Investing for a brighter, greener, more equitable future

Sustainable Investment

Subscribe to Sustainable Investment: eMailAlertsEmail Alerts newslettersWeekly Newsletters
Get Sustainable Investment: homepageHomepage mobileMobile rssRSS facebookFacebook twitterTwitter linkedinLinkedIn

Sustainable Investment Authors: Pat Romanski

Related Topics: Apache Web Server Journal, Sustainable Investment

Apache Web Server: Article

The Powerful Economic Underpinnings of OSS

Open source software has highlighted the fertility of "the commons"

In 1968, Garret Hardin wrote a seminal paper that ran in Science Magazine called “The Tragedy of the Commons.” Hardin defined the commons as a place where multiple people are each endowed with the privilege to use a given resource, and no one has the right to exclude another. Think of a pasture where many farmers can graze their animals. When multiple users have such privileges of use, each user benefits directly from using the resource (one more cow in a farmer’s herd benefits that farmer directly) but the cost of each person's use is borne by all users (the increased use that one cow puts on the pasture affects all users of the pasture).

The result is that each user has an incentive to use the common resource more, and less incentive to curb their use of the resource. The resource is thus prone to overuse and, over time, it degrades - a tragedy of the commons. The idea of a "tragedy of the commons" can be applied to a wide variety of issues including things like air and water pollution, individual consumption choices, and - Hardin's main target - population growth.

Open source software (OSS) development and use, as described by the General Public License (GPL) and other similar licenses, creates a unique situation that yields surprising outcomes when held up to Hardin's analysis. As we shall see OSS effectively turns Hardin’s "tragic" conclusion on its head. Key aspects of open source software development, distribution, AND use enable the open source software movement to create a tremendously positive outcome. Hardin’s analysis applied to OSS shows that tremendous gains in the quality and volume of software development and use are the likely outcomes.

Because this is a pretty long analysis, we'll hit you with some of the conclusions right now. Using Hardin's analysis, we'll show that OSS development is much like the Tragedy of the Commons turned on its head. Unlike the tragedy, OSS leads to a plentiful outcome where the amount of software in the market should continually grow and the quality of that software should continually improve. We'll also show that, as these continual improvement dynamics become more widely recognized, the integration of OSS development and use within business and government institutions will mature, further accelerating the strongly positive aspects of the cycle.

This article was written because Hardin's analysis is an interesting one when applied to OSS and because the implications of the analysis have important messages for public and private organizations considering greater involvement in OSS projects both on the user and the development side.


Perhaps the biggest potential criticism of invoking Hardin's analysis here is that critics would claim OSS is not like the cow pasture or polluted river to which Hardin's analysis is usually applied. That is true. The two important differences are these:

  1. the use of OSS does not have a detrimental effect on the quality of the commons (i.e. OSS cannot be overused)
  2. the use of the OSS commons actually improves the commons (as use of a particular piece of OSS grows, the developer base for the software generally also grows).

So, what is the OSS commons?

On the development/management/maintenance side, the commons is the human-readable source code. It is here that the armies of open source developers spend time adding to, cleaning up, and enhancing the OSS product. On the use side, the OSS commons is the copy of the executable binary that each user gets to use. Like a pasture, this binary is used in different ways by each user, just like a pasture can be used to graze cows or goats. So the commons for OSS are a little more complex than Hardin's commons.

Looking deeper, for the maintenance/development side there is a single commons (unless the code base has forked) and on the use side, there is a weak concept of a commons in that the OSS use commons can be duplicated for the nearly zero cost of network bits or CD-ROM blanks. These differences constitute an exciting turn of events from Hardin's commons as far as the economic underpinnings go. We couldn't have it any better, in fact. If only cars or houses had the same economic attributes as OSS. And, this is just the leading edge of the good deal that is OSS with a maturation of the funding of OSS that will follow in the coming years. While there are good conclusions ahead of us, we can see here that Hardin's analysis can, in large part, be applied to the OSS dynamic with very little modification.

