The acceptance of the Research Data Alliance (RDA) is growing and discussion what role the community sees for RDA, what they expect and what their need was very timely. This discussion took place on the third day of the 2nd RDA Plenary in Washington on 16-18 September 2013.

The stage was setup by a number of questions that have been collected during the first two days. The most popular questions were (including rank by collected votes):
1. What is the output of RDA? Standards? Best practices? Discussions? Certification?  (voted by 86)
2. Changing academic culture to recognise data citation and other non-taditional outputs (voted by 74)
3. Interoperability of similar WGs in different organisations - (voted by 56)
a) Interoperability testbeds (voted by 38)
b) Registries of resources and best practices to prepare for interoperability (voted by 36)
4. How can we cultivate tomorrow's data leaders through RDA? Should there be an early /mid-career BoF focusing on something like innovative approaches? (voted by 39)
a) Governance for forthcoming RDA students (voted by 9)

Here we recollect and look in more details at the discussion on the first two topics.

1. RDA output: Standards, BCP, White papers, position papers or just building community
The RDA represents wide community of researchers and organizations producing data, managing data and offering data services. Data and data related technologies are becoming instrumental for research. How to benefit from joint experience in different scientific domains, share and preserve collected data in each research domain and allow their re-use and integration - are the main questions that drives research communities to join and cooperate around RDA.
Few comments were received about possible RDA role to produce standards and other documents such as Best Current Practices (BCP), White Papers, etc.
Standards and BCP documents would be important output of RDA but standardization process requires a special procedure of the documents preparation and reviewing. People familiar with the process at the different Standard Developing Organisations (SDO) mentioned the IETF (Internet Engineering Task Force) standardization procedure as a good example to follow. In few words it is expressed as “Rough consensus and running code” (see Tao of IETF http://www.ietf.org/tao.html). OGF (Open Grid Forum) provides another example of organization that served research and industry community more that decade and use similar to IETF standardisation process.
Some comments mentioned that standardization process is generally slow and as for current fast technology development and changing research priorities the RDA community a new more dynamic ways need to be found for documenting and promoting common practices and effective solutions in the area of research data compatibility, exchange and preservation. This could be white papers, position papers, RDA reports.
It was clear that such standardization procedures/activities as at IETF or OGF are not a native practice for the RDA community and majority of the attendees who are not originally technology researches and/or Computer Science specialists. So, opinion was expressed that a solution could be looked in, for example, closer cooperation with existing and well established standardization bodies such as IETF, IEEE or ISO. In such cooperation, RDA Working Groups and Interest Groups could propose results of their activity as an input to those standardization bodies. Such practice exists for many national standardisation bodies or industry associations to contribute their BCP, technical reports to IETF, IEEE, ISO or ITU-T. On the other hand, RDA could evaluate and possibly recommend to its community the use of some of the international standards.
Standardisation is an important factor of the new technology development and RDA earlier or later need to define its position regarding standardization activity or liaison with existing SDO for the benefits of its own community.
Ignoring standardization, BCP and other type of outcome that would facilitate technology development will also decrease expected benefits from tighter cooperation between modern science and industry.It is clear that discussion about output of RDA is just started and it will continue at the following RDA Plenary meetings.

2. Building new academic culture
However, majority of comments were related to RDA role in building communities and providing a place for community focused discussions. This was regarded as an important role and an obvious result of the first year of the RDA activity, since 2012 autumn meeting in Washington. But this is rather a first step the community when it growths and crystalizes will look for wider and stronger impact.
One of the main tasks in building RDA community would be promoting a new academic practice, in particular in data citation and recognizing data collection or production as a scientific contribution. Experimental of research data is a basis for any research and in modern science their production requires significant efforts from researches. It would be beneficial if data collected in ne project or by one researcher could be available for other researchers to learn, re-use and do additional research. However, currently there are not much incentives for individual researchers to share their data.
Some commenters feel that introducing data citation and recognizing research data as an academic contribution would help wider data sharing. RDA can be such organization that recognize importance of data publication and citation and promote new academic practice.
Endorsement of PhD research would be another role for RDA by providing a place where new research topics could be presented and discussed. How to implement this if RDA is not hosting or intending to host workshops is not clear.