W3C

Market Data Rights Automation Teleconference

17 March 2021

Attendees

Present
alex, ali, andrew_dunkley, andrew_mor, atiq, belen, ben, caspar, david_sarkis, elizabeth, emily, fernando, gerardo, ian, ian_bone, Ilya, jairy, jeremy_bock, jo, jorge_di_pasqua, josh, justinas_alisaukas, kyle, laura, letitia, markb, marko, michael_aagon, michelle, nigelp, olga, paul, paulc, richard_F, roy, wilson
Regrets
renato
Chair
Jo
Scribe
jo, joshuaCornejo

Meeting minutes

Admin

<jo_> https://w3c.github.io/market-data-odrl-profile/agendas/md-odrl-profile-agenda-2021-03-17.html

Resolution: Accept minutes of last meeting

Jo: noting that there are guests today, it is good to notice the formality <describes the formalities>

Compliance and Redistribution PoC

Ben: we wanted to do an end-to-end demo
… From a consumer to a provider/provider-administrator
… we looked for something that has lots of cross-supply chain interactions
… and if we could replace those with web-based or M2M requests
… we also wanted to look at the end value of this, can we implement some kind of enforcement
… the final area we thought it could touch was the overhead of maintenance
… because there is a significant overhead
… and looking to automate those kinds of changes
… we picked out one particular use case: redistribution
… we know that causes certain amount of heartache
… looking to automating consent of redistribution
… <describes the sequence diagram of the use case>

Marko: need to share my screen
… <describes the proof of concept>

Mark: and you just sucked in this from DataBP via ODRL ?

Marko: yes

Caspar: we've got a fake client application
… the distribution component can understand the ODRL coming from the policy
… <log in to the web application and starts the demo>

Mark: what Caspar has done is initiating a request to the content source
… saying "can we distribute data to this 3rd party"
… at this stage we send a "deny"
… What could happen is that the user could go and create an approval
… <mark describes his part of the PoC in detail>

Caspar: you can see we are still in pending state, waiting for consent
… in the background the access has been pulled down and now it has been granted
… <describes what was downloaded and the access that was granted>

Ben: do you want to change the updating text?

Mark: it is very dramatic
… <goes and changes the entry in his part of the PoC>

Caspar: now that we re-request, we can see the text has been updated

Ben: would be worth showing other use cases ?

Caspar: <shows other use cases>

Ben: we are speaking about a deontic state
… <describes the content of the message>

Mark: maybe stating the obvious, but we have 3 systems in play, moving in real-time

Atiq: demo worked first time, brilliant !
… the real value is the workflow in the exchange between dataBP, deontic and Goldman Sacks
… which is quite an achievement
… <describes the concepts>
… and here we are building an API and how it might work
… the reason why it is important is the business processes and what are the boundaries of responsibilities
… <describes more of the concepts>
… the final part is maintenance/changes ... building that into the process will be a timesaver

Ben: we will pick up some issues later in the meeting, what kind of testing we will be doing moving forward
… <describes some of the steps forward>

Jo: fantastic and successful demo

Emily: <asks Caspar about the uplift in the applications>

Caspar: <describes the different components/processes/requirements for that uplift>

Emily: paraphrasing - there is some uplifting into the apps, and going forward there will be a minimum threshold of how that uplift will take place

Caspar: <describes>

Michelle: you were talking about real-time update in policies
… do you see this as going directly to the exchange?

Atiq: we still have to workout what the right model is
… <describes the gaps/components>

Ben: we are working in a temporal model: in some cases we will receive a message today "this change will happen in 6 months time" ... and the system will work then it is effective

Mark: you can see from the demo I chose a date, but that could have been in the future
… <describes how the PoC works>

Michelle: that is an interesting concept, that it could happen systematically

Atiq: if it reduce toil of the repetitive things we do

Laura: I had the same question as Michelle
… there maybe other requirements that they might put in

Michelle: if you look at exchanges now and that you can add a new subscription, this model could be very useful

Jo: thank you everybody
… you can add more questions via the email list

FISD Webinar in June

Ben: spoke to FISD on the topic of digital rights, my own sense is that the June timing is very good
… rather than talking about the standard itself, focus on what the next steps are
… whether people think a presentation in June is worth it

Jo: and demos

Ben: many demos !
… <continues talking about the elements of use for June's FISD Webinar>
… and if we can show some demos, that's the most potent message

Jo: anyone disagrees ?

Ilya: we should tailor the message carefully for the wider community that haven't been so involved

Resolution: We should pick up on the opportunity to present at FISD

Jo: I completely agree with that

Mark: a lot of the good work we've done has been defining test cases
… and put some of that and communicate it to the broader audience

Action: Ben to call first coordination meeting to prepare for FISD

Ben: commits to deliver feedback in 2 weeks

Interoperability Testing

Ben: one of the things that the work has emphasise is the importance of validation - validating policies, knowing what is mandatory and what doesn't have to be there. I think that as people start doing implementations and PoC's and pull that work and publish it as part of the report on the standard
… and another thing that is on a very early stages is the compliance API's that we are using in this PoC
… and it would be worth to address if the API is also a standard and where is its home
… the work we've done over the past few weeks have shown the importance of validation and compliance. Atiq do you have any thoughts?

Atiq: yes, we should find a way to keep the compliance and not just the ODRL component, the API works and behaves
… and have a clear standard and change management works around it
… so far it has been largely around desktop
… but we should do an analysis on that

Ben: Ilya what is your sense on this

Ilya: you are probably talking code and a definition
… <explains possible places to store the code>

Ben: should we keep collaboration in the test cases

Ilya: such an interesting question for this group
… there will be some interest in legacy organisations depending on the types of problems solved by this specification
… depending on whether or not those API's are wide reaching enough
… what maybe worth exposing is what data exchange is part of this API, so each group can work in what each part means for "them"
… there are many implications and we haven't specify and can run into rabbit holes

Jo: it strikes me that API is a specific technology - we should get the models and try to specify in the context of flows and when (flows) people will be interested in things

Atiq: on the subject of an API as a framework, should we form a breakout group?

Jo: I'll action you for that

Action: Atiq to arrange break out session on APIs

AOB

jo: Meeting reverts to 1100 US Eastern 1600 UK next time
… congrats again to demo-ers
… many thanks once again to Josh for scribing

--- meeting closed

Summary of action items

  1. Ben to call first coordination meeting to prepare for FISD
  2. Atiq to arrange break out session on APIs

Summary of resolutions

  1. Accept minutes of last meeting
  2. We should pick up on the opportunity to present at FISD
Minutes manually created (not a transcript), formatted by scribe.perl version 127 (Wed Dec 30 17:39:58 2020 UTC).

Diagnostics

Succeeded: s/Ben: trying to share my screen//

Succeeded: s/... we wanted to do an end-to-end demo/Ben: we wanted to do an end-to-end demo

Succeeded: s/From a consumer/... From a consumer

Succeeded: s/noticing that there/noting that there

Succeeded: s/What could happen is that the user could go and create an approval/... What could happen is that the user could go and create an approval/

Maybe present: Mark