Why sharing matters more than marketshare to GNU/Linux

Why sharing matters more than marketshare to GNU/Linux


In a recent article, Ryan Cartwright argued that free software isn't playing the "same game" as proprietary software is. He's right—but that begs the question: what game is GNU/Linux playing?

Thirty years of proprietary software thinking have conditioned us to think that marketshare is a critical measure of success, and so we've convinced ourselves that we have to "win" against Windows in order to "succeed". But this is simply not true. GNU/Linux can be a very great success even if it never achieves more than 1% of the installations in the world. The reason is the difference between "power" and "freedom".

The freedom metric

"Success" means something very different when your goal is "freedom" than it does when your goal is "power". Proprietary companies exist to make money, thus power over the marketplace—the ability to demand tribute (in the form of Windows license fees paid to Microsoft)—is the essential definition of success. For free software, however, the goal is to provide the freedom to choose and use free software (the GNU/Linux operating system and its associated universe of applications). For those who seek to create freedom, such monetary tribute is a nice form of applause, but it is not essential to success.

"Success" means something very different when your goal is "freedom" than it does when your goal is "power"

Freedom cannot be forced on people. Freedom to choose GNU/Linux means also the freedom to choose Windows as well. We can argue that it's a bad deal, but we don't have the right to force people to choose one over the other. Nor should we pass value judgements on them for their choice: we don't know the basis of their decisions, nor can we claim superior knowledge of their business. In brief, we don't need power over them! Seeking it is actually counter-productive to our goal of creating a more free software ecosystem.

The most important measure of The most important measure of "success" for free software is when users who want it feel free to use it (Credit: Andy Davison / CC-By-SA 2.0)

Hence, we have no obligation nor need to eradicate Windows in the marketplace nor to consume marketshare. That's because we aren't a massive multi-national corporation which must exercise power over people in order to survive. Marketshare: defined either as sales of services or as total numbers of installations is a nice sign of success for GNU/Linux, but it is in no way essential to the end goal of achieving freedom.

Self-reliance

Freedom is primarily achieved by providing the means for self-reliance. When individuals can provide for their own needs independently, without placing burdens on others, they are more free.

Freedom is primarily achieved by providing the means for self-reliance

GNU/Linux provides self-reliance in the form of software you have complete control over. Instead of having to pay a tribute in order to receive a benefit from a corporate provider, you are able to provide for your own needs using a freely-available product. In practice, of course, you really do this through voluntary sharing networks or "the community" of open source users and developers, rather than trying to operate on your own. However, you are by no means required to participate in this community, and of course, there may very well be more than one such community if one is not universally appealing (as a simple example, there are lots of Linux user groups divided by geography or language).

What does matter: sharing and standards

So what makes you free to choose free software? Essentially, what you need are quality and stability. There is also the touchier matter of interchangeable data format standards. To explain these criteria, let's consider the reasons you might be un-able to use a free software application to do a job:

  • Software crashes due to bugs (quality)
  • Software is too out of date (stability)
  • Can't open the files (data standards)

The first two problems have nothing to do with marketshare for free software programs. They have to do with the level of development activity: How many different platforms and configurations has the software been tested in; how many people are finding bugs; and how much time is being spent on fixing them once they are found?

This photo from the 2006 Linux Kernel Summit shows 73 people. The stability and quality of the Linux kernel depends on the factors that allow these people to share their time towards improving it. Consumer marketshare is only one of those factors, and not the most important (PR photo)This photo from the 2006 Linux Kernel Summit shows 73 people. The stability and quality of the Linux kernel depends on the factors that allow these people to share their time towards improving it. Consumer marketshare is only one of those factors, and not the most important (PR photo)

All of these have to do not with the percentage of people merely using the software in the marketplace, but with the absolute number of people developing and testing the program. In other words, a niche program that less than 1% of the public is using, but which has a strong, highly-motivated group of developers and users may be more successful than a package which is used by 99% of the population, but has few people interested in keeping it working (of course, it's unlikely that such a popular program would not find dedicated users, but this is by no means a closely-correlated relationship).

what matters is the people who are sharing their time to work on the project

In other words, what matters is the people who are sharing their time to work on the project. And that is enabled by the nature of the license, which allows the software itself to be shared. In other words, the strength providing the quality and stability of free software comes from the sharing of community effort and of the software itself. Marketshare, as such, has a limited impact on this for free software (the major exception is when a company makes significant money from the product and therefore decides to share developer time to work on it—but again, there is no strong correlation here: a company is free to free-ride or contribute, regardless of its income from the software).

