Proprietary software is not free 2005 February 22
Recently I have been working on a couple of article and conference presentation ideas that involve Free/Open Source Software (I’ll just use OSS for brevity for the rest of this article). During my research I have ran across a number of articles, some written by University administrators , that point out the OSS is not free of costs once you factor in staff time to test and insatll it.
In order to demonstrate some added costs involved with OSS, Gregory A. Jackson (2004), the Vice President and Chief Information Officer at the University of Chicago, writes that “To use open-source software (rather then the executables accompanied by the source), one must compile, test, distribute, and install it.” Lets examine this statement…
The first thing that jumps out is the part in parentheses. Just because you use a program that is available as OSS does not necessarily mean you need to compile it. While I have compiled my share of programs, I have used hundreds, if not thousands, of OSS programs that I have never compiled. Some of them are in wide use around the world, such as Sendmail, Samba, KDE, Gnome, Thunderbird, Firefox, Open Office, BIND, and many, many more. Even some of the ones that I have compiled to get different functionality, such as Apache, have taken me a much shorter time to compile then I would have been on hold to some mega-corporation who would have just laughed at me once I finally got the hold of a real person and I told them that I want a different option built into their proprietary program. But I digress…. Back to the point, I’m not saying that if you have special needs you won’t want to (or even have to) compile something from time to time, but in many, if not most, scenarios it is not the case, and the assumption that you need to is ridiculous. As already described, the next part of this statement is often not true. You do not necessarily need to “compile it” just because it is OSS. Incidentally, in some cases you may need to compile proprietary software as well.
Lets look at the next part of Jackson’s assertion, testing. Of course you should test a new major install of OSS, but you also have to test proprietary software as well. This is a wash. Except that with OSS you might not have already shelled out a small fortune on a program that doesn’t do what you want.
It probably isn’t even worth discussing, but when was the last time you bought a shrink-wrap package of software that you didn’t need to distribute or install? I know that whenever I get a new version of a proprietary software package it needs to be distributed and installed. If you know of some magic way to get proprietary software on machines without distributing and installing it (and for no cost) that doesn’t work for OSS, let me know about it.
In the same article Jackson brings up a hypothetical scenario. In his scenario an un-specified proprietary program costs $50,000 a year. He compares this to having to pay programmers and systems administrators $60,000 a year to obtain the same level of support. He implies that going with the open source project “has led to the wrong choice.” But has it? How many proprietary software maintenance fees that cost $50,000 a year not require some of your IT staff time? I’d venture to guess that Jackson’s $50,000 a year proprietary system will come close to, if not exceed the $10,000 savings. Not to mention that in the case of many OSS projects, support can be purchased by outside vendors at a fee at a reasonable rate.
While I have used examples exclusively from Jackson’s article, I don’t really mean to point him out. If it was only his article that said this I would have probably just glossed it over and never thought about it again. The fact of the matter is that many administrators like Jackson while attempting to down-play the economic advantages of OSS, often conveniently forget about, or ignore, the additional costs of proprietary software.
I am not saying that when people are looking at OSS the shouldn’t figure on the additional costs… of course they should. When factoring a major acquisitions decision, all the associated costs should be evaluated. This is true weather it is OSS or proprietary. In his conclusion Jackson advises us to “avoid simplistic notions of good and evil.” My advice to Jackson is to avoid the simplistic notation that just because it is proprietary software doesn’t mean there aren’t additional costs. In other words, proprietary software is not free.
Reference:
Jackson, G. A. (2004). Open Source is the answer. Now what was the question. Chronicle of Higher Education, 51(5), B17-B18. Retrieved February 16, 2005, from Academic Search Premier database.