Dear all, this paper has now surfaced
Möller, Steffen; Prescott, Stuart W.; Wirzenius, Lars; Reinholdtsen, Petter; Chapman, Brad; Prins, Pjotr; Soiland-Reyes, Stian; Klötzl, Fabian; Bagnacani, Andrea; Kalaš, Matúš; Tille, Andreas; Crusoe, Michael R. (2017): Robust cross-platform workflows: how technical and scientific communities collaborate to develop, test and share best practices for data analysis. Data Science and Engineering. https://doi.org/10.1007/s41019-017-0050-4 that I recall to have its roots in an idea exchange during a train ride from Saint Malo (where we had the 2015 Debian Med Sprint) to Rennes. The 2016 Sprint in Lyngby brought some flesh to it all, an ELIXIR Hackathon in Trondheim shaped it further, at this year's Sprint in Bucharest we then decided to write this up. Additional authors were picked from comments in threads on the Debian Devel and Debian Med mailing lists or that reached us in person at the ECCB or ISMB conferences and the BOSC and Codefest premeetings. An EU COST action (which is EU money to support travel among peers with similar or complementary interests) came up with that journal and publication costs for the Open Access publication were kindly taken from us. I tend to think that we arrived at something that is not too Debianish and not too much loaded with Bioinformatics. Thanks to a couple of gifted native speakers it is all very enjoyable to read, too. And we should also thank all those above referenced activities and their sponsor that repeatedly bring us together. Strategically, I do not know what the paper will be good for. It triggered a lot in me. For instance we have a dualism with the Conda/BioConda communities that Debian needs to learn from, I tend to think. Also, a thread on auto-built packages to bring closer to our distribution I started here https://lists.debian.org/debian-devel/2017/11/msg00120.html which came up while contributing to that paper. And we need more ideas to develop ourselves to remain attractive to the periphery to our distribution. Maybe we could have synchronised with Fink in a better way to attract Mac people. Or newer virtualisation techniques that are not clouds but mimic a Linux kernel on Windows (Windows Subsystem for Linux (newish) or http://www.colinux.org/ (old, worked nicely!)) should be promoted more. Something I will raise at the next Sprint is a request to think about a Debian Med paper that we should then cite with whatever biological research we perform with Debian Med (directly or containerised). To decide how we should present ourselves in there could reflect/direct what to show on a Debian Med booth on conferences. Cheers, Steffen
