Slashdot Mirror
There is No Open Source Community
porkrind writes "There is no Open Source Community is an Onlamp article about the economics of open source and how most people get it wrong. Really, open source is much more about supply and demand than it is about an activist community or individual drivers (individuals or individual companies) affecting change on society." From the article: "Taking the position that individuals have pushed open source forward leads to the conclusion that a core group of ideological 'believers' is necessary for the continued success of open source software. Businesses unaware of the falsehood of this claim are too afraid of running afoul of the 'open source community' and sometimes make decisions that are not in their financial interests. Both open source-based and proprietary software vendors need to challenge these assumptions."
I should assume that the autohr is trying to destroy open source. If everybody went with their economic interests, there would be no open source.
"Taking the position that individuals have pushed open source forward leads to the conclusion that a core group of ideological 'believers' is necessary for the continued success of open source software."
There's a Non Sequitur right there in the summary; just because an individual may have pushed open source forward in the past does not imply anything about future need.
Contrast this with saying "an individual pushed the invention of a wheel forward, leading to the conclusion that a core group of ideological 'believers' is necessary for the continued success of the wheel" and you see the flaw in the reasoning.
Tell your friends about xenu.net
Here's an anecdote from Richard Stallman.
At a trade show in late 1998, dedicated to the operating system often referred to as ``Linux'', the featured speaker was an executive from a prominent software company. He was probably invited on account of his company's decision to ``support'' that system. Unfortunately, their form of ``support'' consists of releasing non-free software that works with the system--in other words, using our community as a market but not contributing to it.
He said, ``There is no way we will make our product open source, but perhaps we will make it `internal' open source. If we allow our customer support staff to have access to the source code, they could fix bugs for the customers, and we could provide a better product and better service.'' (This is not an exact quote, as I did not write his words down, but it gets the gist.)
People in the audience afterward told me, ``He just doesn't get the point.'' But is that so? Which point did he not get?
He did not miss the point of the Open Source movement. That movement does not say users should have freedom, only that allowing more people to look at the source code and help improve it makes for faster and better development. The executive grasped that point completely; unwilling to carry out that approach in full, users included, he was considering implementing it partially, within the company.
The point that he missed is the point that ``open source'' was designed not to raise: the point that users deserve freedom.
Spreading the idea of freedom is a big job--it needs your help. That's why we stick to the term ``free software'' in the GNU Project, so we can help do that job. If you feel that freedom and community are important for their own sake--not just for the convenience they bring--please join us in using the term ``free software''.
Please help publicise swpat.org - the software patents wiki
Quo usque tandem abutere, Nimbus, patientia nostra?
But remember, if it weren't for these patents and their precious attached IP there would be no progress at all in the software field!
---- Take the Space Quiz!
This article irritates me in the way that most news media coverage irritates me: they purposefully polarize an issue, then present two exaggerated extremes, and try to figure out which one is correct. In the real world, neither is correct, and the truth is somewhere in between.
This article tries to conclude "there is no open source community." They say: "Some software vendors believe that open source is an ideological movement." but say that this is an "entertaining narrative" and that the conventional wisdom (that ideological people drive open source) is wrong.
Why can the middle ground be true? Ideological believers in open source contribute significantly to open source. They evangelize and often they diretly contribute (with code, for instance!). Will an open source project die if the ideological believers abandon it? Will an open source project die if the community stops caring? The answer is (as always): it depends. Many projects are community-driven, so of course they require the community push. Others are driven more by companies, so as long as there are enough companies involved, the project will persist.
I have not finished reading the article, but already I'm annoyed. I find the black vs. white picture it paints a bit boring. The real world is complicated. It is worth making the point that companies should not fall into naive assumptions about open-source... but then again they would be silly to ignore the history of open-source, and the fact that alot of it really is driven and maintained by the community. Use that community to your advantage (but do not be led to believe that they are the final word in every respect).
So is there an Open-Source Community? Yes, of course.
open source is much more about supply and demand
Very true. If there was not a need, OSS would never have gotten started. If vendors had provided good quality, resonable cost software, OSS would not exist.
.signature not found
Either this story is totally fabricated, or the company you work for is staffed by complete morons and will likely go under shortly.
First of all, you don't need a system administrator to install any of those things. Apache, Java, Ant, Eclipse, Tomcat, can all run from your home directory, or anywhere else for that matter. Don't have access to port 80? Run it on some other port for development.
Second of all, Java is not open source in any way, shape, or form.
Third, WTF is your employer doing asking you to write a Java application, but forcing you to jump through hoops to get the software to do it?
Fourth, if this application you are writing is supposed to be deploye don Apache and Tomcat, then obviously the company has already given the go-ahead to use this open source software. So why the hassle?
It sounds like this is either a case of the left hand not knowing what the right hand is doing, or a case of complete incompetance. Neither of which is good for a company.
