How should academic software-projects be funded?
Academic software-projects can be funded, in three basic ways:
- Through (large) grants, such as those by the European Union.
- By industry, as a commercial project.
- As side-projects by individual researchers or volunteers.
All three seem to bring their own unique problems and benefits.
Projects funded by (large) grants, while obviously gathering large amounts of resources, tend to be only credited based on publications, not working software. The grants are research-grants after all. And therefore not much working software is produced. So far I have not heard of many, if any, software-projects funded with sizable research-grants which were successful in producing usable software. That is, other than digital editions, or projects that started out as side-projects and had some momentum before the grants arrived.
Projects funded by industry, on the contrary, go in the opposite direction. They tend to focus on the common denominators of large audiences. Still, quite a few of them are both useful for academics and successful, such as Google Scholar and Mendeley. Their strongest point is that they are valued based on the number of users they serve (through advertising or subscriptions). The problem is, however, that, partially for this same reason, there wouldn't be commercial projects any time soon for smaller groups, such as philosophers.
And then there are side-projects by individual researchers. Most of them are honestly trying to create something which is actually useful for the researchers whose work they want to facilitate. Examples of successful projects are PhilPapers, and various tools for computer-linguistics. Nevertheless such projects have some draw-backs.
- Side-projects are side-projects, and thus cannot gather as many resources (time, money) as the other two kinds. Also the resources that are devoted to the project, often are payed by the individuals involved (less time to work for pay, or less pay).
- Secondly, side-projects are often not taken as seriously as the first two kinds. Even though, given the fact that they can achieve things that the other two kinds cannot, there seems to be a legitimate place for them.
All types of project have their use, and research is important in itself. Nevertheless both commercial and side-projects seem most likely to be the ones that produce useful software.
An interesting idea might be combining the advantages of the three kinds: Funding smaller projects by researchers for researchers, and crediting them, not based on the number of publications they generate, but on the number of academic users they come to serve.