The massive failure of FOSS

The massive failure of FOSS

“The open source software movement has been one of the successes of the digital age” or so says Clay Shirky of New York University's Graduate Interactive Telecommunications Program in the latest issue of Harvard Business Review. Yes, sure, but he's just buttering us up.

Mr Shirky then goes on to mention the positive press Linux and Apache receive in business publications. Subsequently he then trashes the whole open source movement.

Mr. Shirky is writing in a piece called "Breakthrough Ideas for 2007". I'm not sure a small piece on open source is appropriate as a Breakthrough for 2007, maybe 10 years ago; and quite frankly I expected quite a bit more from the Harvard Business Review which in general is an outstanding business journal.

But there it is, Mr Shirky stating the enormous number of failures in the open source movement. He points out that has more than 100,000 open source projects open for collaboration and use and indicated that only a few of them have more than one hundred downloads and most have little activity... this he equates with failure. I.e, those projects that have few contributions or downloads can be considered failures. “A project was proposed but nothing happened.”

But he doesn't stop there. He then continues with the thought that “the vast majority of open source projects are failures” and the media must be wrong for emphasizing the successes of the movement.

If you are in a business that is considering open source software, according to Mr. Shirky you should run for the hills. “Open Systems are a profound threat not because they out succeed commercial firms but also because they outfail them.”

All is not lost however as he goes on to say that the cost of trying new open source systems is low so that the risk is low and therefore it can be useful. He goes to get lost in a social networking example and ends with the thought that if businesses can harness the volunteer efforts of developers who like to learn and experiment for experience sake... go ahead as it makes business sense to exploit.

I read this article several times and have this odd feeling it's some kind of obscure attempt at undermining the value of FOSS. It's obvious that this educator has not researched his subject very well. He needs to look beyond the narrow business journal headlines. I'll bet that he runs his laptop on Windows and has never heard of Ubuntu. He obviously doesn't understand that all those projects on SourceForge aren't a bunch of independent proprietary projects, but a collection of ideas, and innovation. The success is that there are 140,000 projects posted out into the open to stimulate possibility, dialog, collaboration and occasionally some useful code will get picked up. It's about sharing and working together and providing alternatives.

No mention of Joomla or Drupal, Word Press, Zencart or any of the 100s of successful and very useful open source applications, servers, desktop operating systems, mobile apps that have been used to help create many successful businesses.

As far as failures go, I wonder if he has looked at how often the large proprietary vendors have been sued by customers for failed enterprise systems developments, or how often Windows servers crash compared to Apache.

The Harvard Business Review panel of editors missed this one, me thinks!



Crosbie Fitch's picture

Yup. The difference is that we never see the vast swathes of failed or backburner proprietary software projects because they never see the public eye.

Probably more waste behind closed doors because there's less opportunity for collaborative enthusiasm to take the more useful projects forward.

Moreover, dormant projects in the public domain have less reason to be wiped off the slate as an embarrassment - you never know, they may be useful one day...

Chris Holt's picture
Submitted by Chris Holt on

I once, some 5 years ago, saw a report on failures by the big integrators and the number of lawsuits against them by states and countries and various other companies and it was an amazing and colossally expensive amount of failure...hmmm got to look for that report and see if there's an update...

Anonymous visitor's picture
Submitted by Anonymous visitor (not verified) on

Clay Shirky...Another person who writes uninformed nonsense about open-source. (SourceForge is an archive you MORON! Have you not understood why nothing gets deleted on there?)

You know what you do in cases like these? You ignore them.

You put your head down and code like you never code before. If you don't code, you help test or write helpful HOW-TOs or guides. Hell, help a beginner if you want!

My point? No use discussing stuff like this. It doesn't improve open-source software if we talk until the cows come home!

We don't need to defend open-source, because we let our applications do the talking. Those very applications speak far louder than any words of any analyst, report, etc. As long as we improve what we do, those folks can criticise as much as they want.

I think we should start an archive of critics. This way, when they try to backtrack in the future, we can refer back. (We can use the: "But back in Feb 2007 issue of Harvard Business Review, didn't you say...") :)

The best thing about revenge is that it doesn't have to happen straight away. You strike back when the time is right, when they least suspect it. :D

Anthony Taylor's picture

If most of the failed projects have very little activity, that means they are not very interesting to many people. They are essentially trial balloons of an idea.

A corporation has only a few people who talk, and so only their ideas have a chance to become failures (or successes). The Free software world is like a city full of people talking, with each idea becoming a seed. Some of those seeds take root and thrive. Others hardly sprout before withering. Just as thousands of acorns drop from a tree each year, only a few will ever take root and grow. Yet nobody claims the process is a failure.

The idea that anyone would bet a business on a half-started, inactive project on Sourceforge is ludicrous.

Chris Holt's picture
Submitted by Chris Holt on

Loved the acorn analogy Anthony... it's exactly right!

Anonymous visitor's picture
Submitted by Anonymous visitor (not verified) on

The man is confused in looking at software experiments as if they were heavily funded failed business ventures.

"My project is not getting attention." Same is repeated 10 times across Sourceforge. I and some of the other major players in these tiny projects research and make our way to a larger project that is doing similar stuff. Over time many skeletons are left behind but many of their ideas and their developers are contributing to a much smaller number of active successful projects. Small projects R&D help toughen large projects structurally or even work their way in as plug-ins, utilities, APIs, or in some other fashion.

