Trying to run KohGPI but needed to find SPRANNLIB which is not available for what I can see – but found GNU-i-fied version in clibs directory of http://sourceforge.net/projects/rlabplus/files/rlab/

This implementation also depends on the GNU Scientific Library (GSL) which you’ll need to build and install beforehand.

Then to build SPRANNLIB  download and untar the rlab archive. Go to ‘clibs/sprannlib/src’ and type make. If things are in order it will make a library in ../lib called libsprann.a spr_$ARCH.a where $ARCH is your architecture (x86_64, etc).

The .a files can be deployed in your normal place for manual (non-RPM/DEB) installations (/usr/local/pkg/sprannlib for me with symlinks to the libraries in /usr/local/lib). Then go back and build kohgpi but you may need to update the Makefile so that the library paths point to your sprannlib installation location.

Open data

Neil points out that the Science Commons has released a new standard for implementing open access data and also references Deepak’s nice summary of the important points. A standard like this will be an important step towards increasingly more open science and how data is made accessible.

The ISMB open data and open source BoF which never got around to discussing aspects of open data – it was mostly a discussion of different points of view on how obligated one should be to make code open source when it is necessary for a published manuscript. One thing I heard discussed was the idea that the software necessary to access the data needs to be made available, we don’t want to end up with unreadable data because the software doesn’t exist anymore.