Slashdot Mirror
The Cathedral In The Bazaar?
replicant_deckard writes "This opinion piece I wrote to Open explains how dual licensing (simultaneous use of both GPL and proprietary license) works. Dual licensing gives you basically both the support of the community and a profitable Microsoft-like business model. Seems that this model used by MySQL and TrollTech is getting more popular. Now my question is where are the limits?"
This is a great way for software to develop.
It allows companies to sell closed source versions of software that they develop and it also allows the community to develop changes. If the community gets large enough it may even out pace the starting company, or if there isn't much community than the company will just get all the nice security patches that open source is so good for.
Paying taxes to buy civilization is like paying a hooker to buy love.
It only works for software that is a platform for other software, which quite frankly is not the majority of software out there.
For this to be sustainable, you need lots of other companies that like your platform but wants to create proprietary software. This also means that the Free Software Foundation might not like it too much, but that is probably something absolutely everyone cares about.
I've been working on an open-source GIS project for several years and have sold a few copies. I'd love to be able to support myself and a few employees in our venture, and until that happens, I'll just keep working on it, like a prospector searching for a vein of gold.
Personal Strap-On Aircraft for Auction on eBay
I think this is a great idea because you get the best of both worlds. You're able to sell your code to people who want to use it in closed source products AND if it is any good it gets used in open source products. The more it gets used in open source projects, the more companies there will be who come knocking at your door wanting to license it from you.
One of the main values of open source is that it provides a viable alternative to closed source solutions. You don't get stuck with whatever Microsoft or Oracle or Sun wants you to have. It also allows solutions to be custom tailored to fit a particular problem or situation. The downside to it is that it is often difficult at best to actually make money creating open source products. The people who make money are the ones who USE the products to achieve a result that would be expensive or impossible using closed source solutions. IN other words it is the customers of open source vendors who reap the biggest profits from open source.
When open source vendors dual license their products we all win. The vendors win, the open source community wins, and the companies that want to license it for closed-source work win. Everyone gets what they want and that is not a bad scenario at all.
Lee
Muslim community leaders warn of backlash from tomorrow morning's terrorist attack.
Are when you have to employ people on a contract so they can feed their kids.
Although the article is a bit (very) thin, it makes the important point that it is possible to do both (although it doesn't mention the NuSphere problems).
A point that may be relevant though, is that these were mature products which went GPL, i.e. they protected their IP until they had a sufficiently large user base.
Humorous signatures are over-rated.
Sounds dumber than the rantings of a drunken finn at the bar. The whole idea of cathedral and the bazaar is development methods, not copyright law. Certain licenses, "Open Source" or "Free Software" licenses, do help allow for this type of development. The author sounds confused, not insightful. Wonder why he got posted to the frontpage...
The catch, if you are in the business of selling software, is that you need to own the copyright to all of the pieces. You are fine as long as nobody makes a successful fork. If that happens, though, you won't be able to sell that version (except under the GPL).
So you need to keep the outside developers sufficiently happy that this won't happen. MySQL seems to have managed it. And various others. But you better have a large working product before you start. Some people have taken some BSD licensed code, polished it, and re-issued it under new license. That can work. It seems, to me, to be a bit unfair to the original coders (unless those are the people doing the relicensing), but it perfectly legit under the license, and the BSD folk say they don't mind.
I think we've pushed this "anyone can grow up to be president" thing too far.
If I write code and place it under the GPL, then the licensees are the ones who are bound by it. I, as the licensor, am not bound by it.
Lee
Muslim community leaders warn of backlash from tomorrow morning's terrorist attack.
Seriously, why presume that there are limits? There doesn't need to be any. Many grand things, good and bad, came from folks deciding that they didn't want to be limited.
The simple truth is that interstellar distances will not fit into the human imagination
- Douglas Adams
The dual licensing model only works, as a business, when there are customers who will pay to avoid the "viral" features of the GPL. If everyone was happy to use the GPL, there would be no business in dual licensing.
The problem, as I see it, is what happens when someone in the open source community contributes a significant module of code, rather than a small patch.
They've written a piece of code, probably in a separate source file, put the GPL comment at the top and sent it to you. Surely they, by law, automatically own the copyright for this code, and they've released it under the GPL.
So what right do you have to then go releasing it under your commercial non-GPL license?