What about the many applications in a typical distribution? Many are used very often. Some small apps are extremely useful to a minority (maybe to people that get paid 200K/year doing sophisticated work). If a distro with so much useful software for free is the result of a failure, I can't imagine what to think about Windows which comes with so little software in comparison. Microsoft is clearly failing.

What can't be missed is that the active projects, and there are many, "suck" up a large number of developers, end users, and many other contributors. Many people contributing are contributing to successful projects. In the proprietary world, if you find that your project is going to end up in the scrapheap or you can't manage to contribute effectively, your choice is to fake it or take an early leave from the company. Guess what most pick? Now *that* is a waste of money/resources because of not being able to freely move over to a project that is succeeding [but the family has to keep eating].

What about projects with little activity but that are mature and used by a lot of people (this is analogous to a business with little growth or R&D but steady income)? These can never be considered failures as long as they aren't completely superceded by superior products.. even if the activity level goes to zero. And in all likelihood, the last stable version of the project (maybe years old) is mirrored in many places resulting in downloads never seen from Sourceforge, especially true if they are a mature staple of the popular distros.

What about projects that never get much attention but they are very important to the creators? Here, any added help or additional user beats what would happen if working separately. At most, you can just ignore these and the FLOSS community would never be the wiser or hurt by the absence. Some apparently just fail to see that 10 tiny projects are better than zero. Certainly, 100,000 tiny projects are better than zero.

What about the Mythical Man Month? Many times it only takes a few to build good software.

I agree, this person seems to have little experience in the subject matter. Maybe next time we'll have an astrophysicist lecture on Hyacinths and their contributions to society in 12th century Europe.

I think some "management types" are feeling threatened.

Scott Carpenter's picture

I wish I could read the article. I usually enjoy Shirky's essays, but if that's his point--that the failures indicate some problem--then he's way off the mark. I agree with the points made here -- all those abandoned projects indicate fertile ground; compost to support future growth.


cel4145's picture
Submitted by cel4145 on

I wish I could read the article.

I wish the Drupal contact forms were enabled here. Send me an email via the contact form on my site.

Charlie |

Anonymous visitor's picture
Submitted by Anonymous visitor (not verified) on

I am not sure how you came away with the opinion you did of the article. Mr. Shirky is not presenting a positive or negative image of proprietary or open source systems -- only pointing out some of their qualities. In fact he is quite complimentary of one of open source's greatest strengths, which is that it greatly reduces the cost of taking risks. Not every idea or plan works out, but if you don't try and take risks you never get anywhere. He points out that proprietary enterprises have a disadvantage because they have costs associated with taking risks that open source greatly minimizes. And he says that proprietary enterprises may need to learn ways to close that gap if the want to compete with open source enterprises.

Henry Ford said, "Failure is the opportunity to begin again more intelligently." Open source projects have the advantages that they can "begin again" more frequently (see Boyd's Law) and because of their public nature they retain their intelligence better. Conversely proprietary enterprises have advantages, like their ability to attract capital and apply resources.

cel4145's picture
Submitted by cel4145 on

I'd agree that there is some overreaction here concerning Shirky's text. It's the defensive nature of the open source advocate responding to the negative tone of talking about open source in terms of failure. However, Shirky's made some good contribution to understanding open source with his analysis; it's just incomplete. I wrote about it over on my blog.

Charlie |

David Weinberger's picture

Yup, I agree, Anonymous. Shirky was pointing to a strength of Open Source, not criticizing it. Because the cost of failure is so low, Open Source developers are able to take more risks and try new ideas. The result are the many Open Source successes.

The idea that there's value in lowering the cost of failure is novel to most businesses.

Anonymous visitor's picture
Submitted by Anonymous visitor (not verified) on

Guys like this are easy to shoot down.

1. Did he use google today? Yep, bet he did 100,000 linux boxes faithfully
answering his request.

2. Does he own or ever heard of a tivo....probably has one, again powered by linux.

3. Did he get email today, probably delivered by linux and the name requests answered by linux dns servers.

4. Does he own a wireless router? ...probably powered by linux also.

5. Most of the web sites he viewed today where running on linux.

6. His isp surely uses tons of linux to keep his service available.

7. His network admins are probably using mrtg to monitor availability and load to keep his service at peak levels.

of course I could go on and on with that...

As for source forge, there are more uses for it than just downloading some software. It is the worlds largest open source code repository and if most developers are like me I go there constantly to pick out snippets that are useful to me.

Anonymous visitor's picture
Submitted by Anonymous visitor (not verified) on

The difference is that for free programs, "failure" doesn't quite mean the same thing it means for proprietary programs.

"Failure" of a free program means that some time, some effort has been put into the program, but in the end it was decided that there really was not enough interest in the program to see it through. The time and effort are not necessarily lost, however, since it can be taken up by others and completed or transformed.

Chris Holt's picture
Submitted by Chris Holt on

So I know that Mr. Shirky is giving some credence to the open source movement in that it is less expensive to experiment within that structure; but he frames everything in the negative and the negative permutates the article. He just misses some of the key points which commentators above are making and if you didn't know anything about FOSS you'd be a little confused.

Author information

Chris Holt's picture


Chris Holt specializes in consulting for Government and NGO public health and social services organizations about software to assist with case management and patient management systems.
Check out his site at