Editorial: Stable funding for open source software?

Oliver Goldsmith said, ‘Where wealth accumulates, men decay’: this should make many programmers of free open source software some of the healthiest people on the planet. One can only aspire to be of independent means and become what the Victorians might have called ‘A gentleman bioinformatician’; after all, if you’re going to have delusions you may as well have really satisfying ones. The reality of existing mechanisms for funding open source software projects has been brought into sharp focus by the continuing uncertainty over the future of the EMBOSS package. EMBOSS has always followed the principle that scientific software, like any other true scientific venture, should be open to review by the authors’ peers. This implies that the source for such software should be downloadable at no cost or, at least, at a cost that does not prejudice any real or potential scientist. It also begs the question of who should bear the cost of continued core development, given that any large community project requires a stable team to coordinate and maintain it. A survey by the EMBOSS team has canvassed the views of the user, training and developer community for that package. That survey has, so far, elicited over 900 replies; a staggering response even given that there have been over 21,000 downloads of EMBOSS software since its first release. The majority of the responses came from Europe (60 per cent) and the Americas (16.5 per cent). Over 90 per cent of the replies were from academia, only 8.6 per cent being from the commercial sector. Taking into account multi-category usage, the application users made up 83 per cent of the responders, 28 per cent were educators/trainers and 18 per cent were developers. Only 4.2 per cent said they did not find EMBOSS to be a valuable package. Over 86 per cent of these scientists maintain that EMBOSS should remain free of charge at source and, of the remainder, only 0.7 per cent disagreed. On the question of who should bear the cost, 73 per cent supported the public funding model, 1 per cent believed that such a package should be commercially funded and 15 per cent suggested a mixture of both. EMBOSS is a member of that family of packages which have utility in a broad spectrum of biological research projects and whose use is therefore widespread. Is it unreasonable of the users of such software to expect the public, via government funding, to be altruistic in its support? Are they just motivated by greed or by being victims of the vagaries of scientific funding themselves? Or are they, through the survey, hoping to foster enlightened self-interest within the funding bodies? We didn’t ask! What has been encouraging about this exercise is the strong feeling within the community that the open software model is the only one that truly represents the scientific method. Unlike some other sciences, and for many reasons, biology has provided a fertile environment for the growth of free open software. Even if the programmers aren’t healthy, the science still is.