Slashdot Mirror
The Gift Culture in Cyberspace
rleyton writes "Prospect Magazine has an interesting (albeit high level) article about the entire free software movement. Good quotes, but the most interesting is
the end questions that might provoke a bit of debate.
"
While I'd agree with you that most of the OSS companies have been generally well-behaved, the "rift" does indeed exist, mainly between ESR and his camp and RMS and his camp. If you take a look at the license-discuss list that OSI runs, there's been some notable incidents, including:
RMS (correctly) insisting that GNOME is Free Software, not Open Source (and being backed up by Miguel de Icaza, the GNOME project leader)
ESR asserting that RMS could be a better spokesman, but isn't because he doesn't bathe often enough (an assertion that RMS calls exaggeration).
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
What really scares companies like Microsoft about open-source software is the fundamental shift in philosophy it will require them to make. It's not about how secure the software is or who has access to it. Fundamentally, Microsoft wants everyone to have access to its programs, just like OSS preachers. The difference is that Microsoft is primarily a product driven company which provides some limited service for its product, similar to a car dealer. OSS companies like Redhat are a service driven company that sell some product on the side. Service-oriented industries are, by and large, dependent on the product and what scares Microsoft is that if they must turn into a service oriented company, they won't have control over the product and thus fall under many more market influences. Redhat can succeed as long as the product is there since they don't have to sell it (what they sell for $80 now isn't Linux but technical support for Linux). Microsoft, on the other hand, sells Windows and tech support comes with it. If they were forced to open-source Windows or Office, they would lost a very, very large chunk of change, and honestly, do we really need Microsoft for technical support. There are much better avenues for that (like your local computer dealer or manufacturer).
But that begs the question, will open-source survive or even thrive. Right now, OSS has made it because they haven't needed that service. The mainstream market (the people who need support organizations) haven't bought into it. Now that the masses are beginning to understand and accept this concept, you see companies like Redhat thriving because of their expertise, but currently, it's not enough. There's too many problems for the typical newbie to overcome. The support channels aren't strong enough to support it yet. When those support channels strengthen (and the product becomes easier to use, meaning the support channels don't have to necessarily be as strong), then you'll see greater acceptance. If, for some reason, the support level can never truly match customers' needs, then you'll see OSS flounder in minimal usage. OSS will only thrive as long as the service industry can support it. Windows thrives now partially because that service industry is already in place. Microsoft doesn't have to worry about it because they make the money off the product.
The threat to Microsoft is the same as the challenge to OSS. Hopefully, more organizations like Redhat will come along (some may consider that blasphemous) and broaden the support (and customer) base. But OSS can't truly grow on its own unless the underlying infrastructure begins to also develop for the people that use it.
Actually, with that statement, they hit the nail right on the head. See: Why ``Free Software'' is better than ``Open Source'' for more information.
I think it amusing that they made such a point to mention Stallman, then completely neglected to talk about the deeper relationship between Linux and GNU. That's clueless. ;)
As mentioned before, they mean free as in freedom, or free speech, not free beer. It seems as if fewer and fewer people realize what the term "free software" means these days.. See: What is Free Software?.
~ Kish
Surprising for someone who loves to rip into "journalists" every chance I get.. Anything with a license can not be considered public domain because public domain is a legal term with a very specific definition: not copyrighted. Copyleft or any other "free software" license.. hardly counts..
I think it's interesting that they actually noted two different factions amidst the community, as there clearly are. Personally, I fall somewhere between Torvalds and Stallman, which still puts me closer to Stallman than Raymond (ugh). I think the idea of applying free software to games is pretty silly, especially since if I design a whole crew of personalized characters, I'll be damned if I GPL them along with the entire game itself. :)
At any rate, I doubt rivalries or other factional disputes between various groups in the free software community are likely to cause anything to collapse. We all have our personal beliefs (or lack thereof), disagreements, whatever. But that's to be expected of something as anarchic as this. Sure, a lot of Stallman's original message gets clouded a lot of the time, but the point is: it still works. And no one is likely to lay down and let their ideas go unnoticed, so each group is likely to remain rather formidable as far as numbers of people with like-minded beliefs go.
In the end, however, we have better things to do than bicker and argue over the finer points: like code. ;)
~ Kish
Just as advances in flight control and alloys allowed for these planes, so too did advances in communications technology allow for the Open Source software development model.
The proprietary model requires the entire value of the product to be amortized in a single fee. Where Open Source differs is that it enables the COMBINED values of each piece of the whole to be farmed out to people who find it valuable TO THEM. It's a kind of stone-soup, everyone contributes a piece and gets back a whole (which seems to be more than the sum of it's parts, though formally it isn't).