So why do people quote marketshare as if it were the absolute most important metric? Because, for proprietary software it is: if only 1% of the public buys licenses for software A, but 99% buy licenses for software B, then B has 99 times as much money to pay developers. And that's really important for proprietary development projects, because they don't share. By not sharing the code, they not only discourage the desire to share effort, but they actually make it much more difficult (or even impossible) to do. As such, all development and testing time on proprietary software must be paid for. Thus, it is the availability of funds, driven entirely by marketshare, that determines the quality and stability of proprietary software. That's why it's sensible for proprietary software developers to quote marketshare as a selling point.

But those rules don't apply to free software—it relies on voluntary sharing to achieve the same ends. And it turns out that that is much more cost-effective.

The third item, data standards, does have a connection to marketshare, albeit it a tenuous one. If a single supplier wields effective monopoly control of the marketplace, it can also monopolize the formats of data that are used for communications. For example, when Microsoft had such total domination of the word processing marketplace, it created a situation in which a proprietary data format—MS Word DOC format—became a de facto interchange standard. Since this format was secret and only fully supported by MS Word, it created an obstacle to people wanting to use free software. A similar situation exists with respect to various DRM encrypted file formats and video codecs used today.

If a single supplier wields effective monopoly control of the marketplace, it can also monopolize the formats of data that are used for communications

Avoiding monopolies

These are openly hostile attacks on market freedom by those companies. They are, as the law puts it, "anti-competitive", because they put competitors at an unfair disadvantage. This creates a situation in which quality is not sufficient to allow a product to be used. It isn't just free software that suffers from this! This is a problem for all competing software.

Thus, in the interest of promoting both freedom and market efficiency, such tactics, whether intentional or not, should not be permitted. This can be achieved through legal means, primarily by not granting monopoly copyright nor trade secrecy protections for such standards. It can also be achieved by social movements to use standards which are freely available. Finally, having more than one effective competitor in the marketplace will result in natural market forces to encourage use of open formats.

The one issue about marketshare that matters is that there isn't a strong enough monopoly to effectively control the marketplace (Credit: Robert Jorgenson / CC-By-SA 2.0)The one issue about marketshare that matters is that there isn't a strong enough monopoly to effectively control the marketplace (Credit: Robert Jorgenson / CC-By-SA 2.0)

However, even where marketshare matters, it matters only in the need to avoid extreme monopoly: it is not that GNU/Linux needs to have a 90% or even a 50% marketshare to be a "success", it is only that allowing one other software (Microsoft Windows) to have a 90% or 99% marketshare is damaging. A 10% or 15% share of the marketplace would be completely sufficient. More importantly, that 10% or 15% needn't belong to GNU/Linux—it just can't belong to Microsoft (or any other single provider). We're perfectly fine if that marketshare is controlled by Apple's OS X or Free BSD, or some other operating system, so long as it really is independent of Microsoft control.

Success for free software is having the freedom to use it

This is a very different situation than with proprietary software, where "success" is directly proportional to marketshare (or absolute market size). In the end, for the cause of freedom, we don't need a strong marketshare. We just need a free market and enough sharing to get the software developed.

