Dear All,
Please find below my report on the D4Science PMB meeting with the project External Advisory Board (EAB). The meeting was held in Tirrenia (Pisa) on 26th March (full day meeting from 9:00 to 19:00)
The EAB members are:
- Christine Borgman (CB) Professor at UCLA (expert on scholarly communication, eScience, data, library automation)
- Edward Berghe (EB) Flanders Marine Data Centre (expert on data and information management for marine sciences)
- Fabrizio Gagliardi (FG) Microsoft Research
The following D4Science members were present:
- ERCIM: Jessica
- CNR: Donatella, Lino, Leonardo
- FAO: Marc, Joannes
- CERN: Pedro
- ESA: Luigi
- Athens: Yannis
- UNIBAS: Diego (replacing Heiko)
Due to the limited knowledge of the EAB members about D4Science and the complexity of explaining the real objective and advantage of D4Science in few hours the meeting was more a presentation of the project. Therefore most of the time was spent in presentations from D4Science people about the several areas of the project.
From the technology/software perspective there were very few comments/advices from the EAB. Form the perspective of the project vision they were also quite vague highlighting only the need to produce concrete results with our two user communities if we want to attract other communities. Essentially, most of their comments were about (1) the project sustainability and the need to guarantee support/funding for the gCube software and the infrastructure after the end of the project (2) the need to understand the requirements of the project user communities (3) to define exactly who are our real users, and (4) to provide a stable infrastructure. From the "dissemination/political" perceptive the meeting was productive since we disseminate the achievements of DILIGENT and the plans for D4Science to 3 different people from 3 different areas.
You can find the final version of my slides (about the project infrastructure) in EDMS:
- https://edms.cern.ch/file/902203/1/EAB_InfrastructureAndUsers.ppt
You can find below my minutes on the most relevant comments done by the EAB members.
PS: Another meeting about the future of Digital Libraries was organised by Donatella during today and tomorrow with the support of Microsoft. This meeting gathers many people from the DL field from EU/US plus representatives of funding agencies. You can read more about this event here: http://www.grl2020.net
Cheers,
Pedro
=== Intro ===
FG) What happens when the project finishes? Who will maintain the project and the infrastructure?
FG) Even if the infrastructure is guaranteed by the user communities, its important to guarantee that the tools are maintained.
FG) How can you guarantee to the user communities the continuity of gCube after the end of the project?
FG) Have you think how about archiving? preservation?
FG) The infrastructure projects are trying to map to the idea of the research network project of some years ago.
FG) It's important that the software is useful, that it must also be sexy.
CB) The life cycle between users and IT people is very different. Have you any plans to have a social scientistic to analyse this problem in D4S?
CB) It's very important the issue of who controls the data, specially in communities like the humanities.
CB) The vision of sharing is extremely different from community to community, even within the science domain.
FG) It's important to understand the bubget limitations of the funding agencies.
=== Infrastructure ===
no notes
=== gCube & gCore ===
FG) The underling technology (gCore) must be dynamic and as lightweight as possible.
FG) You should move the complexity towards the server, and provide extremely lightweight clients.
FG) From the 137 components composing gCube, how many are core components?
FG) From the non-core components, how many were developed outside the project?
FG) If more development is done... who will support even more components in the end of the project?
FG) You need to start now talking with the funding agencies for funding after D4S.
FG) The EU only "guarantees" funding to HW resources and network, even grid is not guaranteed.
FG) It is still too early for IT user communities to support project like gCube.
FG) From the industry perspective it is also too early to support gCube, the market is still too small.
FG) It's important to lobby at EC level bring the topic of data repositories even more relevant at bodies like ISTAG, ESFRI, and eIRG.
CB) There are new opportunities now starting in NSF for mid-term funding (5+5 years grants).
FG) It's important to have direct talks between EU and US funding agencies.
EB) It's important to have a reliable infrastructure to show and bring more communities in.
EB) How much of the system is really tested?
FG) How much of the gCore maintenance is paid by the project partners and by the EC?
FG) Research centres usually don't support internally the SW they develop, usually it's not part of their mandate.
=== FRM ===
FG) What is the profile of the user that wants to access the FRM data and FRM VREs?
FG) How wide will the future FRM VRE be?
CB) Are you interested in using Object Reuse Exchange (ORE)?
CB) How about provenance issues? Do you need traceability?
FG) How dynamic is the need to create new VREs? How dynamic is the data (data refresh only or data structure update)?
FG) The UN support for the need of tools like gCube could be extremely important
CB) One simple meeting for requirement gathering may not work, several days are needed.
CB) What is the mechanism you establish to collect the requirements?
CB) How do you plan to do the liaison between the real users and the development teams?
CB) Who is actually the "community""? who are the real people? Is it a "in-the-field scientist"?
CB) What is the mechanism to get the feedback from the data consumer? Or from the policy makers user?
CB) You should try to test as early and as possible with the most number of end users.
CB) It's important to bring new audience to the VO. Climate change is related, could this be exploited?
=== EM ===
no notes
=== Dissemination ===
FG) Dissemination must be managed together with the project management.
FG) Dissemination is many times seen as documentation, the next thing to do. This approach is wrong.
FG) The project portal must be as attractive as possible.
=== General ===
EB) A working example application is crucial for the success of the project.
CB) I haven't seen yet what is a D4S user. You should define better who are your real users.
There are minutes attached to this event.
Show them.