During Margaret Thatcher's reign in the UK, she said 'there is no such thing as society'. I find this to be very similar and flawed in the same way. Not everything is supply and demand, tooth and claw. There is room for altruism, generosity and openness too. I find all these in many of my contacts with 'open source' folks. Or maybe I'm just and old hippy, past my sell-by date...
On y va, qui mal y pense!
The article lacks evidence. It spends a great deal of time talking about economics of scale without at any point presenting what specific scale is required for certain effects to occur. Further his timeline is very far off. When open source developed most software were written by a very small number of people living close to one another and then distributed widely by mail. Sure the wide adoption of the internet helped both commercial and open source software use resources geographically far apart but he completely fails to explain why one side benefitted more than the other.
What are the implications for software developers? The obvious manifestation of a lower bar to entry coupled with an increasing number of programmers is that it is getting awfully hard for a developer to charge for software. (Quick, tell me the last time you paid for a bare-bones email client.)
A great example. In 1995 when was the last time people paid for software that had been expensive in 1980? The 1980 office products would be free throw ins by 1995. Small utilities are first sold separately and then get bundled into other larger programs. There proves nothing about scale.
It used to be that a developer could hack up some small utility, pass it around as shareware, and ask nicely for people to send money. While shareware still exists, the trends are not in its favor. More recently, people who hack together a simple utility simply give it away. They don't ask for payment, because they recognize that it's generally a fruitless endeavor. It's not that they give away the software because they think it's a nice thing to do; they give it away because it's the only way anyone will actually notice.
There was never a period of time when shareware was a particularly good model for anything other than marketing. The original shareware authors generally had a plan of:
1) Write shareware
2) Build up a user base (who pretty much don't pay)
3) Use this base to get a commercial vendor interested enough to finance bring the product out commercially
I could go on but this strikes me as a college freshman economics term paper on applying economic ideas to a recent trend, not as a real insight.
That there's Open Source Software (OSS) and really a Free Software community. Thinking about it, the 2 do have difference in application. Hence, there's OSS and free software, where FOSS is subset. That's a business take on this. Unfortunately 98% of the non-technical people can't grasp this, considering the same 98% doesn't understand what a [linux] kernel is.
It is true that the licence needs to be cleared, but surely that would be generally easier in the F/OSS world than the proprietary one.
In reality, the legal team should just go through the major F/OSS libraries then they would have no need to continually ask people about "what ifs". They could have a checklist of things that the software will be used for and you could probably tell in 15 minutes whether or not that licence is acceptable for that case.
In fact, that's one of the reasons I love F/OSS so much: with normal closed source software I have to read and re-read the licences to know exactly what I can and cannot do. With the free stuff, I just look at the name of the licence. I already know my rights and requirements for a fair few of these licences and I save time just knowing that I won't have to try and understand yet another licence in the closed source world.
Remember that by using hte software, you are agreeing to a license of some kind (GPL, Apache...whatever).
No you don't. The GPL doesn't restrict use in any way, and you're entirely free not to agree to any of its terms. If you don't agree, you're not allowed to do any things that copyright law restricts (e.g., distributing it) but then you weren't allowed to do before you started using the software either. Merely using GPL software doesn't mean you have to agree to anything.
On the other hand, if you use any software that has a EULA, an actual use license, then you are perhaps agreeing to something when you start using it. But I've never seen any open source software with a EULA.
I believe posters are recognized by their sig. So I made one.
Incorrect. For most proprietary software, yes, the license attempts to govern use and must be "agreed to" prior to using the software (whether this is legally valid of not I don't know but remain extremely skeptical). Most free and open source software does not include any license governing use (though it does include a disclaimer of warranty). The GPL merely stipulates conditions under which actions that would otherwise be copyright infringement may be performed. And I don't see how any court could decide that a text edited using a particular program is then a derivitave work of that program; please correct me if I am wrong.
I've noticed much free software ported to Windows requires, during installation, that one click "agree" to the GPL. This annoys me to no end because I need not agree to the GPL in order to use the software. Perhaps this common practice has confused you.
You've explained IT experts (sys admins) upper hand in just a few sentences and why they think they are god half the time.
.NET, MySQL setup, ISAPI filters on IIS, or whatever. I'd be blaming your companies policy for addressing the technical needs of the employees in such ways or worse yet the fact you have to "befriend" a sys admin to establish a fairly standard software package on a server for you.
I dont really think that has much to do with open source. I've done dev contracts for a few companies which I have relied on services to be pre established for me, EG a soap interface for instance.
It took nearly 3 months to get the damn thing established for me before i could do what was required. So likewise weather its Java SDK,
Seems like more of an internal political/process issue then software related.
