More on the economics of open source

Posted on February 24th, 2006 by Tim.
Categories: General/Misc., Tim.

 Liu’s article makes alot of sense to me concerning the economics of open source software, by which he means FOSS (Free Open Source Software). The ideas he presents are the foundations of the philosophies of the GPL/FOSS community. A strong argument for open source is that my contribution helps me, and it costs me nothing to share my contribution with everyone else on the planet. When everyone is doing this, I can obtain a large return on my work by participating in the community of open source development.

 There were three conditions that Liu specified that must be present as a catalyst for open source development:

  • All the contributors are building around a common framework.
  • Each contributor can directly benefit from their own work.
  • The cost of providing that benefit to others is minimal.

    •  Specifically, I’d like to look at the second bullet point. This is a very important concept in open source software. Open source projects generally come about when an engineer says “Hey, I could use a program that does X,Y, and Z,” and continues to start writing program Foo that does X, Y, and Z. He will probably start a project group on a site like sourceforge and gather other developers who also want a program that does X, Y, and Z. Economically, if there are n developers working on project Foo, then each of them put in 1/n of the effort, but get out 100% of the benefit. And since it costs nothing to share that benefit with people on the Internet, it only makes sense to share their work with the world, out of the goodness of their hearts.

     But return, if you will, to the original motivation for writing the software. The developers of program Foo began development work because they wanted/needed a program to do X, Y, and Z. If there was some external entity that needed a piece of software that did X, Y, and Z, there would be no motivation for any developers to begin working on program Foo, unless they too needed such a piece of software. This is true of the vast majority of FOSS. Even Linux began as a project where a lone developer said “Hey wouldn’t it be cool if I had a program to do …”

    There are some exceptions. Some people contribute to open source software purely out of the joy of doing constructive work in a group setting. These people are not as common, and do not make up the majority of long term contributors to open source software. Projects often go through phases where development loses some of the fun that it once had, and such developers fall away. In another realm, many of the contributions toward open source software are made purely on academic grounds, and these can in no way be ignored. In other scenarios, however, corporations contribute large amounts of resources towards open source development. In all of these cases, however, their motives can be traced back to the their primary goal of increasing profits or towards  furthering alternative long term goals. For instance, IBM contributes a massive amount of resources towards open source development, much of which is targetted at undermining Microsoft, which at the end of the day, is merely an effort to increase profits. But for the vast majority of contributors to open source projects, the motivation lies in the fact that they gain something personally from contributing to the project.

    Here is where we get to the crux of the matter. All of these contributors fall under a very narrow category: technically competent, technologically oriented individuals, most of whom are computer scientists or software engineers. Therefore, the programs that will be developed under FOSS principles will always be useful to people within this category. This is the real reason why your grandma can’t use Linux. If a group of computer scientists are writing an operating system, they will write an operating system that is useful to computer scientists. They have no economic incentive to write an operating system that is useful to your grandma. And this is the fundamental reason that free open source software must always co-exist with non-free software. In order for entities to obtain software from a group of developers, when said software offers no intrinsic benefit to that group of developers, there must be an economic incentive for this group of developers to develop said software. In a nutshell, you gotta pay for it!

    Through economic reasoning, we can completely understand every decision that Microsoft has made in it’s history. After all, Microsoft wants to make money; no one can disagree with that. Why did Microsoft embrace the Internet? Because consumers would demand software that used the Internet. Why did Microsoft change their philosophy to make security a top priority? Because the market demanded it. This is the fundamental difference: Open source software focuses on the needs/desires of the developers who write it, whereas non-free software focuses on needs/desires of the consumers who offer economic compensation.

    There will always be software which is best written as free open source software, and there will always be software that is only economically viable as non-free software. This is an economic reality.

    8 comments.

    Chris

    Comment on February 26th, 2006.

    I wonder occasionally whether operating systems will always be commercial software. It’s true that Linux was originally built for engineers and computer scientists, but it may be at the point where I, as a non kernel hacker, would start fiddling with it. And of course, as a UI person I would probably spend a lot of work on the interface, making it more usable.

    If academics in the field of HCI (human-computer interaction for all who aren’t Gas) start making it their thesis projects to design friendly interaction systems for open source programs, we are going to start seeing friendlier shells to Linux. Not sure this will happen; just a theory.

    Chris

    Comment on February 27th, 2006.

    I guess to connect this more thoroughly to your idea:

    - Open source software is written for the needs of the designer.
    - Commercial software is written for the needs of the consumer.

    - As more people are allowed to be designers (that is, as the barriers to entry for programmer/designer are lowered), the market and interest in open source will increase.

    Wow, Gas, you just made a good case that what Linux needs to do to capture the consumer market is focus on developer tools. Good job us.

    Tim

    Comment on February 27th, 2006.

    No, I will go one further. Open source software is written for the needs of programmers. Designer is much too broad of a term. As a “UI Person” would you have the slightest idea how to contribute to an open source project? Non-technical management and design positions simply don’t exist in open source projects.

