Date

28 Apr 2016

Attendees

Former user (Deleted)
Afsheen Afshar
David Schachter
Frank Tarsillo
Lawrence Miller
Mike Harmon
Johan Forsell
Johan Sandersson
Matthew Bastian
Hamish Brookerman
Doug Esanbock

Apologies

James Turck
Dominic Williams
Emmet Murphy

Discussion items

Time	Item	Who	Notes
5min	Gather / Welcome	Bruce Skingle	I will call the meeting to order promptly at 5 past the hour
15min	Collaboration Use Cases	All	Discussion of any further proposals for use cases / user stories and examination of scope for specific collaboration between WG members to drive towards a demonstration of the use for Complex Financial Objects in a heterogeneous application environment.
10min	Rendering pipeline proposal	Bruce Skingle	Please see the Rendering Pipeline page, I will give a brief overview of the proposal.
20min	Discussion	All	I would like to sped the bulk of the meeting discussing the rendering pipeline proposal and in particular the idea of enrichment applications. Is the overly complex or does it give us the flexibility we need? Is it insane to imagine that application providers would produce apps which will cooperate in this way?
5min	Any other business
5 min	Close		I will aim to close the meeting promptly at 5 to the hour.

Minutes

Bruce Skingle briefly presented the Rendering Pipeline proposal and there was discussion around multiple aspects of that.

It was agreed that he overall approach seems to be appropriate and that we should proceed with work on the basis of the ideas in that page. There were several areas of concern, and suggestions for alterations as follows:

Enrichment applications should not be able to alter any of the content of the original message payload, all enrichment should be additive. A change to the PresentationML element will be needed to accommodate this.
There is a concern that the performance impact of an additional enrichment step, with a possible round trip to a server in an unknown location. We should ensure that any proof of concept implementation explores the performance and UX implications fully.
There is a concern that this scheme may be complex to administer / configure so we need to look carefully at the UX of the administration.
It may be desirable to allow multiple applications to render the same object, in which case it may also be desirable to be able to influence what each app renders.
It was suggested that standard XML processing tools such as XSLT might be used to perform the enrichment phase.

Action items

Doug Esanbock and Johan Sandersson to chat about whether Dow Jones and Factset can work towards a demonstration of a user case where apps from their two organizations interoperate in a conversation.