And in all due respect the Java SDK comes as a rpm or a deb file in Linux most of the time and requires you to type in a command or click a button (you cant get much easier then that dood). So a 31 step manual is perhaps, some dork in your FOSS has gone to the Sun website and just downloaded some dumb ass doco and sent the thing to you while thinking "here ya go now piss off i have other things to do".
What you have done in your posting is prove something which i've been saying a long time, people blame open source because its hard to use sometimes but really we should be blaming the cheese churning IT industry for producing dimwits under a 6 month time frame by shipping them off to MCSE courses and paying them an upwards of 60K a year to sit of their fat lazy asses. While you can hire a decent industry experienced system admin with a few years Unix experience for 10 or 20k extra and can do the same thing 5 the other idiots can achieve in less time.
Remember a good trade mens NEVER blames the tools, and if he does then hes a fool, though if the tools are broken you have to consider that the tradesman usually picks the tools most of the time and therefore if he cant use them he shouldn't have them.
Further to that i dont want to hear this "oh it has to be easy otherwise no body will use it" because your dealing on a different level here. You don't hear doctors chucking a tantrum because the process of a heart transplant is difficult and only a few people can pull it off.. Thats why they are called professionals and it should be the same in IT, up to this point i cant see why it cant be in many occations for IT.
Remember that by using hte software, you are agreeing to a license of some kind (GPL, Apache...whatever). If you are an officer of the company, you have just created a legal obligation for your corporation.
This is not true. GPL-style licenses are licenses to copy & distribute, they are not licenses to use or install. In the USA, you do not need permission to copy software for the purpose of running it (that covers installation, copying into main memory, etc).
You don't have to agree to a license in order to use most GPL software. You have to agree to a license in order to copy the source code and use it for your own programs.
That more and more open source programs make you agree to the GPL like it was some sort of EULA, baffles me. It isn't.
{Slowly pulls handgun out of pants and pulls back the hammer}
Next time, make sure you get it all the way out of your pants before letting go of the hammer.
See my journal for slashdot ID's by year. Mine created in 2005. http://slashdot.org/journal/289875/slashdot-ids-by-year
Agreed. However I was thinking not so much in the vein of Eclipse (although it was mentioned above) but moreso things like:
* we're writing a program and there's a free library here... what are our requirements in using that instead of writing our own?
* we're wanting to use a web app but want to change / add some features... what are our requirements with regards to our end users?
* we're writing an app that uses GPL code but only for internal use... do we have to provide source code to anyone?
There's just a few questions that even I've been asked from time to time. Having said all that, I think you almost made my point: you mention 'GPL' and knew that you didn't need a licence to use it. By just knowing the name of the licence, you understood your legal rights and requirements.
Now if I asked you: does the developer licence on Company X's component Y allow you to write a competing product, the only way you could be sure would be to read (or get Legal to read) the actual licence. If it was the LGPL, for example, you would know without even having to read the thing...
> Freedom also means that you don't have to make your software "open source" or "Free" if you don't want to.
You're calling the power to take away other people's freedom, a "freedom" in itself. Rubbish. When liberty in an inalienable right for everybody, yes, the "Freedom" to own slaves will be lost. No tear shed here.
Please help publicise swpat.org - the software patents wiki
I'm sure you've heard this before... You can take "freedom" to always mean "your personal freedom". Or you can take it to mean "freedom of society", or a tangled interconnected web of freedom. I am not free to swing my fist into your nose, as the saying goes. Is this a bad thing? Or maybe I am free to punch you, but I then must suffer the consequences of either you punching me or society kicking me out.
Anyway, I don't think "open source" is about freedom at all. Perhaps you are talking about the GPL vs. BSD license debate? From everything I've heard, Stallman is right: open source is based on the "many eyes make bugs shallow" argument, or "many eyes lead to quicker improvements and better software" (I thought The Mythical Man-Month disputed this very assertion?). The executive in Stallman's anecdote was certainly adding more eyes.
As for his point, I did not see too much that's original or any pieces of concrete advice. The Open Source movement has never pushed the four software freedoms over "practical" matters and has always had a fuzzy philosophy based on economics above all else. Other than slapping around a strawman and GNU, I'm not sure what his point was. Mostly he thinks everyone should think like him and pretends that it's true. He does not have any positive advice like, "do this and things will be better for you." The author mostly belittles people with ideological motivation without understanding that motivation or it's importance for his own well being. He summarized in his four key points, here:
Paraphrase: The internet is expanding and that will push Open Source which is just another tool without inherent morals.
The view that there is a core group of altruistic companies and true believers driving open source forward is simply false. The view that open source participants are idealistic Davids fighting against software Goliaths is also false. In fact, surveys of open source participants tend to bear this out.