Turning the Tragedy into a Cornucopia

Here is what is really going on in regards to Hardin's analysis when applied to OSS development and use. The key aspect of Hardin's analysis is that a common resource is subject to overuse and degradation when multiple users have unrestricted rights to use the commons. Because Hardin's analysis addressed both management of the commons and use of the commons, let's look first at management of the commons.

OSS Development

In OSS, management is the same as software development, bug fixing, maintenance, and other actions that improve an OSS project in some way. OSS management of the commons is typically thought of as writing code but it could also include documentation development, debugging or a host of other things that improve the software product or its ease of use.

The foundation of OSS development is volunteerism - via the Internet. There is a direct connection between the level of volunteer developer interest in a particular OSS product and the installed user-base of that project. Sure, certain major projects are occasionally one-man bands such as qmail, the popular mail transfer agent written by Daniel J. Bernstein. In general though, a correlation can be made between interest in use and interest in development. Linux itself is a testament to this ,as Linux development took off in 1991 when it reached a certain level of maturity. It's safe to say that, as OSS continues its growth trend on the use side, there will be a proportional growth in the number of developers ready to pitch in a hand.

The OSS development community is mature in some significant ways that augment the volunteer foundation - this is probably part of a decades-long process of maturation that is ahead. Recent signs of this maturation have been the role big companies have started taking in OSS development. In the past year, a number of big companies have stepped up to contribute significant amounts of existing technology to the OSS movement. In addition, companies are increasingly finding justifiable economic businesses models that include participation in OSS development either through funding OSDL or by directly funding programming teams. As justifiable business models continue to become more commonplace, we will see the OSS development process continue to gain momentum and strength - and from both private and public participants if the development rumblings from Japan and India are any indication regarding OSS industry use consortia.

OSS continually improves because it allows new developers to stand on the shoulders of giants. Sure, there may be cases where bad design or decision-making lead to a step back for a particular OSS program, but for the vast majority of cases, OSS just gets continually better as the various developers continue to apply new code, bug fixes, and other enhancements. Even if all the developers left a project and the project became dormant, the code would probably still be available somewhere for download and because of that, a new developer or team could come back in at any time and continue building on the code base left by predecessors - starting not from scratch but standing on the shoulders of giants as it were. This is something that is simply not done with proprietary software without the code base being sold to a new entity - an expensive and tortuous process.

So, on the development side, we are seeing the beginnings of what could be a huge level of development support from business and industry and we are seeing the traditional source of OSS development talent - the user base - continuing to grow. An exciting time for OSS indeed.

Users Everywhere

On the use side, each user participates in OSS by installing OSS from a CD-ROM or network download and then running the software. Some users are getting pre-loaded installations but the pre-loads originate from a download or CD. No matter the source of the software, for users of OSS, the "commons" is the installed binary and the related configuration information. This definition is a little simple but this is good for now.

Use of the commons in this case has no impact on other users of the commons. About the worst thing possible as far as one user impacting another user is when the mirror you are downloading Linux or other OSS from is too busy and you have to come back later. Other than download traffic, there is no impact caused on a particular user by other users of OSS. A single malicious user can still undertake security attacks on other users but this is akin to a cow rancher using his cows to trample and destroy a common pasture and it is outside this analysis. No matter the species of software, OSS or proprietary, malicious users can attack other users.

Since OSS use is important as a driver for OSS development, let's look quickly at the drivers of OSS use. Way back in the hacker days, when the hacker culture dominated the net and OSS more than it does today, OSS was thought of as software "by geeks for geeks." That perhaps explains the preponderance of OSS on the server side. Linux and Apache, the two biggest OSS projects in terms of impact, are firmly rooted in the server side - not the client desktop, not the embedded space, and not the super computer or mainframe space even though great design is taking OSS into all those spaces. Under the hacker culture, use considerations were probably focused on "scratching an itch." A developer would code up something if it scratched an itch that was currently unscratched. If others had this same itch, the software was likely to develop a user base.

