Effective scientific databasing and data sharing

I am at the TRY workshop in Leipzig, Germany, hosted by iDiv, a new, well-funded German centre for integration and synthesis of biodiversity and ecological data.  TRY is a 5-6 year old project with has tried to bring together a database )or database of databases) of information about plant traits worldwide. Incidentally, "TRY" is not an acronym for anything, but simply indicates the project would TRY a very ambitious task.

The workshop has plenty of interesting science talks and stimulating breakout groups. An interesting underlying theme has been about how to encourage and make effective data sharing, where individual research projects contribute effectively to the global bank of data and the greater scientific good. To be "effective" means the data have to be fairly easily accessible, and to be sufficiently organized and well-documented to be usable.

There seem to be  two main blocks. One is convincing researchers that it worth their time to database their data well. Projects are often required to do this by many funding agencies, but it can still be hard to invest sufficient time to do this well. As well as the moral argument for making publicly funded data available for the collective good, there are many arguments for how databasing can benefit the individual researcher. Many of these arguments are summarized in this blog post by Daniel Falster (applicable to more than TRY):
 http://dfalster.github.io/blog/2013/08/23/making-a-case-for-a-fully-open-trait-database/

The second block is simply prioritizing databasing and sharing when there are so many demands on time and budgets. Most researchers (most certainly including me!) could benefit from thinking about how they will manage their own data right from the start of a project, rather than designing this on the hoof and right and the end of the project. This is something that probably research students are trained in and thought to think about right at the start of their training, so it becomes a habitual way of doing science. 

This paper gives some useful guidance:

White et al. (2013) Nine simple ways to make it easier to (re)use your data.PeerJ PrePrints 1:e7v2 http://dx.doi.org/10.7287/peerj.preprints.7v2