- Seth Finkelstein
How many of the movement leaders actually have a 200+ IQ? How would these people know? And since when did our movement leaders get labeled as being dicks? They always seemed to be pretty nice people, for the most part, if you ask me.. ;)
That said, I naturally agree with Signal 11 (what, you thought I wouldn't? :)
~ Kish
"We all know that the free software movement only benefits the upper(-IQ) crust of internet users. Those of us who are intelligent enough to know how to download Mandrake 6.1 from the FTP site."
Um.. You don't have to be able to download GNU/Linux from an FTP to benefit from free software (which is how you would benefit from the free software movement), as you make implicitly clear in your next sentence:
"The masses will continue to buy software which is where companies like Red Hat and Mandrake will make their money."
Are you confusing free with gratis? Because even though Joe Public is spending some dough, that doesn't mean he isn't getting a better deal because of free software (it's not like free beer! ugh..). I know a lot of (mostly) clueless people who rather like having shell accounts (on GNU/Linux or *BSD systems), benefit from pine (most other email programs that Joe Public is aware of really suck..), etc. etc.
So I'm not sure how you mean that the FSM isn't benefitting anyone but the most intelligent of us. Besides, you don't have to have a certain IQ to know the ins and outs of the free software community. You just have to be aware of it. Even people with low IQs can learn pretty much anything if they apply themselves. Learning comes down to dedication. You can't discriminate just because of some score on a piece of paper. Do we really look at people and guess their IQ? It never comes to mind for me. Ever. How creative they are? Maybe. How smart they /seem/? Perhaps. But an exact rating of their intelligence? No, because you never know..
And my cheap shot of the day: Don't most of the "upper crust" use Debian? ;)
~ Kish
I'd say that you're correct, about modularity's importance to large projects and why.
We shouldn't lose sight of the fact though that sometimes modularity must cut across the grain of a program. Sometimes with very complex software everything really is closely related to everything else, and you have to cut modules where it would be better to leave things whole. Modularity is something programmers are forced to do -- sometimes you get lucky and it also models the problem you're trying to solve -- but not always.
As was probably mentioned before (this post got a lot of responses.. ugh), GNU/Linux is modular. Windows is monolithic. Once a program gets too big (such as Win2k), being monolithic really sucks. It's easier to run a small program and test it extensively than a huge program because there are fewer variables.. not as many things that can go wrong.. less trash to run through the debugger. ;) It has been considered "proper form" to code modularly for quote some time now. The particular "edges" that Microsoft gives itself relies somewhat on being monolithic, however.
Now, as far as Windows instability goes.. I happen to know a few people from good old Redmond, and though I obviously can't substantiate this claim past that (I don't trust MS programmers as far as.. well.. let's just say I don't ;).. From what I understand, Microsoft doesn't really do much of /any/ bug-testing "in-house". That's what their limited beta releases are for. If you ask me, that's kind of silly, but then again.. 1) It's always better to have people test your software that aren't "close to the source".. however.. 2) Since Windows is so monolithic, any one programmer at MS is only likely to know a very small part of it. No one person could hold all that information in their brain (unless they have a photographic memory.. yes I know that's not the correct term for it..).
With free software (I refuse to use the term "open source") most developers release a stable version and a devel version.. That way people who are interested can get the devel version, see if it runs, poke around at the code.. And since they aren't the primary developers, they may notice a number of bugs that the people who actually designed it missed.. And since this happens on such a widespread level, you get a number of different points of view all staring at the same thing from different angles, each with their own coding style, skill level, etc. This kind of development squashes bugs at a rather, well, phenomenal rate. And since it is open to anyone, not just people who sign some kind of beta test agreement, it's a lot more likely to get a good number of interested volunteers helping out. :)
With the Microsoft way(tm), most of the bug-testing is done by, you guessed it, the end-user! (Those limited beta tests just don't cut it..) Not something Joe Public is very fond of these days.. ;) Especially when they charge a $100 for a bug-fix upgrade! Argh..
~ Kish
"The accusations of communism really are not justified, even hard core free software people like RMS are NOT communists."
I believe he once described himself as a combination of a leftist anarchist and a (?) libertarian (sp?). ;)
"It is the SCIENTIFIC METHOD. The "Free Software" "OSS" movement comes directly from the modern scientific method of peer review."
Actually, the free software movement is a direct result of Richard Stallman's ethical views. Yes, it's also a matter of practical concerns, but RMS would rather have free software even if it was technically inferior. See: Why Software Should Be Free (a shorter version can be found here).
The open source movement stems from those like ESR who are.. well, I won't go into detail on my thoughts. The original advocates thought that the term free software scared big business. And it did. However, now the term is corrupted beyond recognition (by.. guess who!? big business..), and also appears to have become the norm these days. How sad. See: Why ``Free Software'' is better than ``Open Source''.
