[Leonardo Candela]

Dear IG members,
first of all I thank you for the useful and interesting discussions stemming from my post asking for authoritative definitions for terms including VREs/VLabs/SGs.
My intent was not to stress the differences among these terms per-se (if any) and to promote their separation. Rather, I was (and still I’m) willing to figure out whether in the various definitions there are distinguishing features worth highlighting. To me all of them have been designed and proposed to provide their users / designated communities with added-value services aiming at offering seamless access and use of otherwise scattered resources (datasets, services, software, computing, etc.). Scattered resources imply to deal with heterogeneity of formats, protocols, technologies, terminologies, etc. There is a very limited / superficial agreement of what the “added-value” services are expected to be (probably because of the diversity of needs characterising the designated communities and application scenarios). 
My request was related to a forthcoming Workshop “Virtual Research Environments: What services should they offer and what e-infrastructures should provide to support them”, 27-28 February 2018, CNR  Pisa 
During such workshop we will discuss some initiatives developing VREs/VLabs/SGs with the goal to reply to the following questions:

What are the distinguishing features (functional / non functional) contributing to make a “system” a VRE?
What are the mandatory functionalities that Virtual Research Environments should offer? What are the desired functionalities that Virtual Research Environments should offer?
What services (and how) “service providers” (including Research Infrastructures) should offer to facilitate the development of VREs?

We are willing to distill the discussions and propose a research agenda (I know it is challenging!).
It will be great if the largest possible group of IG members will contribute to this forthcoming discussion by replying to this post. I’ll take care of passing any message / contribution to the workshop participants on your behalf (with the help of other IG members physically attending the workshop).  
Thanks in advance for any contribution you will propose.

Discussion

[Keith Jeffery]

Leo –
As you will have seen from the discussion this is a complex subject with large scope and various views.
The unifying factor is that a VRE/SG/VL is intended to assist the researcher in their work.
As you know I distinguish VRE from Research Infrastructure (which may have already something looking like a SG) from e-Infrastructure (i.e. computing or sensor resources typically (for computing) PaaS multicloud.
There are two major models:
1. VRE is ‘above and across’ RIs so providing a homogeneous access over heterogeneous RIs. This implies the VRE has a homogeneous superset catalog of the RIs or accesses the catalogs of the RI and ‘on the fly’ makes a homogeneous superset.
2. VRE is ‘within’ each RI. This is fine for that RI since it is efficient. However, for access to other RIs either their catalogs have to be imported and converted to the syntax and semantics of the original RI (this is the [n(n-1] problem) or each RI has to present its catalog in local syntax / semantics and also canonical homogeneous syntax/semantics. There are latency problems and also this makes it much more difficult to compose workflows that use assets from >1 RI and deploy onto >1 platform – if the catalog is homogeneous and in one place the optimisations can be done effectively.
In either case the VRE is distinguished from the RI because of its provision of interoperability and the functionalities provided for the researcher (or educator / learner or manager) – see below.
For me there are many functionalities that should be provided in a VRE.
Single sign on / user registration (possibly from pre-registration at a RI)
Catalog (virtual or physical) of available assets (with associated curation and provenance and metadata concerning rights and restrictions)
Catalog data management functions
Workflow composition and deployment (across sensor networks and multi-platform computing facilities)
Workflow implies execution of (parts of) the workflow e.g. data management, analytics, simulation, visualisation on appropriate RI or e-I platforms optimised for cost/performance against specified non-functional requirements
Access to bibliographic data (e.g. OpenAIRE) and if necessary more detailed sources
Access to research management functions (tracking funding, proposals projects, maintenance of CV, personal bibliography web pages)
Access to office functions (equivalent to the Microsoft office range of software)
Access to administrative functions of the user’s organisation (time management, travel, finance, human resources…)
In general the upper group of functionalities are provided by a portal or SG; the lower set of a full VRE. The upper set can be composed together the end-user to form a VL.
Ideally the workflow management of the upper set of functionalities should be neutral – i.e. provide a generalised workflow description (e.g. CSWL, WSDL, WADL) and then generate a workflow for the target workflow system and for the target platform(s). You will note all this implies that the application (workflow) processing is done at one or more RIs and e-Is. The VRE just provides the homogeneous catalog and workflow composition/deployment.
The development of VREs – in my opinion – implies
1. Continuous dialogue with the various user groups to understand their evolving (and usually more ambitious) requirements;
2. Tracking the roadmap of each RI to understand their future intent;
3. Tracking e-I (platform) offerings and adjusting the catalog to allow accurate workflow deployment
Clearly the RI and e-Is can assist greatly in this.
Hope that helps
Best
Keith
——————————————————————————–
Keith G Jeffery Consultants
Prof Keith G Jeffery
E: ***@***.***
T: +44 7768 446088
S: keithgjeffery
———————————————————————————————————————————-
The contents of this email are sent in confidence for the use of the
intended recipient only. If you are not one of the intended
recipients do not take action on it or show it to anyone else, but
return this email to the sender and delete your copy of it.
———————————————————————————————————————————-
– Show quoted text -From: leonardo.candela=***@***.***-groups.org [mailto:***@***.***-groups.org] On Behalf Of leonardo.candela
Sent: 25 February 2018 07:40
To: Virtual Research Environment IG (VRE-IG)
Subject: [vre_ig] Call for contributions: RDA workshop “Virtual Research Environments: What services should they offer and what e-infrastructures should provide to support them”, 27-28 February
Dear IG members,
first of all I thank you for the useful and interesting discussions stemming from my post asking for authoritative definitions for terms including VREs/VLabs/SGs.
My intent was not to stress the differences among these terms per-se (if any) and to promote their separation. Rather, I was (and still I’m) willing to figure out whether in the various definitions there are distinguishing features worth highlighting. To me all of them have been designed and proposed to provide their users / designated communities with added-value services aiming at offering seamless access and use of otherwise scattered resources (datasets, services, software, computing, etc.). Scattered resources imply to deal with heterogeneity of formats, protocols, technologies, terminologies, etc. There is a very limited / superficial agreement of what the “added-value” services are expected to be (probably because of the diversity of needs characterising the designated communities and application scenarios).
My request was related to a forthcoming Workshop “Virtual Research Environments: What services should they offer and what e-infrastructures should provide to support them”, 27-28 February 2018, CNR Pisa
During such workshop we will discuss some initiatives developing VREs/VLabs/SGs with the goal to reply to the following questions:
* What are the distinguishing features (functional / non functional) contributing to make a “system” a VRE?
* What are the mandatory functionalities that Virtual Research Environments should offer? What are the desired functionalities that Virtual Research Environments should offer?
* What services (and how) “service providers” (including Research Infrastructures) should offer to facilitate the development of VREs?
We are willing to distill the discussions and propose a research agenda (I know it is challenging!).
It will be great if the largest possible group of IG members will contribute to this forthcoming discussion by replying to this post. I’ll take care of passing any message / contribution to the workshop participants on your behalf (with the help of other IG members physically attending the workshop).
Thanks in advance for any contribution you will propose.
—
Full post: https://www.rd-alliance.org/group/virtual-research-environment-ig-vre-ig…
Manage my subscriptions: https://www.rd-alliance.org/mailinglist
Stop emails for this post: https://www.rd-alliance.org/mailinglist/unsubscribe/58982

[Author]

Replies