It's always seemed to me that the most effective way to make a profit off your open source product is to sell after-install support, training, and so on. That way, complex products will always get the larger companies shelling out for these "extras". No, you don't make Microsoft money this way, but you can operate with exactly as many staff as you need to fulfill the requests, and expand (training internally, which is in itself training the trainers!) as needed.
I am a bit annoyed by the constant mis-reading of the Cathedral and the Bazzar. ESR was originally exploring why the Linux kernel (aka the "Bazzar") progressed so fast without any published plan, nor an explicit patch inclusion process, and chaotically open mailing list. This was compared with FSF projects like gcc (aka. the "Cathedral"). Who had explict project goals and schedules, an elite group of patch committers, and a closed mailing list.
CatB is not about Open vs. Commercial (usually closed). I do grant that Commercial is nearly always closed "Cathedral"-like development. But not always. The programming language Rexx was developed in IBM in a process like the "Bazzar". This worked because IBM mainframe community was pretty big and had good communications.
So MySQL having GPL and Proprietary licencesing policies, does not make it both Cathedral and Bazzar. It has nearly always been Bazzar-like (though clearly they had fixed committers and some planning about features).
Sheesh!
-- I am not a fanatic, I am a true believer.
Technically, the means two things. One is that the user can't license the product before he can change it (as he hasn't submitted anything yet), and the other is setting up a reasonable base line to determine whether the user has contributed "enough".
In truth, if there is a very active developer on a dual license project, then the company has every right to simply "gift" a license to that developer as a reward.
Now, whether that reward needs to be declared as income is another question entirely.
Isn't that the whole POINT of open sourcing something?
Here you have what's obviously a good lot of html code (I'm no coder, so I don't know the exact level to which the kde stuff is used) which has become good enough to attract the attention of Apple, a large developer. They've jumped in, are using the code (as they and anyone else are free to do) and made changes to it to suit their needs, and fixed bugs. They've put it back into the pool of open code, and anyone ELSE is now free to do the same as them, just starting from a base that's not only improved by having more hands working on it, but has had a stunning level of free marketing. A mindshare increase, code improvements, and more widespread use. To me that's the whole point - the good stuff survives and is made better.
Basically, if you don't want to GPL your programs, you have to pay for the privilege. The $$$ from sales of proprietary cygwin licenses are used, to some degree, to fund Cygwin development at Red Hat.
Your question is pretty subjective. "Stuff that matters" to me might not matter to you.
I read newspapers and their
--K.
Sig: Bad people happen. Try to avoid being one of them.
First of all, you can really only dual-license if there is something to sell. That means dual-licensed open source software needs to have one of the less "business-oriented" open source licenses, like the GPL. But for some software, the GPL just isn't appropriate. For example, for a GUI toolkit, if it is GPL'ed, that limits its adoption.
Also, with dual-licensed software, companies have other considerations in mind than just creating technically the best possible software or the best software for the users. They may want to try to dominate a particular market niche and exclude competition, for example.
I think in the case of Troll Tech, both of these apply. Companies like Sun and IBM have chosen Gnome because they did not want to hand the keys to commercial applications for their desktops to some small company that they have no control over. Furthermore, Troll Tech has a strong incentive to have something like Qt/Embedded take over the entire display--if people could use free or competing toolkits on something like the Sharp Zaurus, it would threaten their revenue stream.
I think dual licensing can make sense in some circumstances, but I think it's risky. If you are considering using dual-licensed software, I think you would do well to look at the license, business, and track record very carefully. Personally, I think for MySQL it works, but for Qt, it doesn't.
Dual-licensing works extremely well in one specific case of open-source software, namely that where there is no confusion about copyright ownership. If a project gets numerous submissions from external contributors, they either have to sign-over ownership, or the project has to fork. Neither are really satisfactory.
But for projects working ex-Cathedra, where external developers provide feedback and perhaps patches but no significant functional extensions, the dual licensing model can work very well indeed.
My company has made open source tools since 1995 or so, and we've now switched from a BSD-style license to a GPL / or commercial license model. Our motivation was (a) guided by Stallman's comments about BSD licenses helping commercial companies at the expense of open-source projects, which we agreed with, and (b) the desire of some customers for some kind of commercial framework.
Since 1995, we've found that there are very few people able and willing to contribute to our software, even though there are thousands happy to use it. Possibly because our tool are somewhat complex internally. So, the dual-license model fits our process perfectly.
Sig for sale or rent. One previous user. Inquire within.