Towards A Real Business Model For Open-Source Software

[ModplanMan]

This topic has gone wildly astray, but meh, I'll dive in.

Being able to choose between open technologies and free software and proprietary software is barely even a choice if you want to maintain a completely free and open market for software and ideas.

Choice isn't just about how many players work with x codec, it's about allowing the creation and maintenance of choice in situations where it hasn't previously existed or is becoming limited, in which free software and open technologies inherently encourage. Proprietary software doesn't - it helps limit and reinforce the decline in choice, and helps further the creation of new monopolies where neither businesses both new and old nor users are able to break it - certainly not within a reasonable time frame to not be harmful to both.

A society that allows one individual to encroach on another individuals freedom is not a free society. To stop that from happening you create rules and laws that stop people stepping over those barriers, but irrespective of those laws everyone has a guaranteed set of freedoms. To define freedom as entirely about choice is a myopic view of freedom.

[o0max0o]

This topic has gone wildly astray, but meh, I'll dive in.

You're right, we're astray but it's too fun.

A society that allows one individual to encroach on another individuals freedom is not a free society. To stop that from happening you create rules and laws that stop people stepping over those barriers, but irrespective of those laws everyone has a guaranteed set of freedoms. To define freedom as entirely about choice is a myopic view of freedom.

No, I am sorry. Freedom is freedom, independently of the social pact chosen by a group of people.

By your definition no society is free (not America, nor any of the European countries) since only a few hundred people decide what the rest of the population is free to do or not... that's not it, they decide what is legal and what is not!

Again: freedom is unrelated to social pacts and government. You are free only if you are free to disrespect the law: in the same moment the law becomes an absolute barrier to your freedom of choice you are not free anymore (in the pure sense of the word).

[kraftman]

I believe in people's freedom, you believe in what you call code freedom. I believe that Open Source can be competitive even with permissive licenses, you don't.

You believe in some developers and companies freedom. In few areas it can be competitive even with permissive licenses.

If there has been no pro-GPL article that might tell you something about the editorial choices (until now), or about the fact that it is so understated that it needs no further explanation.

Not the same things. Believe me, there are many people who don't understand a thing and they are able to believe others in everything they say.

How does the GPL guarantee that there will be contributors?

If someone uses GPL and shares he must be a contributor. It's, so simple.

In order to contribute you have to agree to the license, if you don't you can't contribute.

If you don't agree you're not using GPL or you're braking the law. Stop playing, because it makes you to look stupid.

So the basic assumption the GPL makes is that there are other developers that agree with it. The same assumption that other licences do, without crippling anybody's freedom.

Playing strawman now? GPL makes there are contributors and BSD do not guarantee this. GPL also gives the code freedom, because it mast remain available for community and BSD does not

Now, you might think that you would look like an idiot, but what if I choose to release my code in such way, consciously, and I am well aware that somebody might make a commercial product out of it? What if I am happy with it? How is it wrong? It's my choice.

I don't care about you. I was talking about community or developers, companies who want to compete with others and who weren't be happy.

I understand what the GPL is and what it means

. I just hate that anybody who supports the GPL thinks those who like to use other licenses are complete idiots.

Not the ones who like to use other licenses are idiots. I explained it didn't I?

So what, I am not talking about MS or Apple, I am talking about people, contributors. That's the difference. You believe everybody out there is just trying to fool us, I believe there are people that will help no matter what.

Good for you

A distribution license is a choice, just like any other. So why should I be "the idiot" for making a different choice?

You'll be an idiot if you want to compete with MS or Apple and you decided to use the BSD license, so MS and Apple can just take all your project advantages.

He does when he releases code under the GPL. If I want to contribute I have no choice but to agree with his view of the concept of "software freedom".

If you want then you're not forced.

[kraftman]

because it mast remain available for community and BSD does not

must* of course.

[perpetualrabbit]

Why linux is not a success on the desktop?