As the OSS user community has grown, we can say that the primary drivers of user adoption are now ease of installation and features. So, if a piece of software has useful features and is easy to get and install, then it is likely to be used. The Linux distributions are generally thought of as focusing on the installation side of things by making it easy to take any reasonably standard PC, server, or laptop and install a nice version of Linux on that hardware. The easier it is and the better the features, the more people will use it.

On the feature side, we also see some other user dynamics on the demand side. Specifically, software is complementary to other software. For example, because Apache and Linux are more complementary than Apache and Windows XP, we can see that as people adopt Apache, they bring Linux or one of the BSDs along more than XP. Similarly, as people decide to move to Linux, Apache use also grows more than IIS, the Windows-based Web server. Another complement to software is hardware. In this sense, as hardware gets cheaper and more powerful, that should drive increasing use of Linux (and all software in general).

Summarizing the user side, we see ease of adoption and the features of Linux and all its software complements as driving use. Because use by one person has no cost for other users and in fact use by one person has benefits to other users through development, OSS is like a big snowball gaining momentum with every day as use feeds back into development which leads to new features which lead to more use. These dynamics are key parts of the powerful economic underpinnings that drive the OSS movement.

But How Great is the Future Going to Be?

The next part of Hardin's analysis was how bad the tragedy would be. Hardin basically said that, if left unchecked, the resource would degrade until it was unsuitable for its current use. A way to state Hardin's claim more generally is that, as the resource degrades, the uses that it can be applied toward become fewer and fewer and the remaining uses tend to be less economically profitable. A resource that could have been used for either ranching or a parking lot may only be useful as a parking lot after it has been degraded.

In the OSS case, we have to look at how far the good of the OSS commons will extend and how quickly.

Because OSS generally makes continual improvements in features, bug fixes and so on, it's a like a ratchet wrench that only turns in the correct direction. Flipping Hardin's conclusion of degradation to one of improvement, we see that the OSS resource should generally become suitable for new uses over time. This can be thought of as portability and new applications. Portability is one area where OSS shines, with the BSD flavors being ported to over 1000 platforms. The presence of Linux in both supercomputing and a wide array of embedded devices shows vast advances in the hardware applicability of the software when compared to any proprietary platform. In other areas like GUI desktop environments, OSS has found the challenges a little steeper but there is no doubt that these challenges will be met in time.

The other aspect to this cornucopia is how fast that ratchet wrench is turning. To that issue we can say that, today, the wrench is turning surprisingly fast on many fronts. For example, bug fixes have been shown to come out about 6 times faster for OSS than for proprietary equivalents. It’s also pretty astounding to think about the strides and maturity of Linux, a project that started in 1991 that now exceeds comparable proprietary offerings in many (but not all) ways. So, it's safe to say that the wrench is turning quickly now and will only accelerate over time as the user base grows and as business and industry involvement in OSS continues to mature resulting in greater levels of development involvement.

Fueling OSS Even More

Lastly, Hardin focused on ways the tragedy of the commons can be avoided. Since OSS leads to a good outcome, our analysis may instead focus on how to make a fantastic future better than it might otherwise be. Here, we have to look at solutions for managing a resource that has an improvement and momentum all its own. I think the closest analogies here are the economy itself. In the U.S., government economic policy continually focuses on building an environment for sustainable economic growth.

We want our economy to grow but not so fast or so slow that we have under-employment one year and then over-employment the next. So, we have collectively decided that we want our government to take an active role in managing the state of our economy. OSS is not really much different. By having more organizations take an active role in the development and use of OSS, we should see more support for development and use and we should see features and quality levels accelerate to new levels.

We are now starting to see some of these signs with the profitability of Red Hat, the solid funding for OSDL, the use consortia announced for consumer electronics in Japan, the distribution consortia announced by India. There seems to be clear interest in making Linux distributions at the country level with government involvement.