Of course, freedom may not be everyone's goal for free software. Some people may simply want market domination, and there are companies like Canonical (maker of Ubuntu) who will benefit from increased adoption of GNU/Linux. Their financial successes—at whatever level—are beneficial to the overall success of GNU/Linux, because they mean that more money will be spent on developing GNU/Linux. However, even if such companies fail, GNU/Linux will not: development will continue whether there is financial backing from sales of services or not, so long as there are enough people who need the software enough to share their time in keeping it working and providing new capabilities.

Category: 

Comments

Paul Gaskin's picture

"Freedom cannot be forced on people."
- Terry Hancock

That is true. Freedom is not imposed on people.

"Freedom to choose GNU/Linux means also the freedom to choose Windows as well."
- Terry Hancock

That is true as well, but I'm more concerned about people's freedom not to choose windows. That's why it's important to push back against their monopolistic, anti-competitive business practice.

Consuming Microsoft's market-share is vitally important for the purpose of giving people the freedom not to choose Microsoft Windows, because Microsoft has so much money, they're subsidizing the hardware manufacturers to block users from access to choices other than Microsoft.

"We can argue that it’s a bad deal, but we don’t have the right to force people to choose one over the other. Nor should we pass value judgements on them for their choice: we don’t know the basis of their decisions, nor can we claim superior knowledge of their business."
- Terry Hancock

If I were a business owner, I would not permit my employees to use Microsoft products. If I were the owner of a business, I would have the right to choose what operating system my employees used on my internal network, and Microsoft products would be an unwelcome and insecure intrusion into the privacy, security and integrity of my network. So in that circumstance, I could require people to choose free software over Microsoft.

If I were in charge of choosing the curriculum of a computer literacy class or a computer science course in college, I would also feel entitled to require that the software they learn should be available to students with full respect for their user freedoms and available at no cost.

If I were in charge of setting standards for public infrastructure, I would be comfortable requiring that voting machines must not be proprietary software, and that other public infrastructure not be entrusted to proprietary (and implicitly inferior) software.

"In brief, we don’t need power over them! Seeking it is actually counter-productive to our goal of creating a more free software ecosystem."
- Terry Hancock

I agree, we don't need power over them, but we need to curtail their power over us. Seeking to effectively resist oppression is not counter-productive to keeping a free software ecosystem. It's okay for tactics to be reactionary in some cases, but the strategy needs to be proactive.

Terry Hancock's picture

If I were the owner of a business, I would have the right to choose what operating system my employees used on my internal network

So, you assert that there's nothing wrong with companies that require you as an employee to use Microsoft Windows? You are in a sense, correct, but I personally chaff at that kind of treatment.

If I were in charge of choosing the curriculum of a computer literacy class or a computer science course in college

You assert similar power for school administrators.

If I were in charge of setting standards for public infrastructure

Again, you assert that such organizations should have that kind of power over you.

You then assert that your proposed decisions are right (as if these "ends" justified the above "means"):

I would also feel entitled to require that the software they learn should be available to students with full respect for their user freedoms and available at no cost

I would be comfortable requiring that voting machines must not be proprietary software, and that other public infrastructure not be entrusted to proprietary (and implicitly inferior) software

Well, maybe -- those are fairly good ideas. But there are people who believe just as strongly in the opposite decisions, and can make their own arguments for why various proprietary softwares should be used. As a no-brainer example: many school curriculum planners would insist that it is much better for students to learn the particular software programs they are most likely to use in the business world, which right now, means Microsoft software.

Would you want them to have that level of power over you?

Power is a two-way street. Freedom is the centerline of that street. Don't reach across it, unless you want people reaching back.

In each of these cases, the take-home point is not that those people should have the power to make those decisions and enforce them on other people, but rather precisely the opposite: that because they do have power over other people, their decisions should be open to public scrutiny and democratic decision making.

The teacher should listen to his students' opinions on what software they should learn. The choice of voting software should be subject to approval by the electorate, and (one could argue) the source code should be available for public review. And so on.

The good news (from your PoV), is that you should have the right to agitate for those decision-makers to make decisions in favor of free software. The bad news is, the proprietary lobbyists have the same right.

