I attended this years ECMTB 2014. Apart from many interesting sessions on mathematical modeling, I attended this minisymposia on software for multicellular simulations. All the speakers had done a great job developing software for the application community, but the impression was that they managed to do this despite being productive methods researchers, not as a consequence of it. We have had a similar experience with the software we develop, and it raises a very relevant question about the feasibility of developing scientific software. Arguably, professionally maintained software based on the latest methods research is essential in order to disseminate the results in the application areas, but how to secure funding to enable a long term maintenance and development of software? In CS oriented fields such as Scientific Computing, research infrastructure usually implies hardware and clusters. Arguably, good scientific open source software should also be viewed upon as important research infrastructure, and some infrastructure funding could then be used to increase the incentive to maintain software.