Re: Interop meeting - today


Adrian Gropper
 

A CCG thread presents more rubrics for interoperability. Specifically, I'm suggesting we try to keep VC interop separate from DID interop. https://lists.w3.org/Archives/Public/public-credentials/2020Jul/0108.html

So, indeed, let's start with the hard questions first. What's the hardest question?

- Adrian


On Thu, Jul 30, 2020 at 9:22 AM Kai Wagner <kai@...> wrote:

Hello Adrian,

thank you for jumping right onto the thread.

The comparison you make is quite nice. Still, I think the profiles will be a bit more complex, especially when it comes to semantic- and trust interoperability/acceptance.

Looking at the technical interoperability, I believe this should not be seen as a narrowing of scope, but as a focus on outcome. In the real world, SSI vendors will not be successful because of their particular technical implementation, but because of their ability to deliver value to as many use cases as possible (while sufficing trust and compliance criteria).

Bringing USP driven competition to the layer of basic SSI interaction will likely lead to market dynamics that make interoperability less likely and might even lead to re-centralization and monopoly dynamics.

As said, I am looking forward to find a consensus on this, but agree with Orie that we need to start with the hard questions to set the scene and agree on expectations.

Best

Kai

Kai Wagner
Partnership Development

+49 176 83 588 604


Twitter: @kai_dentity


jolocom.io info@... Our events
Twitter Telegram GitHub YouTube Blog

Jolocom GmbH, Waldemarstrasse 37A, 10999 Berlin, Germany
CEO (Geschäftsführer): Joachim Lokhamp
Amtsgericht Charlottenburg (Berlin): HRB 158758 B
Am 30.07.20 um 13:11 schrieb Adrian Gropper:
The Jolocom definition reminds me of WiFi or Bluetooth where seeing the SSI logo would be all one needs to expect a certain level of functionality and where differences in the features of a WiFi or BT implementation are minor and not market differentiators. 

So, your SSI definition approach is fine but would such a narrowing of scope to many be a handful of “profiles” be acceptable to our community.

- Adrian

On Thu, Jul 30, 2020 at 5:17 AM Kai Wagner <kai@...> wrote:

Hello Taylor, @all

thank you very much for picking up the "define interop" discussion again.

At Jolocom, we have been thinking about this for a long time and it is one of our main decision making factor as we move on in developing our SSI technology.

While I hear the points made by Daniel Hardman on building interop only where relevant populations are present, I can not agree with it. In fact, the approach described by Daniel might even lead to a discussion on compatibility, rather than interoperability (see wikipedia citation below).

"interoperability imply Open standards ab-initio, i.e. by definition. Interoperability implies exchanges between a range of products, or similar products from several different vendors, or even between past and future revisions of the same product. Interoperability may be developed post-facto, as a special measure between two products, while excluding the rest, by using Open standards. When a vendor is forced to adapt its system to a dominant system that is not based on Open standards, it is not interoperability but only compatibility."

At Jolocom, we have defined interoperability like this:

"All agents / participants can communicate and interact with each other directly, without intermediaries, regardless of the agent / wallet implementations being used. The participants do so by supporting a common set of interfaces, and communicating using open and well-established specifications. Ideally this communication should not be facilitated by a 3rd party provided / maintained infrastructure, as this bears the risk of a lock in effect and centralization of the larger infrastructure."

With this definition as a starting point, we have tried to operationalize interoperability by describing the result of interop in the following statements:

  • As an Issuer/Verifier I can present one QR code to request interactions, regardless of what wallet/agent is on the other end (I never have to care).
  • As an identity subject/holder I can use any wallet I want, I can move across wallets, including encrypted backups.
  • An identity Holder can present a Verifiable Credential to a Verifier, regardless of which implementation the verifier or issuer uses, or which wallet the Holder uses.
  • As a Verifier, I can resolve DIDs from any DID method and use the DID Document to verify signatures, regardless of which DID method the document is from.
  • As a Verifier, I can check the integrity of the Credential regardless of the implementation used by the issuer of that credential as well as the holder.
  • As a developer, I can use any implementation of SSI tools (i.e. libraries, CLI tools) and expect it to perform its role with other deployments.

The above points are only a start and feedback is much welcomed. Ultimately, we think the discussion on interop need to lead to a modular and open ecosystem where interoperability is achieved by following standards and specifications, rather that "welding things together", which is handled in a "different IIW community".

I hope we can bring the interop-working group to work in that direction.

Looking forward to your thoughts and ideas.

Best

Kai

Kai Wagner
Partnership Development

+49 176 83 588 604


Twitter: @kai_dentity


jolocom.io info@... Our events
Twitter Telegram GitHub YouTube Blog

Jolocom GmbH, Waldemarstrasse 37A, 10999 Berlin, Germany
CEO (Geschäftsführer): Joachim Lokhamp
Amtsgericht Charlottenburg (Berlin): HRB 158758 B
Am 29.07.20 um 17:07 schrieb Taylor Kendal:
As follow-on to the “define interop” discussion. 

Is it possible/reasonable to land on a common denominator (highest level possible) which could broaden/narrow based on groups represented?

The ability of a system to work with or use the component parts of another system.


Or a bit more specific:


The ability for different systems, devices or applications to connect, in a coordinated manner, within and across organizational boundaries to access, exchange and cooperatively use data amongst stakeholders.


Then the inevitable moving targets fall under various sub-domains (foundational, structural, semantic, organizational, etc.) 

Also, as far as “outputs,” I think there’s intangible value in simply having a neutral venue for building/sharing context. 
*Recommended practice: join DIF interop wg :)

Food for thought...

Taylor Kendal, CPO


From: Balázs Némethi <balazs@...>
Sent: Wednesday, July 29, 2020 6:00:39 AM
To: interop-wg@dif.groups.io <interop-wg@dif.groups.io>
Subject: Interop meeting - today
 
Dear all,

A few reminders about today's meeting and next week's:
  • The nomination of chairs will close - if you have not, please nominate using this form
  • Election of chairs will happen over email.  You'll receive information after the meeting
  • It will be open for one week
  • Ballots will be checked against the attendance records of this group: 2 of the last four meetings when ballots close, i.e., last two meetings, today, and next week's.
  • During today's meeting, we'll do a little more issue review on the charter, a little discussion of the ballot, and the rest of the time will be to discuss workitems and roadmap
Best regards,
 --

Balázs Némethi
Operations @ DIF

Join interop-wg@lists.identity.foundation to automatically receive all group messages.