Paul Gaskin's picture

"...there are people who believe just as strongly in the opposite decisions, and can make their own arguments for why various proprietary softwares should be used." - TH

Terry, suppose that instead of you and I discussing software, people were talking about slavery and abolition. Suppose one person had just made the case for abolition of slavery and someone who were a bigot and an ignoramus (not at all like you) responded by saying something like this: "there are people who believe just as strongly in the opposite"?

The answer is that those people who believe just as strongly in the opposite are wrong.

One of the popular criticisms of liberals made by "conservatives" is the lack of moral conviction or moral clarity. The "conservatives" call it "moral relativism" or "moral ambiguity".

In this case, hypothetically placing yourself in the shoes of your ethically deranged opposition creates confusion rather than clarification.

I remember as a child when commercial television and vending machines found their way into my public school. At the same time, school lunches were reduced in quality and availability.

There is a long tradition of academic freedom which is being ravaged by private interests. Now, where is the federal funding for public schools? Where is the federal funding for college education?

All of the public infrastructure of the United States is under attack across the board. It is the work of people like Grover Norquist who wants to "drown the government in a bathtub".

I'm really not inclined to consider your hypothetical question about the proprietary software being imposed on me in the same manner I would demand free software.

In the private sector, companies can and do require applicants to use proprietary software. They have that right to do so.

In school they also require students to learn proprietary software quite often. In school, they should not be permitted to impose proprietary software on students.

In the infrastructure of our voting system, they should not be permitted to use proprietary software.

For voting infrastructure, public school children's curriculum, I have zero tolerance for proprietary software. I'm not willing to entertain the idea that the opposition has a leg to stand on in this debate.

They can take their proprietary software and shove it!

Ryan Cartwright's picture

Terry, suppose that instead of you and I discussing software, people were talking about slavery and abolition. Suppose one person had just made the case for abolition of slavery and someone who were a bigot and an ignoramus (not at all like you) responded by saying something like this: "there are people who believe just as strongly in the opposite"?

From Answers.com

big·ot (bĭg'ət)
n.

One who is strongly partial to one's own group,
religion, race, or politics and is intolerant of
those who differ.

Bigotry is not limited to one side of any fence. Those who opposed slavery did not suggest the imprisonment of the slave owners without giving them the opportunity to change and neither should we.

Statements like the following one could easily be perceived by those you oppose as bigoted. I am not saying you are a bigot I am saying that in this context the label could be applied by both sides and that makes its use in an analogy worthless.

The answer is that those people who believe just as strongly in the opposite are wrong.

They may be but if you believe "Freedom is not imposed on people" then surely you have to allow them the opportunity to be wrong. Now some of us may believe strongly that--because of certain practices by proprietary companies--certain parts of society are not being given a fair opportunity to express their freedom of choice but if you are saying the answer to this is to impose free software on them then you are going a step too far in my eyes.

In your previous comment you list a lot of "if I were" statements, all of which suggest that you are not or have not been in those positions and so have the luxury of being able to make statements such as the ones you made. In reality to impose anything on a group of people will generally bring about resentment towards it (unless they agree with it). That resentment can breed dissent, which can ultimately breed revolution. It's something the music "industry" are having to learn with regards the removal of fair use by DRM. If you impose any key decision (such as what software to use) without either explaining to, consulting with and winning the approval of those you are imposing upon, you may well end up doing more harm than good to [y]our cause.

Terry is correct, if you want people to choose free software, imposing it upon them is not the way to win their hearts and minds. Educate them, coax them if you want to but if you remove all other options you will likely alienate them. This is a lesson the proprietary software world is having to re-learn and it is one the free software community would be foolish to ignore.

Paul Gaskin's picture

I was specific about my examples. I would not suggest that people in the private sector could legally or should have any software choices forced on them by regulation. Therefore, the business owner has the power to choose the software because its all happening in the private sector.

However, another specific example I gave - voting infrastructure, should absolutely not allow the choice of proprietary software.

Can you imagine why I might think so?