Whoever may be looking at an OSS development investment should be looking at the economic dynamics of OSS – that it is perpetual, always improving, and always growing into new applications. Investments should be thought of as opportunities to join forces where interests align to accelerate development in a particular direction. The ROI analysis on making these types of investments should surely include the significant on-going costs associated with maintaining a proprietary software infrastructure. Perhaps the holy grail here is in aligning businesses and governments toward accelerating a desktop replacement for Windows using an OSS model. Linux on the desktop for U.S., European, and Australian business and government is something the author is actively working on.

What About the IT Industry and the Economy?

Now, we have mentioned the word profit in a few places above. What will the impact of the OSS movement ultimately be on national and world economies? In general, OSS will tend to continue to wring cost out of IT by allowing developers and users to continually stand on the shoulders of the giants that have come before us. Development is a big focus now but as the movement matures use may take the forefront with issues like configuration, transition strategies, hardware integration (as opposed to today’s hardware compatibility), and other types of non-development activities will mature largely in an open way and will continue to wring cost out of IT and information activities. When Sun’s CEO, Scott McNealy, claims IT costs can go down by 90%, it will not be with a company like Sun doing the same things Sun does today. We should be left with a minimal cost structure around the most popular hardware/software/network solutions while remaining IT expenditures will focus on custom software development and proprietary innovation in spaces that are outside the core of today's PC / cell phone / consumer electronic space.

As cost is wrung out, end users will see savings and expansion in the types of products, information services, and features in a wide array of technology and information industries. Stretched to its limit, this analysis would support a claim that a new user in Argentina might latch onto an OSS project through his or her use of Linux. That user later becomes a developer and contributes a code snippet that leads to software that enables the company who makes your car to virtually eliminate long distance phone charges. You end up saving $100 on your next car due to OSS software having to do with phones. Odd but certainly within the realm of possibility.

A Final Thought

To counter the "tragic outcome," Hardin suggested that humankind either evolve to understand the larger picture of the effects of multiple users acting on a common resource, or that the commons be managed and protected in one of two ways: through privatization of the commons or through government regulation of the commons.

Hardin preferred evolution to protection, but he knew that humankind was slow to evolve in such a basic way. Protection has proven to be the case, as we have needed laws to curb our tendency to destroy common resources. Perhaps OSS is an example of human evolution on some level coupled with or augmented by the Internet and the fortuitous economic attributes of software. It certainly seems that way to me.

More Stories By Paul Nowak

Paul Nowak first used Linux in 1995 while migrating from Sun to Linux at the University of Michigan. He used Linux in subsequent IT projects including web, telecom, telemetry and embedded projects and is currently CIO of a small professional association based in Washington D.C.

Comments (19) View Comments

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.

Most Recent Comments
Rene Vouri 03/19/04 03:24:30 PM EST

and now to something completely different:)
in the article you wrote, that bugs are solved up to 6 times faster in oss than in propietaere systems.
i would like to know on which study you base that statement.

Paul Nowak 10/31/03 11:18:21 AM EST

Mick, I think those are excellent ideas and insight. If you toss on your international public policy hat one more time, what kinds of suggestions do you have for improving outcomes for the OSS movement? As the OSS matures and continues to forge it's way into the mainstream, what types of policies and approaches are likely to succeed? Could be anything from training on how to lead an OSS project to something like the treaties you mention above to pull OSS development in new directions or better coordinate country-level policies toward OSS use.


Mick Wilson 10/31/03 10:31:35 AM EST

Focusing on the historical 'reality' of Hardin's idealized commons may not be doing us much of a favour. I suggest it could be more instructional to take at look at the 21st century version.

A number of resources that our forebears treated as inexhaustible are now subject to international consensus-based treaties - think of the atmosphere (Montreal and Kyoto protocols, regional transboundary pollution agreements etc.), the oceans (UNCLOS, regional seas programmes, hazard waste shipments), biological resources (migratory species, biodiversity) and so on. I am not suggesting that these treties are equally worthy or effective, but they do represent a shift from the view of an inexhaustible "commons" to one where recognizing finite magnitude leads to collective management.

