Free software will win. Eventually.

Short URL:


Free software (eventually) works better than proprietary software; why?

Making dramatic statements always implies a need to "back" them (or "prove" them) with facts, data, statistics. However, a statement like "Free software works better than proprietary software" is so broad, anybody can prove it and disprove it at will. It depends on which angle you take, which area, and what your comparison terms are. However, I would like to add an important keyword to that sentence: "Free software works better than proprietary software". That easily-missed word shyly hiding in brackets makes all the difference.

That extra word --eventually--implies that really, it's only a matter of time before it becomes better.

Some background information

In 1995, when I introduced GNU/Linux to the company I was working for (which was also my first employer), the company's senior system administrator (called "Peppe", believe it or not) looked at it and laughed at me. "It's got some basic stuff, but you do realise it's a toy?". Peppe was right. He was used to AIX and SCO Unix, and back then GNU/Linux really was indeed a bit of a toy. The networking stack was there, but it was rather basic. The LibC (a basic library that allows you to actually use your system) had major limitations. The Linux kernel only ran on i386 processors. It was a toy--a fabulous one, but still a toy.

Fast-forward 12 years: Linux--the kernel--is an astonishing piece of software. Its firewalling stack is "stolen" (sometimes legally, some other times maliciously) by several major hardware makers (especially routers and modems). GLibC is a state-of-the-art library which supports multithreading. The kernel runs on... well, pretty much anything--in fact, it might soon power your toaster.

While Peppe was "right" at the time, he definitely failed to consider two very important parameters: "potential" and "growth rate". Or, more simply, he missed out on that crucial word: eventually.

How did GNU/Linux "get there"? An example: Linux firewalls

I mentioned that the firewalling stack improved so drastically in Linux - how did it get there? Well, it was definitely a bit of a bumpy road.

Firewalling became possible in Linux 2.0 thanks to ipfwadm. It was a bit of a revolution: for the first time, it was actually possible to run a firewall that was based solely on Linux. However, it was evident that ipfwadm was very limited and people were getting more and more excited about creating firewalls with Linux. When I heard that they were going to rewrite everything in order to overcome ipfwadm's limitations, I thought: they are crazy, they must have lost it. But, they hadn't: when Linux 2.2 came out, ipchains greeted everybody, a new revolution started in the Linux kernel. ipchains solved a lot of problems in terms of flexibility. However, that wasn't quite enough: there was yet another rewrite, called netfilter/iptables, which turned the "toy" into a fully-featured firewalling system.


My own experience: Drigg

When Digg came out, I absolutely loved it. I loved Slashdot and the fact that they would be just so picky about what went on their home page. However, Sladhdot didn't seem to have enough subcategories--it was a site for geeks after all. When I saw Digg, I realised how good their idea was, and how far that could go. I was also amazed that for quite a while there was no easy way to create a Digg-like site. The first attempt to create a Digg clone was the Spanish site Mename, which was in Spanish only. Pligg sprouted from Mename (whose author was kind enough to release the source code under the GPL); phpdugg joined the army. We ran FSDaily using Pligg to start with, but while we were immensely grateful to the Pligg developers for their work, we wanted something different--something based on Drupal. So, I sat down and I developed it. I would have never been able to develop Drigg without the experience I gained using Pligg and without the solid base of Drupal (which I now more or less inside-out programming-wise). Drigg took a lot of development hours. I based it on a karma module I had written for Free Software Magazine. FSDaily ran on the first beta of Drigg--beautifully. Since then, I have more or less rewritten it (splitting it into separate modules) and release it to the broad Drupal community. The first version of Drigg I wrote at the very beginning was a bit of a disaster: it was the original Karma module on steroids and a fattening diet. I changed it, improved it, fixed the countless bugs reported through the fantastic Drupal system. The current version of Drigg probably won't win any "efficiency awards", but the code is solid--right now, there are no open bugs in Drigg.

Drigg got there.


Free software: the unstoppable revolution

People who work and live in specific markets always have a terrific insight on how big a market is going to be. In 1999, my neighbour told me that Western Australia was about to start booming like anywhere else in the world, thanks to the mining sector. He suggested I buy as many houses as I could regardless of the debts and the mortgage. I bought my house, and then witnessed the craziest boom in Australia's history. My neighbour knew what was coming because he worked in the industry. He told me that all of his mates knew it, and that to them it was so obvious, it was odd that nobody else knew. I got confirmation that if you eat and breathe mining every day, and you have intuition, you will be able to predict what the next step will be. He did, and got it right.

So many people who eat, breath, live free software today know that it's unstoppable. Microsoft's executives, locked in their corporate rooms, didn't have the means to realise free software's momentum. When they suffered from the effects (Linux was being installed where Windows NT should have been), they started fighting back--clumsily, noisily, ineffectively. They didn't have the insight to figure out what was coming from the free software world.

