Milking The GNU

The FSF freed Software. What about the people who wrote it?

Milking The GNU is using Red Panda a new kind of (Free) News Browser for Firefox.
My Photo

About

Stay Tuned

  • Search about open source NEWS
  • SUBSCRIBE to this blog's feed

Quick Jump

  • Latest
  • Humor
  • Open source references
  • Patenting issues
  • FOSS Business models
  • Companies fighting the GPL
  • Liberty Bell
  • Milking open source
  • Miscellaneous

Good Readings

  • ZDNet Open Source
  • 451 CAOS Theory - open source
  • c|net Blogs: Matt Asay's
  • Linux Journal
  • Dirk Riehle's Software Research
  • Digital Majority
  • Linux.com
  • Building FOSS for Society
  • Law & Life: Silicon Valley
  • RedHat Truth happens
  • Open Source Law Blog
  • NonProfit Open Source Initiative
  • Scott Mace's
  • Ryan Tomayko's
  • Fabrizio Capobianco's
  • Ars Technica
  • Roy Schestowitz's Boycott Novell
  • Roberto Galopini
  • Groklaw
  • Jerome Boyer's agile archi.

Archives

  • November 2009
  • June 2008
  • May 2008
  • April 2008
  • March 2008

eos-james-dixon-pentaho


This page contains detailed comments about the post Equitable Open Source:

Hello James, thanks for talking the time to comment.

Disclaimer: I am a founder of Pentaho - business intelligence software using commercial open source model

First I am glad that there are people out there with passion about this topic.

I agree that is necessary to agree on a goal if you are going to do this. I think that if an commercial open source company is unfair to its community members, they are likely to shift to another community where their contributions are welcomed. This is also the case for all open source projects: the administrators and core developers of an open source project can be equally unfair (in my experience).

Well, you're assuming that forking is easy and that it plays its role in regulating relationships between communities and open source companies. But short of a handful of examples, it's not that simple at all:

  • Most forks fail. So the process in itself is not easy or obvious.
  • Key project developers are often employed by the company in question.
  • Often Trademarks and sometimes patent or license issues add to the confusion
  • Dual-licensing models  prevent another company to step in and to take care of the community (when such kind of symbiotic relationship has developed)
     

Can I assume that it is not a goal to drive commercial open source companies out of business? I hope so.

Glad to see  the community is not the only one getting paranoid :) No, indeed one of the goals is to make the relationship more balanced.

If this is the case then things that constrain their ability to make money from their commercial customers or their ability to increase their commercial valuation should not be included.

Well, things are evolving. There are plenty of different business models out there;  tomorrow will see plenty of new experiments as well.  I'm sure some models will be preferred but that doesn't mean that one must  be chosen against all the others.

I see a couple of issues with this whole thing:

1) Why judge commercial open source projects differently from non-commercial ones?

Because the goal is not to label/assess projects per se but companies. Ideally, large companies contributing to large open source projects will get their EOS label too.

To be honest I have found it easier and more satisfying making contributions to JBoss than to Apache. Some open source projects make it almost impossible to contribute. You could take a look the rating system that Bernard Golden's describes in 'Succeeding With Open Source' as a potential measure of all projects.

Once more the goal is not to assess project but rather the relationship between OS vendors and their community.

and much more importantly: 2) Why are you focusing on the commercial open source guys? Why not the free-loaders? You've found an easy target so you've stopped looking for better ones. There are thousands of commercial companies around the world who profit from all open source: they provide services, support, training, certified stacks etc. Search Google for 'Apache support' and look at all the sponsored links. Some of these companies do the right thing and employ full-time committers to contribute back to the projects they profit from, but many or most of them do not, and even the ones that do spend a very small fraction of their income on this. If you contribute to Apache many, many people profit from your work and contribute nothing in return. In contrast the commercial open source companies have full time engineering teams dedicated to writing open source software, they have full-time community facilitators, they host events and awards to honor open source contributors, and they create lots of lots of open source software in the full understanding that a very small fraction of the people that use the software will ever contribute as a paying customer. The difference with commercial open source is you know the people trying to profit from it. Come to OSCON, JavaOne, or LinuxWorld and you can meet them.

Actually this rant (you've got to admit this is one) gives even more credit to this EOS project.

Not only  do you sound incredibly defensive but you're also forgetting the role of those zillions of open source support companies. There are many ways to contribute. Hiring commiters is only one way to contribute and even then the practice can even be  pushed too far (if a company ends up owning all the key developers, the community ends up having little control).

There is nothing wrong with small companies offering support, training etc.  Doing  so they contribute to the diffusion of open source. Many of them would not have the means to hire  contributors anyway. And without those companies, forking would be even harder. Just like bigger companies, they have their place  in the FOSS eco-system. Qualifying them of  free-loaders is really missing the point and quite unjustified.

It is true that commercial open source companies are typically for-profit businesses. If you think businesses have no right to make money from open source then you need to fight to get all open source removed from the internet.