I would liken the OSS movement to one of these global environmental matters - participation by individuals (nations) from all around the world, a sense of local action influencing a global issue (think of ozone-depleting gases being eliminated from your local deodorant), a sense of a collective public good being served through self-interested action by like-minded individuals or nations. And few, if any of these treaties are backed by punitive policing but rather by perceived standing in a political arena.

So what does this tell us about the OSS movement? Perhaps that it can be seen as one element of a emerging global propensity for multi-lateral management of at least some types of shared resources, in this case an intellectual domain.

Paul Nowak 10/24/03 01:46:46 AM EDT

Greg, *I* said Hardin was controversial. I didn't say you said it.

I respectfully disagree that every commons that has ever existed has been managed by consensus. Many commons simply don't get managed. This is often a cultural and economic reflection but major cities throughout the world are maintained at very different levels. The public places in these cities are a commons but they get very different levels of management. Is this a consensus decision? No. Keeping the streets clean just is not high enough on the priority scale in much of the 3rd world to make it on the list.

Population is the same. Few countries have effective population growth management either because of cultural blocks or other reasons. So, I just don't don't buy your arguement that every commons is managed by consensus when many are simply not managed at all.


Greg 10/22/03 11:23:43 PM EDT

I did not say Hardin was deficient. I said his model was deficient.

I wrote about his model above, on 18 October, the sixth comment on this page.

Paul Nowak 10/22/03 07:35:44 PM EDT

Greg, So please offer a tidbit of Hardin's analytical deficiencies. I think Hardin himself is very controversial but his analysis of the commons is a living part of economic analysis and resource management.


Greg 10/20/03 06:43:04 PM EDT

Sure, Paul, however if your frame is fallacious, then whatever argument which you hang off it can never be valid.

You give Hardin a prominent place in your article, a prominence which gives his faux commons a spurious credibility, if allowed to stand uncorrected.

In any case, as you recognized when you adopted the metaphor, the Free / OpenSource software movement is a collective effort to define a software commons. Hardin's model and especially its deficiencies teach lessons which will likely be relevant to the ways people perceive and nurture it.

Informed discussion of Hardin's model is quite germain.

Paul Nowak 10/20/03 03:13:19 PM EDT

It's been nice to see all the informed discussion about this article. I've wanted to respond to several comments but more importantly, I want to re-focus attention on the main thrust of the article. I don't think it comes out strongly in the article but Hardin is really just a framework that gives structure to a discussion for describing AND managing economics of OSS. The real thrust of the article is here:

"Whoever may be looking at an OSS development investment should be looking at the economic dynamics of OSS – that it is perpetual, always improving, and always growing into new applications. Investments should be thought of as opportunities to join forces where interests align to accelerate development in a particular direction. The ROI analysis on making these types of investments should surely include the significant on-going costs associated with maintaining a proprietary software infrastructure. Perhaps the holy grail here is in aligning businesses and governments toward accelerating a desktop replacement for Windows using an OSS model."

My next article will be about ways of "accelerating development" of OSS in targeted ways. We've already seen a number of models for OSS development some of which are emerging and that I therefore want to focus on more (OSDL, IBM programming teams, animation studios) and others which are huge but have been long standing foundational sources of development contributions (individual volunteers).

Paul Nowak

Ewing Caldwell 10/19/03 06:19:36 PM EDT

Before a commons can be successfully created, all the impediments to one must be identified. At present, the
accepted structure is of "property" and individual "property
rights." Copyright and Patent law reinforce this accptance.

What if:
a public body (a govt qango (Quasi-AutoNomous Government
Organisation ---eg: NASA :-)) be established to register and
manage copyright and patent. An inventor, an author (book,
screenplay, software program, cooking recipe, etc) registers
their work with this Intellectual Property Registry (for want of a better name) and deposits their work with it. The *ownership* of the copyright and patent remains with the
author/inventor. It cannot be signed away.