Why is Linux not a success on the desktop?
------------------------------------------

* To begin I would like to counter with saying that for many people Linux _IS_ a success. I gave my parents an older computer, and put ubuntu on it. It does anything they need: web, internet banking, word processing, printing, burning CD's, etc. They had a windows computer before, but it was so slow and so full of viruses, it was just frightening. Then they got a macbook, but after a few years it broke down, and was out of warranty. On their pensions they cannot afford another Mac. So Linux is a big success for them actually, and in no way are they held back in doing the things they want. In fact, they can do lots more then they ever could with windows or the macbook.
* Millions of people around the world use linux as a desktop system, but they are hard to count. How can you count them without clear sales figures, right? Only statistical methods apply, but they are unreliable.
* Linux is not pre-installed on computers like windows or mac, and 95% of people (statistic made up, meaning only the vast majority) will never put a new OS on their computers. This means that most people never get to choose to install Linux. The only way most people start using a new version of Windows or OS X is when they buy a new computer. Half of the people even never update their OS within a major version, i.e. they do not install Microsofts' service packs. With OS X this is somewhat easier, and Mac users will update more often, but still I bet that over half of them have no plans to run Snow Leopard. The only way that Linux will ever be mainstream is when it comes with the computer you buy. Only 10% or so will ever consider doing anything else than what other people are doing. We are a species or mostly stupid sheep, followers, don't ever forget that people.

But Linux is doesn't have many games, proper open source 3D drivers, multimedia gadgets etc.!
-----------------------------------------------------------------------

* The graphics and sound infrastructure in linux is a big mess. Way too many developer groups not working together enough is the biggest problem, followed by not having access to necessary documentation for programming the hardware.
* Here I think is an opportunity for a concerted effort by a few big teams of experienced programmers, with proper funding and a good plan. This will need to be funded, because people like that are scarce, therefore not cheap. Linux needs something like OS X corevideo, coreaudio and coregraphics. The many little projects there are, like for instance Gallium3D, DRI, and ALSA, Pulseaudio and all the others that overlap or leave gaps need to be consolidated and work together. Following a good plan with clear milestones, and with proper funding.
* To get there, the Aladdin license it not such a bad idea at all. You either pay for the latest and greatest driver infrastructure that works for the latest and greatest hardware NOW, or you wait half a year, or one year, but not more, and you get it for free under the GPL. I find that very reasonable, I having spent say 300 euros on a new videocard or high-end soundcard, I would not mind to spent say 50 euros on the driver infrastructure. Not at all. And neither should you mind, or you should just use the slightly older version. The half-year (or so) delay of releasing the code under GPL might also be enough to get nvidia to give their hardware documentation under NDA.
* When important driver infrastructure and software libraries for sound, graphics, video etc. are in place, games are easier to port to linux, so they will come. Linux can be a gaming platform. The video library mess and audio mess are reasons given by Adobe why Linux still has no proper working 64 bit Flash plugin. The one there is now really sucks, as you all know.
* There are other areas lacking in Linux land: wifi drivers, printer drivers, easy color management support (it exists, say lprof, but it is not easy like in a mac), gimp not close to being a photoshop, compatibility with iPhones, digital cameras, lots of other gadgets. These things could also be developed in the same way, or by the same as yet imaginary company.
* Once gaming comes to linux because the infrastructure for it is there, Linux is mainstream. Gaming systems get used as normal PCs by the parents of those gamers, say, and Linux will gain more and more users. Once it gets market share, all the stupid little gardening, bookkeeping etc., meaning the tens of thousands of applications that windows has but Linux lacks, will come too.
* And with becoming mainstream the viruses and other malware will come to Linux too. Linux may be less vulnerable a priori, but that only means that malware creators for Linux will need to work harder, be more sophisticated. With mainstream use come lowering of the collective IQ of the Linux users, and the stupid masses are the targets of the organized crime that are after the identities, banking passwords, money of those masses.
So be careful what you wish for.

Patents will destroy Linux/open source long before it becomes mainstream!
-------------------------------------------------------------------------

* They will not. They did not destroy Apple, Microsoft, Sun, IBM etc. Worst case it means that you will need to pay royalties for using a distro.
* There almost certainly is going to be a fight, with Apple, Microsoft, this `mpeg-la' on one side, and Google, Redhat, maybe IBM, Canonical, Redhat on the other. Oracle hopefully sides with the bright side.
* With this Bilski lawsuit, and lots of criticism on the patents systems of USA and EU, there is bound to be a reform in that area too. If only because big and upcoming tech producing countries like India and China are never going to abide by these patents. Asia generally does not respect copyright, let alone software patents. And they shouldn't. The same goes for copyright on music and movies. Sooner or later the corporates are going to loose, because sooner or later there will be a system like bittorrent which cannot be traced back to an individual or IP number, will be completely encrypted, cannot be blocked because simply working at port 80. A system that can only be blocked when shutting down the internet completely, or forbidding all encryption. Some file sharing system that is totally dark and can be used with total anonymity, not distinguishable from normal ssl traffic (to https:// sites). It will come if the MPAA and RIAA will keep doing what they are doing.
* So the market will change and a sane way of getting pay-for content to users and still get paid. Like some kind of flat fee system that does not exclude the little indie record companies. This will make DRM and things like h.264 obsolete sooner or later.

