mailto:g.ken.holman@gmail.com>g.ken.holman@gmail.com) 
> <<mailto:g.ken.holman@gmail.com>g.ken.holman@gmail.com>; 
> Steve Capell 
> <<mailto:steve.capell@gmail.com>steve.capell@gmail.com>; 
> public-credentials 
> (<mailto:public-credentials@w3.org>public-credentials@w3.org) 
> <<mailto:public-credentials@w3.org>public-credentials@w3.org>; 
> <mailto:sam@prosapien.com>sam@prosapien.com
> > *Subject:* Re: 4-Corner Credential Interoperability Model
> >
> > Cool, do you have any code to run this?
> >
> > On Sat, Feb 4, 2023 at 19:38 Michael Herman 
> (Trusted Digital Web) 
> <<mailto:mwherman@parallelspace.net>mwherman@parallelspace.net 
> <mailto:mwherman@parallelspace.net>> wrote:
> >
> >     If you use some DIDComm imagination, it's 
> easy to see how the 4-Corner Credential 
> Interoperability Model is highly complementary 
> to supporting a layered VC model ...here's a glimpse.  "More news at 11..."
> >
> >     -----Original Message-----
> >     From: Steve Capell 
> <<mailto:steve.capell@gmail.com>steve.capell@gmail.com 
> <mailto:steve.capell@gmail.com>>
> >     Sent: Saturday, February 4, 2023 4:24 PM
> >     To: Michael Herman (Trusted Digital Web) 
> <<mailto:mwherman@parallelspace.net>mwherman@parallelspace.net 
> <mailto:mwherman@parallelspace.net>>
> >     Cc: public-credentials 
> (<mailto:public-credentials@w3.org>public-credentials@w3.org 
> <mailto:public-credentials@w3.org>) 
> <<mailto:public-credentials@w3.org>public-credentials@w3.org 
> <mailto:public-credentials@w3.org>>; G. Ken 
> Holman 
> (<mailto:g.ken.holman@gmail.com>g.ken.holman@gmail.com 
> <mailto:g.ken.holman@gmail.com>) 
> <<mailto:g.ken.holman@gmail.com>g.ken.holman@gmail.com 
> <mailto:g.ken.holman@gmail.com>>; Christopher 
> Allen 
> <<mailto:ChristopherA@lifewithalacrity.com>ChristopherA@lifewithalacrity.com 
> <mailto:ChristopherA@lifewithalacrity.com>>; 
> <mailto:sam@prosapien.com>sam@prosapien.com <mailto:sam@prosapien.com>
> >     Subject: Re: 4-Corner Credential Interoperability Model
> >
> >     I think the 4-corner model is the 
> opposite of the VC model.  Itā€™s a EDI message 
> routing framework that delegates the problem of 
> trust (ie trader identity verification) to EDI 
> hubs.  Hubs perform a ā€œserviceā€ to their 
> authenticated subscribers to map messages to 
> the format they need - so forget about document 
> integrity (how do you maintain a signature when 
> the document is transformed at two 
> hubs?).  Thereā€™s nothing verifiable about the 
> message that eventually lands with the receiver 
> (through 2 hubs).  Instead the parties have to 
> trust that the hubs have properly identified 
> their clients and have not lost anything in translation
> >
> >     Itā€™s also expensive (every message 
> attracts  two clip-fees).  Iā€™ve never seen 
> this model achieve any significant uptake 
> except in cases where it is mandated by a national regulator
> >
> >     Kind regards
> >
> >     Steven Capell
> >
> >     Mob: 0410 437854
> >
> >      > On 4 Feb 2023, at 10:49 pm, Michael 
> Herman (Trusted Digital Web) 
> <<mailto:mwherman@parallelspace.net>mwherman@parallelspace.net 
> <mailto:mwherman@parallelspace.net>> wrote:
> >
> >      >
> >
> >      > ļ»æA colleague of mine, Ken Holman 
> from Ottawa, recently produced a conference 
> paper on the topic of the 4-Corner Credential 
> Interoperability Model. It's originated in the 
> OASIS Universal Business Language (OASIS-UBL) 
> community where they were experiencing problems 
> with spec compliance, interoperability, and 
> low/slow technology adoption rates. UBL is a 
> set of 90+ schemas defined for the most 
> commonly used business documents used in commerce today.
> >
> >      >
> >
> >      > The 4-Corner Credential 
> Interoperability Model is a simple, efficient, 
> low-cost approach to enable different 
> communities or trading associations to easily 
> interoperate through the exchange of different 
> types of credentials. For example, this would 
> be a stepping stone for enabling a layered 
> approach to how verifiable credentials are 
> defined, used, and exchanged - an opportunity 
> to move away from a "one-size fits all" niche 
> VCDM specification to a layered model that addresses the dual needs of:
> >
> >      > a) the layering of simple, more 
> generic models through to more complex
> >
> >      > niche credential specifications,  as well as
> >
> >      > b) an interoperability model that is 
> simple, secure, efficient, and low-cost.
> >
> >      >
> >
> >      > Live Presentation: Case study of a semantic library underpinning the
> >
> >      > 4-corner model for document exchange 2022-11-08
> >
> >      >
> >
> >      > 
> <https://www.linkedin.com/posts/gkholman_ken-holman-case-study-of-a-sem>https://www.linkedin.com/posts/gkholman_ken-holman-case-study-of-a-sem 
> <https://www.linkedin.com/posts/gkholman_ken-holman-case-study-of-a-sem>
> >
> >      > antic-library-activity-7001714323223977985-IKbA
> >
> >      >
> >
> >      > Case study of a semantic library underpinning the four-corner model
> >
> >      > for document exchange
> >
> >      > 
> <https://doi.org/10.1075/da.2022.holman.four-corner-model>https://doi.org/10.1075/da.2022.holman.four-corner-model 
> <https://doi.org/10.1075/da.2022.holman.four-corner-model>
> >
> >      > Proceedings of Declarative Amsterdam 2022 (7 and 8 November 2022)
> >
> >      > Available under the CC BY 4.0 license.
> >
> >      >
> >
> >      >
> >
> >      >
> >
> >      > Best regards,
> >
> >      >
> >
> >      > Michael Herman
> >
> >      >
> >
> >      > Web 7.0
> >
> >      >
> >
> >      >
> >
> >      > <winmail.dat>
> >
>
>
>
>--
>Steve Capell


--
Contact info, blog, articles, etc. http://www.CraneSoftwrights.com/m/ |
Check our site for free XML, XSLT, XSL-FO and UBL developer resources |
Streaming hands-on XSLT/XPath 2 training class @US$125 (5 hours free) |
Essays (UBL, XML, etc.) http://www.linkedin.com/today/author/gkholman |

Received on Monday, 6 February 2023 14:02:29 UTC