The IPR then offers a licence (one or more) for the works, at auction, to publishing houses to print, distribute and sell, or contracts directly with various publishing houses. The IPR collects the royalties (and ensures they are paid!) and forwards them to the author/inventor.

I see the advantages of this approach as:
1. the authors and inventors are paid their royalties.
2. A fair and reasonable royalty can be set
3. the authors and inventors retain their "property rights"
- they can't be signed away
4. works enter the public domain in their correct time
5. works do not drop out of sight when they drop out of
popularity. The IPR is a permanent public repository
anyone can access any time ...

Music and software are covered in the same way as books.
This way the RIAA can't sue 12 year olds for IP violation.
Instead, it will be the IPR office .

(Speaking of the RIAA, it is a perfect example of the evils
of monopoly rights. We need to recognise and control these
rights properly if they are not to become the out of control
monsters the RIAA has become.)

Scientific research is a global commons. It is a commons
protected by a well-known methodology and guarded (jealously
and assiduously) by its custodians: research scientists and

This will maintain the rights of the IP producers, the members of society and more properly limit the powers of
large companies/amalgamations of capital.
(NOTE: I am *not* suggesting "state ownership" but responsible community control to protect individuals from
being ripped off by corporates.) And it should go a ways
towards developing and broadening the intellectual commons.

layton davis 10/19/03 05:17:37 PM EDT

It is true that there are many differences between the various "open source" licenses. I agree that these are mainly derived from differences in beliefs between the individuals creating these licenses. However, I still believe that the GPL has had an influence that affects the practice of using the other open source licenses.

For example, for most people who buy a linux distribution, be it Red Hat, SUSE or whatever, much of the software on that CD is covered by the GPL. As those people become familiar with the GPL, they tend to assume that all of the open source licenses operate similerly.

Even those of us who understand many of the differences will treat BSD licensed projects the same way we would treat GPL'd projects for one of two reasons 1)It's simpeler or 2)out of courtesy to the project coordinator.

Back to the idea of a "commons" in land use. There are some principles which the open source community has discovered and which I believe apply here.
1) the principle of ownership. Ideals of RMS asside, the GPL is the open source license which most clearly identifies ownership and a set of rights belonging to that ownership. It also puts very strict limits on how users of that which is owned can use it. Believe me, If it was truly "free" Microsoft would have included much of that code in their Windows product, just like they have appropriated BSD code.

2) the principle of returning value to the community. I don't recall this being encoded into the open source licenses, but the article "the cathedral and the baazar" pointed out that this is an almost universally practiced law among open source participants.

3) Voluntary participation. This is from my own observations. You don't have to use linux. There are still many commercial operating systems on the market other than MS Windows. and some of them have stability that matches, and in some cases(as of linux kernel version 2.4.20) still excedes the stability of linux.(although not by much)

I believe that all of these principles should be followed in setting up a land use commons.

Ewing provided links to some very informative material which I believe could work awesomly well, if it were actually implemented. For those who don't want to go read it for themselves, basically the idea is dropping all taxes except for the land tax(land only, not to include improvements) and that the tax be slightly below the rate currently set by real estate appraisors to rent the property in an un-improved condition. Along that line I would suggest the following tax exemption, which, if you observe the rural communities now in existance would not ammount to much. That is to allow a 1 acre exemption from tax for those who lived outside of any city limit and did not use any community provided utility(eg water or electric). I would suggest you read for a better idea of the logic behind the suggestions.

Short of this, I wonder if there is something that could be done within the confines of our current tax and property ownership conventions that could create a commons.

Kevin Bedell, Editor in Chief of LinuxWorld Magazine 10/19/03 01:14:36 PM EDT

Just because a project uses the GPL as its license does not mean it is or is not 'open source'.

In many ways the differences between the Free Software and Open Source software movements have more to do with social goals than the software itself.