They were growers who were meant to predict the mining boom, but--unaware of what was coming--found themselves more or less suddenly with no workforce available (the mining sector took the lot).

Ubuntu's (and GNU Linux's) bug #1 -- Microsoft has a majority market share will get fixed.




Terry Hancock's picture

The real reason why free software is destined to take over is that it is highly conservative. Once the community figures out how to do something, then it knows how -- forever. That's because we share the information.

An application that came out a little too early or only served a certain niche because it was missing some crucial element may very well sit unnoticed for years. But because it's available under a free license and you don't have to go asking permission or re-inventing the wheel, somebody can just pick it up and run with it. It's this enormous library of freely available material that makes that possible. With free software you really can "stand on the shoulders of giants".

The proprietary software model throws a wrench in the works by sequestering the enormous investment of labor under private ownership and secrecy. Information is lost. If someone else wants to write the next advance on the state of the art, they first have to re-invent the state of the art -- which is HARD.

So the proprietary industry is constantly being thrown back to the starting gate, while the free software world just keeps on plugging.

The only reason why there's any competition at all from the proprietary world is that the ability to earn enormous profits makes high levels of capital available, and thus it is possible to run really fast towards a solution by hiring lots of programmers and paying them enough so they can work full time on the project.

But there are hard limits on that kind of development. There are some things you can't build in a day, no matter how many people you can afford to hire. And even when it's feasible, it's wasteful. The proprietary industry literally throws away billions of dollars inventing things that are already invented.

It's kind of a hare-and-tortoise race. It's not as sexy being the tortoise, but at least you win.

Albertde's picture
Submitted by Albertde on

My experience leads me to agree with you. All complex software tends to suffer from bugs either from programming errors or from unforeseen interactions. In the proprietary world, there is no real incentive to kill the bugs. I remember once working (briefly) as a technical writer in a software development firm when a programmer suggested that they do a new release that would just clear the accumulated bugs in the software rather than develop new bug-ridden features. The project manager said no way. Customers are unwilling to pay for a new release that would just eliminate the bugs - the software was supposed to be bug-free, they were only willing to pay for the new features!

RedRat's picture
Submitted by RedRat on

While I would like to agree, and more importantly, hope that what you say will come to pass, somehow I must disagree. If the question were "Can Free Software be Better?" then I would say a resounding yes! Why? Because proprietary software needs to be multi-functional, i.e., it has to appeal to the widest possible market in order to make bucks for the software company writing it. Free software can be very simple and therefore efficient in that it zeros in on a specific problem (after all that is why the programmer had to write it).

However, other than ego, what is the programmer's motivation to extend or improve his program? Once he has written the program and it does what he wants, there is little reason to extend or modify it. He has made it free so he has no reason to do this. I will admit that there are people out there who might say I want to make the best operating system and make it free (Linus Torvald for example), his motivation was a realization for a need. Perhaps all programmers are that way, at least to some extent.

I am relatively new to Linux (actually I got started back in the early Red Hat days, then fell away until just recently). I have been looking for a reasonably easy video editing program for Ubuntu but have found most are limited or just plain don't work. I finally found one in Kdenlive but can I be assured that if I find limitations in it, they will be fixed?

Terry Hancock's picture

"The" problem with this argument is the "the"! :-)

You state things as if there were only one programmer who can make these changes. And proprietary licenses indeed force something like this to happen (at least they make the pool of people who might work on the code very small).

But with free software, ANYONE can make those extensions. Odds are somebody needs the same feature you are missing and has the knowledge to fix it so that it works. And of course, if no one needs the feature, then it probably wasn't that important. So, you see it's not necessarily about motivating the original programmer to do something he otherwise wouldn't be interested in -- it's more about getting the people who are interested to contribute the necessary code.

Free software is user driven because it's user written.

The only cases where this gets weak are the ones where the users of the program are almsot entirely people who can't program. That has to be pretty extreme though -- surely most artists don't program, yet programs like Gimp and Inkscape are doing just fine. Because there *are* some people who are interested enough to write that kind of software.

There is, of course, one very noticeable kind of sequestering that does go on in the free software world, and that is the boundaries based on programming language and libraries (e.g. the Gnome/KDE rivalry or Python/Perl/Java, etc). So steps towards resolving these conflicts are important (e.g. cross-platform libraries, automatic translators, foreign function interfaces, etc). But this is so much less limiting than the proprietary industry, and there often are technical solutions.

amazingkip's picture

It's all a matter of time. Proprietary software cannot compete in an open world market. What proprietary software companies need to realize too is that there is a ton of money in the open source market; look at the $800 million acquisition of MySQL by SUN. $800 million for an open source company? That's huge. Check out, they talk a lot about how the Linux world is like the Japanese automaker, able to produce more for less.

Author information

Tony Mobily's picture


Tony is the founder and the Editor In Chief of Free Software Magazine