100 likes | 181 Views
Policy & Organisation. eHumanities address Software and Tools Sustainability KNAW Trippenhuis , Amsterdam, 10 April 2013. Questions. What services are to be sustained? Who (which organizations(s), groups) are or should be responsible for what aspects of the matter?
E N D
Policy & Organisation eHumanities address Software and Tools Sustainability KNAW Trippenhuis, Amsterdam, 10 April 2013
Questions • What services are to be sustained? • Who (which organizations(s), groups) are or should be responsible for what aspects of the matter? • What consequences can be drawn from that for the future (model of) funding of maintenance, accessibility, etc.? • Is a separate (sub-)organization required; should software/tools be collected actively or passively? • (how to go about the funding for sustainability of software that resulted from past projects –for which nothing was organized yet-)? • Are there in policy/organizational solutions ehumanities’ specifics?
Jan Odijk: What services are to be sustained? • speaks from experience in CLARIN/CLARIAH. • different classes of tools and services, reflecting the research lifecycle: • software to support data creation • processes to enrich data • store data and generate metadata • annotation tools • analysis tools • visualisation tools • make the different classes of tools interoperable • it only appears afterwards whether a tool is successful or not • Main goal: re-use of the tools
Neil: making tools more generic is not a viable strategy, unless the leap is small • Jan: nevertheless we see that some linguistic tools are being used in historical research • Jacob: is it feasible to rationalise tools for eHumanities? • Jan: yes (example – Stevin project, followed by CLARIN) • Neil: is it possible to identify the best tools that are used by the community? • Jan: yes – for some tasks (quality/flexible/supporting team/modifiable) • Demetrius: how can you make the user community bigger? • Patrick: who are your clients? individual researchers or bigger (b2b) projects? • Jaap: if you want to be really innovative, the tool is usually not mainstream and very much dedicated • Jan: we need first of all an overview (registry) of the tools that are around. • Jaap: and this is of course dynamic. And we need criteria to decide whether to maintain the tool or not
Marc Dupuis: Who (which organizations(s), groups) are or should be responsible for what aspects of the matter? • Many things are moving in the area of (open access to) publications and data, and different groups are claiming a role • Does not the same apply to software? • Is it the funding organisations that should take care of this, or is it the research organizations themselves? • Jan: distinguish different tasks • Making it visible/accessible: stable centre • Bugs, improvements: a dedicated community • Neil: you need to invert the question. The research community should be in the center; how to organise the decision? From that comes the next step about funding. • Patrick/Jan: discuss about requirements about software documentation • Marco: maintaining the outcome of the Catch+ programme is an issue (report). You cannot evade the question about who.
Patricia: most tools of Catch+ are maintained by the (heritage) organizations where they were developed for their own use. • Peter: is enforcing maintenance feasible? • Neil: it is, but… If software is successful, it takes about 7 years before it is generally taken up. Key time for follow-up is 2-4 years after ending the project.
Peter Doorn: Do we need a Dutch Institute for Software Sustainability? • An existing organisationor a joint effort? • community driven (CLARIN, Catch)? • across communities and disciplines? • Big Science is more self-supporting than the humanities! • Candidates? • How to pay? • Which functions? • expertise • software preservation / maintenance / exploitation / running the service? • setting criteria for sustainability: a “Seal of Approval for Sustainable Software”? Software Sustainability Plan? • active or passive acquisition of tools?
Neil: some striking parallels with thoughts in the UK (eg. 5 stars of software quality; software management plan) • Jan: CLARIN has CLARIN centers • Peter: but many of them have trouble with maintaining externally developed software for free • Patrick: a coalition between NLeSC and DANS, possibly extended with others (CLARI-N/-AH, DEN, Catch, 3TU.DC?) seems very well possible • Jaap: CLARIN and Catch are front runners, followed by other humanities • Proceeding with quality guidelines and management plans: international collaboration would strengthen it. • Neil: you could also use quality criteria to decide which software to support and which not (6 criteria: research impact, timeliness, enthusiasm) • Jan: It helps to have a core set of very successful, widely used tools that everyone wants to be compatible with
Who pays • Jan: CLARIN centers get some financial stimulus for preparing to curate data and tools according to CLARIN standards