Also, another specific example - public school-children's computer curriculum. Proprietary software should not be accepted in public school for the same reason there should not be Coke and Pepsi machines in public schools.

The school consists of a captive audience of children who get preyed upon by mega-corporations like Microsoft and Coca-Cola.

The citizens of the U.S. are deprived of our freedom to vote as Americans by a proprietary voting machine company such as Diebold which has used proprietary software to steal elections for George Bush.

I don't have a say when some government contract is awarded to Diebold. I'm part of a captive market with proprietary software imposed upon me.

This is an issue of moral clarity in the case of our voting infrastructure. There are no valid arguments to be made for using proprietary software in our voting infrastructure.

I take your point about the literal meaning of the word "bigot", but I was using it with the connotation the word has acquired from its use in discussion of ethnic issues.

There is always someone who might say one thing or another, but that is not necessarily a good reason for one to doubt oneself.

A coherent mind does re-asses deeply held understandings every time a critic disagrees. That would be too reactive and disruptive to the integrity of a coherent mind.

I understand Terry's point about our internal sharing being more essential than our outward evangelism. My point is that we cannot cede the battle over market-share against proprietary software.

Ryan Cartwright's picture

I would not suggest that people in the private sector could legally or should have any software choices forced on them by regulation. Therefore, the business owner has the power to choose the software because its all happening in the private sector.

When I said "In reality to impose anything on a group of people will generally bring about resentment towards it (unless they agree with it). That resentment can breed dissent, which can ultimately breed revolution." -- to which I think you are referring here, I was not talking about legislatively imposing software choices upon business owners. I was referring to the practicality of business owners or their IT managers imposing such decisions upon their staff. Yes I know it is done all the time with proprietary software but IME that brings about the same resentment and dissent and eventual dislike of the software being imposed. If you are a business owner, you are of course within your "rights" to make such a decisions without consulting anybody. But I happen to believe business leaders have a responsibility towards their staff as well and dictats rarely go down well. Sometimes though, yes, the business owner has to make the tough choice but choosing free software is always about more than the bottom line and if you want your staff to "get it" then imposition of free software is not a good place to start.

With regards software choices in voting structures, personally I believe that technology and an anonymous but secure voting system don't mix very well. I am a luddite on this matter and would rather not have electronic voting machines at all. I am yet to be convinced that an electronic system can register that I have voted and keep the way I voted anonymous. But if we have to have them then yes an open process requires systems that are not proprietary.

I did understand the way you were using bigot but I don't think you can use it wth that meaning in this context, thus the only other definition we have is the one I gave. To make an analogy between proprietary software and human slavery might add gravitas to our argument but it also cheapens the true horror of daily life for those in real slavery. Proprietary software companies might lock users into contracts and attempt to keep their custom by foul means or fair but should the customer wish to stop using that company's products, they can.

As for ceding market-share, my views on this are well documented on this site. Indeed they sparked Terry's post. :o)

cheers
Ryan

Paul Gaskin's picture

The reason I brought that up is not to compare software freedom and or lack thereof with actual human slavery.

Today, software licensing isn't an issue of such obvious moral clarity as the issue of people owning other people.

Perhaps in the future, lack of software freedom may actually become a form of control used to impose real human slavery. Perhaps it is already happening in places like Iraq.

The reason I brought it up is to show that when the conversation is about an issue more obvious moral clarity than software licensing is today, the idea that there are two sides to an argument doesn't really change the obviousness that one side is ethical and the other side isn't for most observers.

My point is that you can always find people who disagree with even the most obviously righteous positions on ethical issues.

That doesn't mean those who disagree should be held in equal regard and we should be forced to debate them as if they were just as ethical as ourselves.

This is one of the major things which went wrong with the media in recent years which has lead to the terrible circumstances we see in Iraq.

Sane people were forced to debate against crazy people because corporate-media pundits presented two sides to every argument, even when one side was obviously deceptive, insincere, and unethical.

I don't think we should hold the proprietary software advocates up as if they are on equal footing in this debate.

