Comments by and on behalf of PLAN-E

03-08-2018

Comments by PLAN-E on the recommendations in “Prompting an EOSC in Practice”.

Patrick. J.C. Aerts, Chairman PLAN-E

1. PLAN-E reconfirms it’s full support for the realization of the EOSC, but we have some serious concerns in the proposed implementations, regarding matters insufficiently or inadequately addressed.
2. We support most of the comments made by Ron Dekker from CESSDA concerning the nature of the document, intended to accelerate the implementation process.
3. The two elements of concern by PLAN-E are:
   • Inadequate researcher involvement
   • Insufficient attention for the importance and role of software.
4. Concerning the user involvement:
   • The EOSC should be implemented to serve all researchers from all research domains. Also the many research domains that so far did not require national or international ICT or Data infrastructures to do their research competitively. Data from those domains are not less important, nor for that discipline nor for other disciplines. Lacking data from those domains may lead to scientific bias or unbalance if only data are going to be used or addressed of the disciplines that are close to the present core or EOSC activities.
   • Many of those disciplines are not able to articulate the translation of their ICT-or Data needs into practical procedures or physical infrastructures.
   • Rather the eScience communities, academic researchers that do have practical experience in translating scientific (data and ICT) requirements for those domains through the support that they provide should be given a visible role in the stakeholders forum if only for this purpose.
5. Concerning the importance and the role of Software:
   • It should be recognized that digital data have no meaning but through software.
   • Software should not only be open, as stated in the recommendations, but should also be
     •  
     •  
     •  
     • Energy consumption aware.
   • For data to remain FAIR as long as the data are relevant, the software required to produce, read, interpret or handle the data must be maintainable and executable on relevant platforms.
   • Software Sustainability requires software to be designed and written according to minimum requirements with the goal to render the maintenance possible at minimum cost and technical efforts.
   • For proper referencing software in scientific publications and to mitigate the risks of proliferating erroneous published results in case of software errors, software needs to be properly version controlled. In terms of priorities this requirement seems from a practical point of view to be no less important than the software being open.
   • The same applies to documentation. In fact, software should be FAIR with a software specific interpretation of the FAIR principles.
   • Nowadays the energy consumption of ICT, including data handling, transport etc. is a real socially relevant topic which is not mentioned in the EOSC recommendations. We strongly advise to take the energy consumption of data processing and processes into account by addressing the energy consumption as part of computer coding conduct.