Refeds


Subject Re: Work Items discussed during the REFEDs meeting
From Ingrid Melve <ingrid.melve@xxxxxxxxxx>
Date Mon, 14 Jun 2010 10:44:51 +0200

On 07.06.2010 14:21, Licia Florio wrote:
> Could you please provide your comments to the proposed list of topics by
> Friday at latest?

Apologies for extending Friday...

> a. Raising REFEDs profile
> --------------------------
> REFEDs flywheel, website, etc
> 
> b. Established points of contacts with dependable communications for all
> major R&E federations and VOs
> ---------------------------------------------------------------------
> The main challenge will be to keep this list up-to-date (the current
> REFEFs wiki has got already an extensive list, but mechanisms should be
> defined to ensure that the data on wiki are kept up-to-date).
> This is particular relevant for moving VOs related efforts forward.
> These include discussions with VO leadership, specific plans for
> rollouts and cutovers, and contacts with those federations whose
> countries house major research groups.

This is important, both for the higher eduation and research VOs and for
the federations.

> c. Working with publishers
> -----------------------------
> Andrew reported on the initial calls between a some of REFEDs
> participants and publishers. Andrew noted that the publishers when
> approached were very positive about the REFEDs. They felt that REFEDs is
> the only forum where their needs can be discussed in an international
> context.
> 
> As engaging with the publishers can result in discussing a variety of
> issues, Andrew proposed that he would, in consultation with the
> publishers, prepare a list of topics that would then be mapped to REFEDs
> roadmap. Some of the issues identified cover licences, usability, and so
> on.

I would like to see this activity as a more general "domesticating
applications for federations", but with the first deliverable centered
on communicating and understanding publisher's needs for integration of
federated information resources.

> d. User Interface
> ----------------------------
> This includes rough consensus on common language for use in privacy
> managers, work on harmonising user interfaces, consensus on common
> language for use in privacy managers.

Does this include the deep linking issues, or do they fall under e) ?
How do we relate to on-going work like UMA?
 http://kantarainitiative.org/confluence/display/uma/Home

> e. Discovery
> -------------
> Decide on a specific plan for eduId (note that eduID is just a temporary
> name, no agreement on the name has been reached) and foster the
> deployment. The discovery problem is being worked on in many different
> arenas within the Internet identity community, including ULX, OpenId,
> Infocard, etc.
> The liaison with the publishers could be beneficial for this.

There is strong overlap with d, since user interface is the key to good
discovery.

> g. Attributes
> ---------------
> Semantic of eduPerson: based on the discussion followed after Mikael's
> talk use-cases need to be collected and discussed.

We are all using eduPerson, and then some stuff.  If there is to be work
on semantic interoperability, we need first to state some ground rules
for federations for higher education and research:
 - use the eduPerson schema
 - quite a few federations use Schac in addition
 - most federations add other schema (in our case: eduOrg, eduOrgUnit,
norEduPerson, norEduOrg)
 - semantic interoperability today is provided as Best Effort, no
guarantees for the same definition
 - federations MUST publish their guidelines for attributes

When that is done, the discussion may start about the content, but let
us not forget in the REFEDS context to show the shared framework (and be
grateful to the guys who whipped up eduPerson for us all to enjoy and
share ;)

> e. Looking at ways to exchange verified metadata
> -------------------------------------------------
> Create a WG to define a short-term approach to providing metadata for
> shared community resources (e.g. the Refeds wiki, the spaces wiki, the
> Czech medical atlas, etc)
> 
> eduGAIN is not the answer to everything, other solutions should be
> looked at. Several options are already available; the output of this
> work item should be the reach consensus on which solution(s) to use.

Reach consensus or understanding mechanisms as the first step?

> f. LoA
> --------
> - waiting for the results coming from JISC study on deployability of
> both InCommon Silver and Kantara;
> 
> - next steps for federations to start deploying higher LoA


Ingrid