The Free Software movement had its roots in Richard Stallman's initial effort to create a 'free' operating system. This effort led him to quit a job at MIT and begin working on it full-time. His primary goal and motiviation was to create something that ALL USERS had free and open access to and that could be inspected and modified by the users themselves.

He quit his job both to dedicate all his time to this as well as to ensure that MIT would have no possibility of copyright or patent claims based on any of the work he did.

I recently had the opportunity to interview Richard and spent time discussing this exact point.

It was really about power in many ways. The power to know what the software on your machine is doing, and the power to change the software if you wanted to stop it from doing things you don't want or to add features to make it do what you need.

One of the great examples Richard used regarding the value of free software today is 'spyware' -- code embedded in your systems you use that monitors and reports on your activities.

For example, if users wanted to, he reasons, they should be able to modify Microsoft products to limit the amount of data that MS can collect on you and the applications you run. The central tenet is that this is a 'right' we should have -- a right that we as a user community have not claimed.

The goal of the Free Software movement is to provide free versions of software that does *everything* you need. Free operating systems, free utilities, free productivity applications, etc. Many even believe that software inside your cars, in your cell phones, in voting machines, in your tivo box, in your entertainment systems, etc should all be free. We have the right, it is reasoned, to know what this software does.

[Look for the interview of Richard to appear in an upcoming issue of LinuxWorld Magazine.]

The Open Source movement, on the other hand is based on sharing source for more pragmatic reasons. While user control of the software is certainly part of the reason for Open Source, it is not necesarily the primary one.

Many of the reasons software is developed as open source have to do with the fact that it in many ways is a superior model for developing software. It allows the users themselves to fix bugs in the software, allows greatly expanded testing ('many eyes on the code'), and can speed innovation because it facilitates the adopting of new ideas from so many different sources.

The principle method by which software is considered 'open eource' is if it is licensed under terms that meet those specified by the Open Source Institute ( ). The terms software must meet to be considered 'Open Source' are located at

These terms were developed by Bruce Perens along with Eric Raymond and a few others. [Bruce is on the advisory board of LinuxWorld Magazine and we have reprinted a recent speech on 'the state of Open Source' in this month's issue. We also featured an interview with Eric Raymond in our October, 2003 issue.]

LinuxWorld Magazine will continue to provide information on both these communities. Our February, 2004 issue, for example, is planned to focus on nothing bu profiles of free and open source software projects and applications.

Ewing Caldwell 10/19/03 06:27:39 AM EDT

"The Tragedy of the Commons" is a myth. It was originally circulated as propaganda by landowners in England several centuries ago to justify the enclosures and privatisations of the English commons.

Anyone who presents this argument as a justification for anything is to be read with suspicion. It is a plausible sam which is used over and over to justify an invalid case, because it worked so well for the English landowners.

Greg said it well. I like Layton's idea ... see for some interesting
material on those lines.

ttempting to scan you about?

It is used often to because it worked over and over again for the English landowners.


BO 10/18/03 04:11:15 PM EDT

The GPL is OSS

From the article

"Open source software (OSS) development and use, as described by the General Public License (GPL) and other similar licenses, creates a unique situation that yields surprising outcomes when held up to Hardin's analysis."

Is completely accurate.

The GPL is most certainly OSS. The source is required to be open to all users. It was open source before there was an Open Source Initiative. The distinction is that OSI is not always GPL.

From the OSI Web site:

"The "classic" licenses, GPL, LGPL, BSD, and MIT, were the most commonly used for open-source software before the Mozilla release in early 1998. The Mozilla Public License has since become widely used. Many other licenses have been submitted for review and approval by OSI. As you can see, the list of approved licenses is growing."

See the website, as Alice Rightly says it is confusing.

I really liked the article, I hope this bright spot in one area doesn't deflect the criticism of Hardlins work as it applies to things like copyright and the public domain in other areas.

Greg 10/18/03 03:44:04 PM EDT

"a place where multiple people are each endowed with the privilege to use a given resource, and no one has the right to exclude another" is not a commons; it is a figment of Hardin's imagination. He defined something which cannot be managed, and must be misused.