Perhaps "slave" is a rather strong term in this case, and I can't recall if Stallman ever described it that way or not, but it's closer to the truth than most. The free software movement, however, was /not begun due to practical concerns, but rather ethical ones!!/. See: The GNU Project.
~ Kish
When software is just written as a tool or to accomplish a general task, OS works beautifully. Just as peer review in science tends to result in research that withstands scrutiny and means something, peer review in software development tends to result in higher-quality, more secure software (I definitely fall into the "pragmatic" group in the article).
But it doesn't work for everything. In-house applications whose structure must reflect internal processes may not work as open source, because it reveals too much internal information. Additionally, as many have pointed out in the past, games have too short a life-cycle to really work as open source. The engines fit the model, but the game as a whole really doesn't.
This isn't a slam on open source software. I prefer it in almost every arena, especially when I'm concerned about security. But we need to recognize that there are situations where it doesn't work. Closed software is not the same as slavery, folks: there are times when keeping the source hidden does work and is appropriate.
"You can never have too many elephants on your team."
It's wrong because it presumes that you need to be intelligent to contribute to this movement. It's wrong because intelligence is NOT the primary attribute required to make a meaningful contribution to this culture, and it's wrong because ESR said it. It is simply a wide misconception that you need to have a 180 IQ to program - Microsoft is proof of this. *rim shot*
If you guys want to talk about the technical community, talk TO the technical community before you publish - don't go making blind assumptions about what you need, or do not need, to be a member of this community. When in doubt - ASK!
Do I seem upset? Yes, that's because I am. That statement does more to undermine the free software movement than virtually any other - we accept contributions from anybody. If it's good, we'll take it. You don't need to be an Einstein to join us.. all you need is dedication, and time.
--
I'd say it suits all, but fits some better than others. As a general note, Open Source code is only released when it is correct - I can't bring to mind any new release of any OSS that is not a definite step forward, yet I can easily bring to mind many patches from commercial suppliers which exacerbate what they were trying to fix. I think this is because OSS programmers aren't under pressure to bang out the next version - proof that Sales & Marketing departments are inherently evil and counterproductive to good code. By this nature, OSS software is not suited to very fast moving technologies, yet is excellent at instances where programmers have the time and opportunity to get it right.
What does it mean for the economics of cyberspace?
I can see less & less commercial vendors out there flogging proprietary code, which could be dangerous. If there are less professional programmers, there will be less students, so who will actually write the Open Source Code? There will have to be a big shift in corporate attitude to prevent this. Companies will sell mostly services, and give away the applications free. So cost models will shift to reflect that; most of the cost models I have seen in industry are heavy on the "system" cost and mostly have support services bundled in. But will users pay heavily for something as ephemeral as support? If you buy a system, you have something shiny & new you can show your shareholders. If you buy support, you have a voice at the end of a phone line that shareholders never see.
How can free software coexist with hard-nosed profit-maximising companies?
Closed Source will almost always be the poor cousin of Open. Even if companies base their code on existing Open source and charge for any additional reworking, there will still exist a "free" version within a few months. Companies won't dare invest in innovation with Open source, in case someone picks up the idea, rewrites it and gives it away.
Business which depends on Open Source (like RedHat) must keep the OSS community on side, or will rapidly perish.
Strong data typing is for those with weak minds.
The article is bland, but the questions interesting. I have thought this before releasing my own code as OS (cpuburn http://users.ev1.net/~redelm FYI).
First, people will code to solve their problems. Second, if you've already written the SW, the cost of releasing it OS is minimal. Only if you could sell it for significant $$$ would you hold back. Or planned that all along (software companies).
But the Unix design is very modular. While a whole Linux/*BSD OS _is_ worth alot of money, the components are worth very little sold individually. Fortunately, they are also easy to write.
So I would argue that OS will work for modularizable code, but not for large monoliths such as MS-Windows applications. Even large Xapps become monoliths, and saleable. How much would you pay for Lotus 1-2-3 for Linux/X?
Sometimes, strategic considerations cause companies to release otherwise saleable code as OS [Netscape]. I expect this is the exception, not the rule.
-- Robert
Since I obviously don't have access to the Windows source code, and in fact haven't even seen Windows 2000, I'm going out on a limb here :P However, I can answer in one word:
Modular.
Most free software projects I'm aware of are highly modular and compartmentalized. This is obviously nice aesthetically, but also necessary for the project to even exist: if the developers are geographically dispersed and working in their free time, no-one (well, almost no-one) can be an expert in every part of the system. It is therefore *essential* that the program be broken down into bite-sized pieces, or no-one will be able to comprehend it, much less work on it. Keeping a clean separation between components has the side effect of tending to contain bugs within a code module, therefore making debugging much easier.