    This isn’t just about UI either. For instance, there will always be tradeoffs concerning ease-of-use versus security, and in a project run by computer scientists, what do you think they will choose?

    Improving developer tools would help, but I still don’t think it will change the dynamics of open source.

    The Blog of Justice » FOSS and the handicapped

    Pingback on March 19th, 2006.

    [...] A few weeks ago I wrote this article about the economics of open source and how it affects the types of software written. Apparently I’m not the only one to have noticed. [...]

    The Blog of Justice » FOSS and the Handicapped (Part 2)

    Pingback on March 23rd, 2006.

    [...] After writing an article about the economics of open source, I keep finding random commentary that agrees with the point I was making. This time it was a post at Rick Schaut’s blog that was linked to by Raymond Chen. [...]

    Chris

    Comment on March 23rd, 2006.

    Claiming OS software is always written for the needs of programmers is certainly true for a majority of projects, however you extend this claim to be a universality: “Non-technical management and design positions simply don’t exist in open source projects.”

    I’m willing to believe that this is true to an extent, and even in a majority of cases, but I think to claim anything as tautological–”Open source is written for the needs of programmers”–and then extend that argument to conclude that open source software will never be designed for the needs of anyone but programmers is incorrect, by your own arguments:

    A) The commercial developement of open source to undermine Microsoft’s position, IBM for example, would seem to be a perfect example of a situation where a company would take a direct interest in not simply producing software for programmers, but rather for the lucrative enterprise market. IBM would not be seriously undermining MS unless it targeted MS’s consumer base; thus IBM (or, say, Red Hat) would have an interest in offering features–everything from security to ease of use and interoperability–that appealed to customers who would then stop using MS equivalents.

    The vast majority of open source development may be for programmers, and thus cater to programmer needs, certainly. However, the idea that these programs will never cater to the needs of non-programmers does not explain, for example, Firefox or the OpenOffice initiative. Unless you want to claim that both are written for programmers, your original thesis.

    Just as you noted that corporate contributions to FOSS were an exception to the “by programmers, for programmers” rule of personal gain, you also noted academia. While you dismiss these contributions as unimportant at the end of your paragraph,

    “But for the vast majority of contributors to open source projects, the motivation lies in the fact that they gain something personally from contributing to the project.”

    The fact is that corporate contributors, who are concerned with the interests of non-programmers, are both a major wellspring of the open source initiative. This can only be explained by claiming that there is an economic incentive to OS devlopment that you don’t note in your article.

    B) So the question must be this; why is there an incentive to develop free software? Personally I think I should write a new entry about this, but I would say that one possibility is the ancillary profit motive. If Red Hat charges for its OS, but it is offered as FOSS, then how can it expect to find paying customers?

    Red Hat’s explanation for this suggests an evolution in the concept of software not previously considered:

    Enterprises simply can’t afford software that doesn’t come with full assurance of reliability and performance, and a 24×7 support infrastructure they can trust. They require comprehensive technology integration. All the pieces need to fit.

    To cater to an enterprise market for software, Red Hat is selling not the software it offers, but rather a guarantee or standard of reliability. This is the same as how, say, plumbing is offered to a regular home. We are not sold the intellectual property of the pipes and fixtures; we are sold the guarantee of reliability. The software itself might as well be free; one can retain plumbing without getting regular releases of new plumbing, but the work required to maintain it and keep it usable falls to the company providing it.

    Ehh, this is too long. I shall rewrite it. Why did you stop going online?

    Chris

    http://www.businessweek.com/technology/content/dec2005/tc20051228_262746.htm

    The Blog of Justice » How can you make money if it’s free? (Profit models of Free Open Source)

    Pingback on March 23rd, 2006.

    [...] A while back Gas wrote an article that criticized open source software as not offering the necessary incentives to drive people to develop for deaf, blind, grandma, and other types of non-programmers who use computers. (In response to this article I wrote before that). [...]

    The Blog of Justice » Is free software really free?

    Pingback on April 2nd, 2006.

    [...] Squeegie and I have written a number of articles on the topic of the economics of open source. One thing that I think neither of us has directly addressed is that there is a great deal of disagreement about the use of the terms “Open Source (OS),” “Open Source Software (OSS),” “Free Open Source Software (FOSS),” “Free/Libre/Open Source Software (FLOSS),” and “Doesnt Everyone Notice That Acronyms get Longer Free/Libre/Open Source Software (DENTAL FLOSS),” etc. Our good friend Richard Stallman (also known as RMS) has been vocal about distinguishing between these terms over the years (but RMS has been known to be… eccentric…). Originally, there was little confusion over the terms; Open Source refered to projects such as GNU, Linux, and Apache. As companies sprang up to profit from open source software, and as licenses appeared that were incompatible with the GPL, there arose a need to distinguish between these various classifications. Generally, people simply lump all of them together under the category of “Open Source”. [...]

    Leave a comment

    Comments can contain some xhtml. Names and emails are required (emails aren't displayed), url's are optional.