In the mean time, it is really important to get the Linux software infrastructure for 3D graphics, video and audio in a good state. At least as good as Windows and OS X, preferably a lot better. It is possible, with good leadership and proper funding, and a good plan. Not even that much funding, but still probably somewhere in the tens of millions. And a strong legal arm, to shield the programmers from patents lawsuits and stuff like that. It can be done in a year or two, and I think Google, Redhat, Canonical, Novell, IBM, Oracle and more are the companies that will gain enormously from it, and should therefore fund this idea.

[timrichardson]

Why is Linux not a success on the desktop?

No one has challenged the definition of success. The original author seems to equate success with making lots of money. By this definition, Linux is a complete failure, but if you have to wonder if software based on copyleft open-source licensing would even consider this to be the applicable benchmark.
Based on the criteria of providing a modern operating system which is open source, Linux is not only a success, it has completely dominated its competitors. Linux market share is tiny, but so is BMW's market share. Both are products that skilled users go our of their way to use.

People contribute to copyleft opensource software for lots of reasons, but one of them is that they see a system where co-operation is guaranteed: where they have their contribution exploited. Subscriptions models are not open-source and don't encourage users to become developers. Don't call it open source. It's not very unusual actually; lots of products have "community" versions which are open-source and enhanced versions which are closed source. I personally would steer away from such products if there was an open source equivalent; open-source projects only need a couple of good developers to do amazing things, and I'd much rather support real open source. But that's just me.

In the app-store world, we see that a lot of small developers can make a good living by selling a $2 app and it seems that open source is not making much impact here. For niche products, open source may not be the way to go.

[nanonyme]

If someone uses GPL and shares he must be a contributor. It's, so simple

This is a bit unclear.
If someone shares the original piece of GPL code, they are redistributing. (covered by license)
If they are making derivative code and sharing it, they are contributing. (also covered by license)
If this was what you meant, I agree.

[ABBY]

This is a really excellent read for me. Must admit that you are one of the best bloggers I ever saw. Thanks for posting this informative article.

[Hoodlum]

I'm responding to the article rather than the comments.

The business model seems viable. I think it certainly has potential for profit. However it does not seem desirable if software freedom is a concern, for several reasons:

Firstly, the general point that you are moving away from software freedom to create this model from the very start. While BSD licences give you more absolute freedom (imo) this model would create a situation where the practical freedom (of the code, users etc) is less. A majority of the interesting development would be closed. Closed development processes and reduced transparency. This would certainly add a new dynamic to the BSD vs GPL debate (which I personally don't care about, just making the point).

Secondly your suggestion of the code being closed for years; this results in closed solutions being unavailable to many and free alternatives being developed before they are opened, massive redundancy. There is already a lot of redundancy in FOSS, we don't need more. This presents other problems too. For example after 5 years the newly opened code, even if it wasn't obsolete, would be dumped out as a "finished" project. The code released would be completely foreign to everyone outside of the closed developers, who have likely moved on to other projects. This prevents you taking advantage of the beneficial aspects of open development. This issue alone has killed otherwise excellent projects in the past.

Reading the article I get the impression this is a band-aid fix for a problem inherent with open source development. I applaud your attempt to improve the situation, especially for smaller projects such as the BSDs and for them maybe it would work. For bigger FOSS projects, with rapid development I don't see this as viable. Free alternatives to any closed project would be created and improved before the code could ever be released. If the BSDs became more successful this same situation would probably occur there too but then again, maybe that's enough? Just to get into a position where the service model is viable for you?

[Hephasteus]

In the last 4 quarters, Red Hat brought in $750 million in revenues, Microsoft brought in $60 billion: that is almost two orders of magnitude more. Red Hat isn't some new startup either, it's been around 17 years. If there were so much money in open source, they'd be making it by now and grabbing huge market share.

Ya ya. Microsoft had 23 billion cash on hand at the start of the downturn. Yet it still goes to the market and bonds out 3.5 billion dollars last year. You know what the closed source model has that the open source model doesn't. Big fat lies. Until you grab micorosft by the throat and shake it till it fears for it's own life it won't tell you the truth. Until you actually learn what deception is and how it's employed. You will never understand how workable business models fail while unworkable business models succeed. You read charts and study data that is lies every day of your life. It doesn't make sense and it will never make sense till you learn to call it out.