My feeling just from using Windows is that things are much more tied together -- 'integrated' if you will. There is an effort to divvy things up, but it feels like this division isn't clear enough, especially in Win9x (where you have all sorts of weird hacks involving DOS programs interacting in strange ways with Windows programs..)
Small discrete programs lend themselves really well to free software development, perhaps one reason that free software has been traditionally associated with UNIX-like systems.
Daniel
Hurry up and jump on the individualist bandwagon!
I think a lot of this has to do with the basic design of both systems.
The core of Linux was designed as a clone of the Unix core, a system that, whatever its faults, has undeniably stood the test of time.
The core of Windows was a kludge, designed to give a crude graphical interface using hardware that really wasn't there yet. From inadequate disk space to insufficient memory, machines running Windows 3.x were not ready for any kind of elegant solution.
Each subsequent version of Windows has attempted to add features to the original. For instance, Windows for Workgroups introduced networking, Windows95 introduced the 32-bit subsystem, and Windows98 introduced the abysmal web-based user interface. But the basic core - the use of DLLs in the C:\windows\system\ directory, the lame "multitasking", etc - remained. In other words, the new features attempted to paper over the deficiencies of the old system. What they did was like adding a remodelled kitchen with a Sub-Zero refrigerator to a house with a crumbling foundation.
The result is, unsurprisingly, an unmaintainable mess. It's only partially upward-compatible and doesn't really work very well.
The biggest strength of the Linux development model is that it would not allow an abomination such as Windows to exist; nobody would be willing to pour their love and/or spare time into such an effort. Instead, development proceeds in a modular fashion, and interactions between, say, X-Windows drivers and the basic operating system are avoided. That goes a long way towards avoiding crashes, too.
D
----
Yes. Eyeball count is only part of the story. The other parts are motivation and ability.
If you're working in a software house, you've been given a chunk of code to write/debug. Probably a poorly defined target. You'll do that, and go home at 5pm if you can. Fixing problems in other people's code likely won't bring you much reward. Worst of all, you don't really have ownership of anything.
Contrast that to OS: People have a problem they need to solve. They debug or write code to solve their problem, and share it because it's easy to do. They _own_ their work, even if Linus rejects the patch, they can still use it.
Ability is a very controversial subject. Suffice it to say that I don't think you can hire contract programers as skillful as OS volunteers.
-- Robert
"The free software movement raises some big questions. Does it provide a new method of producing high-quality software, or does it apply to only some programme categories?"
It only applies to two categories well. First, it's very useful, and will most likely replace proprietary software altogether, in the area of infrastructure and "commodity" software. This includes web servers, text editors, kernels, compilers, etc. These are the projects that thousands of people code on. Some things that aren't "commodity" classed yet, could be in the future if an Open Source version makes it so. For example, office suites aren't a commodity yet, but the release of KOffice and Workspace could very well make it so.
The second category is "hobby" software. Someone has an itch and scratches it. It's written because someone wanted to have fun writing it. This software doesn't get the benefit of ESR's thousand eyes peering at the source code. Often, only one or two people ever code on it. But sometimes these programs will mushroom out and take on a life of their own.
Other categories of software will be a mix of Open and closed source programs. I don't see Open Source taking over more than a small minority of games or vertical apps.
"What does it mean for the economics of cyberspace?"
I hate buzzwords. If by 'cyberspace' you mean the internet and its sucessors, it has already changed the economics of it. The infrastructure of the internet is already Open Source. The commodity software of the internet (browsers, media players, commerce solutions) are rapidly becoming Open Source now. Javascript, perl, HTML and XML have open source code right now, it's only a small step to make them true Open Source.
"How can free software coexist with hard-nosed profit-maximising companies?"
For infrastructure/commodity software, it's the proprietary companies that will have to learn how to coexist. For other categories, Free Software developers are going to have to learn the rules of the game: marketing and targeting. People who buy vertical applications don't choose their application out of ideology. They choose based on suitability to their needs. This will mean that Open Source solutions will fail if the proprietary solutions are targeted and marketed better.
"And will the movement, riven by tensions between Open Source pragmatists and free software fundamentalists, fall apart under the pressures of success?
They movement may fall apart, the concept of open software will not. It existed with the first mainframe and will continue to exist until we replace the generalized computing device with something better.
It's important to note that every movement has it's "true believers", the "core group", and the "hanger-ons". True believers are always in the tiny minority, whether it be a political party, a religion, or anything else. Hangers-on are always the largest group but they rarely contribute anything. The core is what drives the group. They aren't fanatics but they aren't followers either. If they feel that Open Source has to change, right or wrong, it will do so
"Open Source works in practice; will it work in theory?"
If it don't, change the theory.
A Government Is a Body of People, Usually Notably Ungoverned