Every commons that has ever existed has been managed, by consensus, by the people who depended on it for life.

Societies, which don't learn to manage their commons sustainably, die quickly. Usually, however, commons are destroyed when armed might displaces consensus management.

A farmer and every one of his neighbours knows how many cows he needs to survive and how many square feet of turf he needs for each cow. This establishes the lower limit, where arguments become bitter and knives leap from sheaths. It is only the stranger, or the neighbour blinded by greed and a bigger knife, who does not care who starves.

In the case of Linux, SCO is a stranger, not armed well enough to take the commons by force, who hopes by bluster, banging his sword on his shield, to scare off defenders or perhaps attract a few to stand by his side.

Historically, one must distinguish invaders who are desperate from those who are greedy. Often the latter will encourage or propel the former to attack first.

Sheep, BTW, are rarely pastured on common lands. They are sent out to 'waste lands' because they destroy the turf.

Joseph 10/18/03 02:51:06 PM EDT

Hardin's analysis only applies to rivalrous public resources, and free software is non-rivalrous.

Rivalrous means that if I am using something, I block you from using it. Common land is like this, if I graze my sheep on a particular square foot, you can't also graze your sheep on that same square foot. Free software is non-rivalrous. If I am using a piece of software it doesn't stop you from using it.

This distinction appears in elementary ecenomics textbooks, and I am glad Paul Nowak has discovered it.

Interestingly, people appear to have different standards for appropriate use of rivalrous/non-rivalrous resources. Many people engage in music file-sharing who wouldn't eat your lunch...

michael 10/18/03 12:22:53 PM EDT

i think that your thoughts on open source are correct but i think that the pasture is just a bit bigger than open source and you have to look at the software industry as a whole. then you can truly see the results that oss is have and will continue to have on the "pasture". I like the ideas presented but i think they are a little narrow in scope. thanks, michael

Alice 10/18/03 12:17:17 PM EDT

The GPL is not OSS!

This may sound nitpicky, but it's an important distinction. The GPL is the brainchild of Richard Stallman, founder of the FSF- the Free Software Foundation, GNU software, and software licensed under the GPL and related licenses, is "Free Software". Not "OSS". And it refers to "Free as in Freedom", not free of cost.

"Open Source Software" is a term used by the OSI- Open Source Initiative,, starring Eric Raymond and other luminaries. There are many OSI licenses. Some are GPL-compatible, some are not.

Both the FSF and OSI are real live official organizations, with boards of directors, lawyers, copyrights and trademarks, volunteers of all kinds, and specific goals and missions.

Your article makes a number of excellent points. Perhaps a followup could be on the distinctions between Free Software and Open Source Software, and the goals of the two groups? Hardly anyone gets it right, especially in mainstream media.


Paul Nowak 10/17/03 04:35:55 PM EDT

Layton, please give an example of a structure for land use that would be contributory and voluntary.

Software has the advantage that what happens with the commons that users interact with cannot flow back in a negative way to the commons that developers interact with. While quality of a land-based commons is going to be a dynamic balance of use degrading it and management/maintenance improving it -- both of which stregthen and weaken over time, OSS quality is like the ratchet wrench of continual improvement.

I back your idea of applying aspects of the GPL to resource management but start us with an example. Some possible hints are to somehow make the percieved attributes of a land resource more like that of software.

Paul Nowak

layton davis 10/17/03 04:14:47 PM EDT

There are fewer differences than you may immagine. The tragedy of the commons assumes that there is no rule to put limits on the users of the commons. OSS does have rules on the use of the commons, and it is called the GPL. Even the BSDs gain from the GPL, because it has tought people to think in terms of what can be done to improve rather than to use the commons. As this article pointed out, there will always be abusers.

What I wonder is this -- what would happen if a commons were developed in land use with a structure of contributory and voluntary use similer to what the GPL has provided in the software world.