Scientific data processing will increasingly often occur in a distributed, federated landscape of repositories and services providers as described by several working and interest groups working and large scale data/research infrastructures. These data and service providers need to offer their services across national and discipline borders to all a variety of researchers to tackle not only the grand challenges, but also the many small research questions. To come to an open and flexible landscape of services providers we need to build trust mechanisms of various sorts. In addition we can assume that within a few years highly automated methods will be used to aggregate virtual collections from different data providers, to create and execute distributed workflows including services from different service providers. In such distributed scenarios it is important to improve trust, to support machine-readability at all levels and to allow for all kinds of checks reporting on the state of the "system as a whole " for example to detect and report errors. In particular in the discussions about data fabric it became obvious that a registry of "trusted" repositories based on common specifications seems to be essential for a seamless functioning of such a distributed system. Large infrastructures based on machine processing such as EUDAT and ComputeGrids are using such a registry, other infrastructures such as EPOS and ELIXIR are working on a specification for such registries and others will start as well. Also in the world of repositories for published data registries are being defined with the same to task to indicate to users which registries are subject to quality checks. There may be more initiatives that are working on such a repository registry. Therefore, it is now a good moment to interact between all these initiatives •to discuss whether we need such registries and for which functions and •if a need is being indicated, to discuss steps to come to agreed specifications to increase worldwide interoperability. This BoF session will have two parts: (1) First we will give all initiatives that are working with such registries a short slot to present their work (but only if documentation has been made available beforehand) and (2) Second we will have an open discussion where experts are motivated to make comments of all sorts. The aim of this BoF is to find out whether a concrete action such as a Working Group should be started and if so what it should address. The BoF request is supported by experts from a number of WGs and IGs and from EUDAT and Grid experts. It also includes the experts of re3data who started a registry for trusted repositories in the realm of published data.
1. Welcome and introduction (max 10 minutes)
2. Short presentations of existing registries that presented documentation beforehand (max 35 minutes)
3. Open BoF like discussion (45 minutes)
Requested by: Keith Jeffery, Gary Berg-Cross, Larry Lannom, Daan Broeder, Raphael Ritz, Timothy Dilauro, Tobias Weigel, Rainer Stotzka, Rebecca Koskela, Jane Greenberg, Giridhar Manepalli, Tiziana Ferrari, Heinz Pampel, Beth Plale, Giuseppe Fiameni, Mark van de Sanden, Johannes Reetz, Damien Lecarpentier, Yunqiang ZHU, Peter Wittenburg, Herman Stehouwer
Contact Person: Peter Wittenburg