Open Source vs. Free

I am working with a group of developers who have built a Laboratory Information Management System (LIMS) based on the Open Source principle.  I spent a lot of time researching the availability of inexpensive, flexible LIMS that we might deploy as part of the project in Liberia.  There were about 30 variants of Open Source LIMS and about half that many commercial versions including some SAAS.

Recently there’s been a blog that was brought to my attention about the ‘cost’ of an Open Source solution.  The blog was focused on LIMS and so it was worth reading.  The author came to a similar conclusion that I did about open source LIMS in general, in that there have been very few successes.  Of the 30 or so I looked at most had been abandoned after the grad student left the institution where they worked or the organization’s funding stream evaporated or the single author decided to pull the plug and abandon the effort.  There are a few that continue including the one I selected for our project in Liberia, Bika.Lims and Bika.Health.   I’ll cover that selection process in some other posting.

What I wanted to focus on here was the disconnect between ‘free software’ and an ‘open source based solution’ which I think the blog author missed.  The blog author’s point was that the bulk of the cost of a system implementation of something as complex as a LIMS isn’t in the cost of the software but rather in the cost of implementation.  In general, the author got that part correct.  Implementation of any system involves a lot of challenges, a lot of people, and a lot of costs.   The actual cost of the software license for a piece of commercial software may be a fraction of the overall cost of implementation when you factor all the labor for training, physical implementation, support networks, etc., etc.  The point is well taken but may miss one of the really important issues around Open Source, namely flexibility.

Let’s take the work I am doing in Liberia right now.  We have very difficult circumstances for our deployment.  We have a very tight budget, difficult environment (poor power, lack of internet, no air conditioning, low computer literacy rates, etc., etc….), lack of continuation funding, and a challenging model to try to fit the LIMS into.  Liberia is recovering from two civil wars and the horrendous Ebola outbreak.  The outbreak exposed the poor health care system within the country and was a wake-up call to the rest of the world at what a risk that puts everyone in.

The LIMS that was needed had to be able to operate in this environment.  It also had to be tailor-able, it had to be able to be adapted to the specific needs of the health care system of Liberia. Most importantly, it had to be sustainable.  When I evaluated commercial applications for LIMS what I found were very sophisticated systems, a fairly high startup cost, a high annual support cost, and an inflexible system.   Given that both an ‘Open Source’ and a ‘Commercial Licensed’ software package would have the same implementation costs (i.e. the training, etc. I mentioned earlier), the cost associated with licensing and annual updates/support was a significant factor in my decision making.  Why?  Because I knew we had to customize the software to make it fit for Liberia.  Even if the commercial system vendor would agree to do the customization, that cost would be above the cost of licensing.   What if instead, I could take those costs for software (LIMS, database, operating systems, etc.) and instead apply that budget line item to customization?   To go even further, what if instead of paying licensing fees I instead took some of that money to train local Liberian’s on how to customize the LIMS and become contributors to the Open Source project itself?

In fact that is what we are doing.  We selected the Open Source Bika.Health platform not just for its functionality, its zero cost for licensing, and its developer network but also because it is Open Source, meaning it is growing and accepts contributions to the software from developers around the world trying to solve problems.  Additionally being Open Source meant we were allowed, actually encouraged to experiment with the software, with the platform it runs on, with pretty much everything to make it work for Liberia.

Yes, we are incurring some additional costs in developing our own implementation for Liberia.  Yes, we do not have a corporation behind the software development adding new features (which we probably won’t use in Liberia anyway) and yes we have to think about support but we have to think about support anyway as the funding for running the LIMS is based on a grant and that grant will expire.  So focusing on that eventuality it became clear that if we can train Liberian staff to support and grow the system, to adapt and contribute to the project, then they may even become the regional experts on the project in West Africa and support others in implementation.

Yes, implementing a software system is not inexpensive, easy, or free and so you must be careful of what your actual objectives are.  Are they to deploy and run?  Are they to provide a quick win and then leave them in the lurch or are they to provide a tailored system that fits the current needs and the capabilities to grow that system themselves in the future.  If that’s the objective, then an Open Source solution might be exactly what is needed.