Surveys don't bear this out. The average free software project is created by someone who just wants things to work and has no interest in monetary returns. Other surveys also bear out the importance of freedom for those who are using free software. The free software community has grown much larger in recent years and it still contains many people who are ideologically motivated. If he thinks their work is unimportant, I'd like to see him do without GNU's GCC, and other tools.
If he thinks that the movement will continue to grow without freedom, he's very wrong. The DMCA, software patents and other issues have a real ability to stop both free and open software dead. A very easy test of this is to look at licenses that are open but not free. An extreme example, and the limit of amoral "open software", is Microsoft's initiatives. This is really just an extension of the cross licensing cesspool which was created when a bunch of greed heads tried to scoop up the whole world of computing back in the 80's. Other less than free licenses form a spectrum that attracts more or less participation. Without software freedom, open source would quickly fall on it's face because no one wants to particpate in things that are owned and controlled by others.
The internet will continue to be pushed and expanded by government and major publishers with more or less freedom for it's end users. Free software will continue regardless.
If he thinks he can ignore the good advice the FSF offers, he's dead wrong about that too. I don't think they ever claimed to be the one and only driving force of free software. They understand that it's users writing software that gets the work done and that they can only do that if given the freedom they need. They are a loud and sensible voice for that freedom, and have created a very popular model, the GPL. Freedom is very important to a larger piece of the Open Source community than the author would like to realize.
Friends don't help friends install M$ junk.
Some random thoughts about complexity. I don't have a coherent argument though:
The author seems to assume that the more programmers there are, the more a software project will advance. In my experience, though, a small, dedicated team of 1 to 4 programmers can outperform the entire rest of the world in 99% of interesting cases. On page 3, for example,
The author seems to equate an increase in complexity with an increase in functionality. It's true to some extent, but it also makes maintenance harder. To maintain or even improve any software, you need people who understand that software and, more importantly, who understand each other's changes. Which is why it's so nice to have a small group who can meet and talk and make decisions together. And to be productive, those people have to have a really good reason to:
So far, I have seen these qualities mainly in commercial teams, with a few prominent exceptions in the open-source world.
The bicycle is a worse example...or it will be until someone builds a matter duplicator. The picture example is pretty dead-on, the only thing it leaves out is the status of transient copies made during the execution of the program.
OTOH, what is the legal status of a photograph taken by a respected professional of a painting by a painter who was unskilled and had never sold a painting? Practically, the professional would get the permission before taking the picture, unless the painting was in a small part of his composition, in which case...fair use? But is this the legal requirement? It would be difficult for the painter to demonstrate ANY financial harm, and he probably hasn't registered his copyright. Imagine a photo which the photographer titled "Art Class" which contains numerous paintings which are "works in progress" in an art class (after the class was over and the students had left) at a local high school. Say the photographer got the permission of the teacher, and the students weren't even informed.
Now imagine a software library containing an assortment of algorithms programmed by unknown programmers of varying levels of skill, such as used to be commonly available at various schools. (Maybe they still are, though I doubt it. But you can find such collections on the web, if you happen on them.) What is the status of that library? What if one of the authors happens to find that his work has been included. Yes, he would have given permission, probably, if he'd been asked at the time, but at the time no one *thought* of asking for permission..and not it's been circulated for years.
These aren't simple problems, and legal changes have made them ever more complex, and dangerous. Copyright law isn't as bad as patent law, but it's bad enough. At least patents still expire, but Mickey Mouse will slave eternally.
I think we've pushed this "anyone can grow up to be president" thing too far.
No, Tripwire and sniffers have proved insufficient. They are imprecise, and far more difficult to use than it is to read source code. Consider this: how many hundreds of thousands of people leave college each year being able to read source code? And how many of those can use sniffers to learn what data is being sent by an application what wants to keep that data secret? Very, very few.
This is the reason that these methods only find spyware in very mainstream applications: there are very few people who can and do use them.
Sniffers? It's my network? Not when the data's encrypted.
Laws have also not proved effective for stopping abuse of software users. Remember that End User License Agreement? No, of course not, but you agreed to it, and you waived your rights and you said it was ok for Apple, RealNetworks, and Microsoft to run spyware on your computer.
And when you require software developers to disclose what information is being sent, how is that audited? One European country passed such a law about website cookies, now every time you visit a commercial website from that country for the first time, you're asked if it's ok for them to store information about you so that they can provide better service to their customers, etc. etc. etc. (well, actually, most websites have ignored the law, but anyway). KaZaa's agreement said "we can use your computer as remote storage and can use your processor for stuff" - and everyone (with insignificant exceptions) agreed to it.
You cannot get around this problem by bolting piles of numerous ineffective ideas together. There has been no proposed solution that even comes close to free software.
Please help publicise swpat.org - the software patents wiki
Um ... the original design of the US (small central government with most laws enacted by states) doesn't exist anymore. That ball stopped rolling a long time ago.
-russ
Don't piss off The Angry Economist