You or your company must really feel bad if the mere suggestion of giving an Equitable Open Source label to companies is triggering the You're against people making money stance! At this point this post is an offer to talk about the issues. No reason to feel threaten.

The reality is that many businesses make money from open source. Even more companies profit (by way of reduced costs) from using open source. This is all fine. Many contributors to open source projects work in IT departments that are trying to use the software to cut costs. This is how it should be as they are the people with the use cases.

You seem to have fixated on the wrong target. I suggest that people should only contribute to open source projects (commercial or otherwise) that they find valuable. If a commercial open source project is useful to you why do you care what % of their total IP it represents? It is either useful to you or not.

You know that's funny but when I come to think of it, the exact same arguments are employed by companies that feels threaten by the mere mention of fair trade or equitable commerce. I'll give you that open source developers are not 11-year old kids working in factories but the fact that the exact same defense is argued comforts me in this idea that EOS is a concept much needed.

If everyone contributes fairly to the projects that they find the most useful, natural selection will take care of the rest. In the mean time please consider focusing your efforts on those silent free-loading, profit-making companies that only consume open source and contribute nothing.

Why do you care so much about free-loaders? Your Natural Selection principle should take care of them, right?

I hope this did not come across as a rant. All the best,

All the best too. But yes, you did rant.

James Dixon, Chief Geek, Pentaho

Posted by: James Dixon | May 19, 2008 at 11:17 AM

-----------------------------------------------

Thanks for taking the time to reply to my non-intentional rant. My responses:

Well, I'm happy to discuss the issue, that's the whole point.

If your intent is purely to focus on companies then I suggest a change of name. The term 'Equitable Open Source' does not indicate this intent to me as it could apply equally well to projects as well as companies. I think that was the basis of my initial reaction: why take something generically useful (a badge to recognize open source projects that do things right) and only apply it to commercial companies?

I like the term "equitable" in that it is nicely reminiscent of fair trade. Fair trade  and Equitable Open Source  are both about  qualifying the relationship between the producer and the consumer/end-user.

When I talk about community members deciding to leave I am not talking about forking necessarily. For example if you don't like the way JBoss treats your contributions to JBoss AS there are other application server and servlet engine projects you can go to. This is the case for all the main commercial open source offerings that I know of.

Yes you can but it would have been better to have a global appreciation of the relationship between the company and the project BEFORE to write those hundreds of lines of code.

In regards to the 'paranoid' and anti-money statements there is a very vocal minority that have extreme views about the commercialization of open source. In order to remain balanced an effort like this needs to avoid being influenced by, or associated with, extremists from any side. I think you would need to be very explicit about that otherwise the credibility of it will wane.

I think this initiative is largely orthogonal to the fsf vs. OSI conception. Depending on the case it might well be that using a specific business model or a specific license enhances or hinders the relationship to the community.

I also agree that many of the small companies offering support etc for open source do participate in the advancement and proliferation of open source. I agree that there are many way to contribute (in the Beekeeper model I talk about 12 main ones).

A very interesting paper indeed.

But I have come across companies that use open source internally and in their products but which forbid their staff from contributing. I have also come across companies that embed open source into their products and yet who publicly warn about the risks and dangers of open source. This doesn't seem right to me.

It isn't and I'm sure that an EOS label would have to address those issues.

The free-loaders don't concern me unduly (this is what karma is for). I'm just pointing out that, in my opinion, if someone is going to spend some time and energy by focusing on a group of companies to ensure their behavior is fair by the standards of open source
communities I see a bigger bang for the buck by looking at those companies who consciously contribute little rather than those who contribute a lot. But, after all, its your time and energy.

But that's the idea. Remember the label is community centric. If a company has a negative relationship with a community/project there is nothing that prevent to have a label as well. A sort of thumbs-up/thumbs down albeit with a scale in both cases for instance.

I think the critical issue is the goal.

I agree, it has to be defined precisely before to proceed forward.

Many commercial open source companies have to strike a daily balance between adoption and monetization. They need the adoption of an open source community (individuals) and they also need the business of customers (companies). This balance changes over time for any given company. I don't think it will be effective to judge companies at different stages in their growth on where they are striking the balance.

I understand the conundrum but the problem is that of the company. A good EOS score might prevent a company to switch to a business model which otherwise would have been much less favorable to the community.

I think that openness, access, licenses, relationship with the community and relationships with other companies are valid measurements. The other issues you list are just part of the business model in some cases and I don't think its your intent to judge based on that (or is it?).

No, the business model is just a part of it. But its impact on the relationship to the community is important. The issue is also largely centered around control. Who ultimately controls the project that is supposed to be controlled by the community?. Or if you want to have a more pony tailed opinion :) what freedom does the community really have? What freedom do the contributors have?

James Dixon, Chief Geek, Pentaho.
Posted by: James Dixon | May 20, 2008 at 11:00 AM