The free software community should seek to prevail over proprietary software in the free market by non-regulatory means, and we should seek to have proprietary software pushed out of middle schools, high schools, and essential public infrastructure by government regulations.

Ryan Cartwright's picture

That doesn't mean those who disagree should be held in equal regard and we should be forced to debate them as if they were just as ethical as ourselves.

The idea that somebody should be held in less regard to ourselves simply because they disagree with our viewpoint is a dangerous one. Their argument may well be considered invalid or of less importance to the debate but that does not give us the right to disregard them for holding it. Rather do we not have a duty to explain why we find their argument is invalid rather that let them continue in ignorance?

The free software community should seek to prevail over proprietary software in the free market by non-regulatory means, and we should seek to have proprietary software pushed out of middle schools, high schools, and essential public infrastructure by government regulations.

As much as I'd love to see only free software used in such institutions I feel we differ on the idea that a certain viewpoint should be banned simply because we consider it "bad" for them. Again I would rather see them be given opportunity (which granted they are rarely given now) to discover free software and utilise it. I am not one who subscribes to the idea that the way to beat proprietary software is to leglislate it out. Legislation should not dictate which software is to be used, it can and should dictate that all options be fairly and openly considered within such bodies.

Paul Gaskin's picture

I never singled anyone out as a person. The people I referred to were only referenced by what they believe, not who they are, so that set of people is as transient as the thoughts in those people's minds.

Also, in the specific quote of mine which you referred to, I hadn't even specified what ethical belief they disagreed with.

Of course I never suggested "a certain viewpoint should be banned".

I'm in favor of legislation which prevents certain public infrastructure and public services from depending on private corporations.

I'm against privatization of the military, for example. I'm also against allowing proprietary software in the curriculum of public school children as I mentioned before.

The reason for this is because the proprietary software does not belong to the user, it belongs to someone other than the school children who use it. It thereby subjugates those children to the control of the private entity which owns the software.

Free software does not do this to school children, therefore it is like comparing apples and oranges.

I agreed with you when you wrote this:

http://www.freesoftwaremagazine.com/columns/can_we_please_stop_fighting_fud_fud#comment-76041

Ryan Cartwright's picture

I never singled anyone out as a person. The people I referred to were only referenced by what they believe, not who they are, so that set of people is as transient as the thoughts in those people's minds.

That's the point, you are dismissing people because of what they believe. You specifically said they (not their viewpoint) should not be held in equal regard because of their viewpoint. Irrespective of the view in question no free society can ever claim to be free if it treats those with a certain viewpoint as lesser members of that society. You may feel you were saying that their view should be disregarded but what you said was that they should be and it was that which I disagree with.

I agreed with you when you wrote this:

That was a long comment ( it's a habit of mine :o) ) but I guess you mean this bit in particular "I personally feel that the free software community does have to compete with the proprietary companies ... but it’s more about winning hearts and minds than revenue and it is in this field that my concern lies.". I still do believe that but the context is important, I do not believe free software has to or should attempt to compete with proprietary software on fiscal or market share grounds precisely because those grounds are dictated by the proprietary software companies. Free software should compete with proprietary software on its own terms and within its own strengths. Whether we like it or not a product which is not governed by unit cost can never compete in those terms with one that is -- it's pointless.

Paul Gaskin's picture

It is better in general to focus on the ideas rather than the people who are presenting them.

There are people, however, who've established a track record of malicious deception.

Even while it's considered bad form to focus on the messenger rather than the message, we still must hold each other accountable for cases in which a pattern of willful deception has been established.

Regarding your comment which I brought up - I read it wrong. I meant to say I disagree with it and I agree with the guy who said free software ought not to be compared to proprietary software because it's like apples and oranges.

Author information

Terry Hancock's picture

Biography

Terry Hancock is co-owner and technical officer of Anansi Spaceworks. Currently he is working on a free-culture animated series project about space development, called Lunatics as well helping out with the Morevna Project.