Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
The GPL: A Technology Of Trust

By grout in Op-Ed
Sat Jun 16, 2001 at 11:58:19 AM EST
Tags: Software (all tags)
Software

Microsoft's attacks on the GNU General Public License (GPL) prompted me to analyze its technical merits, using insights from the book 'Nonzero' by Robert Wright. Since I'm a fan of Open Source for its pragmatic benefits, my own conclusions surprised the heck out of me.


Society is built on exchange. One particular form of exchange that we're genetically wired for is reciprocal altruism: speculative generosity with expectation of future payoff.

Open Source is a textbook example of reciprocal altruism. But this leaves the Open Source community vulnerable to parasitism. (This term comes from game theory; I'm not trying to insult anyone.) In a small group, trust comes from repeated interactions, and personal experience is adequate to recognize parasites and avoid them. But in a large group, interactions between any two people are often indirect and/or infrequent. Something more than experience is needed to engender trust between people who don't know each other, and who may never even meet.

Therefore, any large group must evolve a technology of trust. If it doesn't do so, it will fall victim to rampant parasitism, which will cause inefficiency, which will eventually bring stagnation and failure to compete -- that is, death.

The GPL is a technology of trust. Contributors to GPL'd projects trust that the GPL -- which depends on law, itself a technology of trust -- will prevent parasitism. They trust that if they contribute to a project, they will have access to the valuable goods built on their own work. So, while GPL'd projects can have forks, they can't have proprietary forks. And that makes all the difference.

This analysis may seem simple or even obvious. But its implications are far-reaching.

1. The GPL will eventually dominate Open Source (if it doesn't already). Both analysis and observation point to the GPL, or something like it, as the destiny of Open Source. More than any other current license, the GPL discourages parasitism; thus it enhances efficiency; thus it helps a culture outcompete rivals whose technologies of trust are less advanced. By making its host culture successful, the GPL -- or some future license built on it -- will finally win out.

2. We must preserve the GPL, for the sake of the community. When Microsoft attacks the GPL, it would be tempting for those of us who don't identify with ``Free Software'' to use as our primary reply that ``Open Source is more than the GPL.'' That would be a mistake. The GPL's peculiar strengths are crucial in the Open Source community's competition with other cultures who would love to see Open Source, let alone Free Software, gone and forgotten.

3. The GPL is good for business. Companies that use the GPL are neither foolish nor stupid. They simply want to trust that other companies won't be able to take unfair advantage of them, and the GPL gives them that immediate security while simultaneously allowing open cooperation. And in the general case, the GPL is a friend of business because it makes new and better efficiencies possible, and economies thrive on new and better efficiencies.

(On the other hand, we can agree with Microsoft that the GPL is bad for their current business. We can then proceed to use Microsoft's favorite word as we reply: Innovation won't stop just because you're not ready for it. The printing press was a good thing, after all, even though it forced professional scribes to change their business model. Adapt or die.)

In summary: We in the Open Source community need to stand with the FSF and defend the GPL against all comers -- not merely as a tactical move, but because the GPL is a valuable technology of trust. To outcompete other cultures, we must adopt technologies that work. And the GPL works.

    -- Chip Salzenberg <chip@pobox.com>, member of the board of the Open Source Initiative

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Related Links
o &lt;chip@p obox.com&gt;
o Open Source Initiative
o Also by grout


Display: Sort:
The GPL: A Technology Of Trust | 65 comments (54 topical, 11 editorial, 0 hidden)
How do we know? (4.00 / 2) (#1)
by chromag on Fri Jun 15, 2001 at 10:43:21 PM EST

To outcompete other cultures, we must adopt technologies that work. And the GPL works.
I was under the impression that it had not been tested in court. How do we know it works if it's never been legally challenged?

Though perhaps that's your point.


--

-c
dump the zeros


Court testing is not the point. (4.00 / 1) (#2)
by grout on Fri Jun 15, 2001 at 10:56:47 PM EST

The GPL seems defensible, according to legal authorities I trust.

If by some fluke the current GPL is found by some court to be partially unenforceable -- which I do not expect to happen, by the way -- then a new version of the GPL will certainly be produced by the FSF before the ink is dry on the court judgement. And whatever changes are made in that new GPL, we can be sure that it will still require distribution of sources with binaries. And that's they key feature that makes the GPL our future.

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Can't parasitize in this situation (4.00 / 3) (#3)
by khym on Fri Jun 15, 2001 at 10:57:30 PM EST

Society is built on exchange. One particular form of exchange that we're genetically wired for is reciprocal altruism: speculative generosity with expectation of future payoff.

Open Source is a textbook example of reciprocal altruism. But this leaves the Open Source community vulnerable to parasitism.

Seems to me that parasitism is only really a problem when the parasite uses up something. In the case of Open Source, if someone were to take a BSD licensed product, make some improvements, and change it to closed source, the original is still there; nothing is used up. Since it's closed source, open source developer aren't going to be "stolen" by the fork. The only thing that might be "stolen" would be some user base. A "tragedy of the commons" type situation just can't apply to Open Source.

--
Give a man a match, and he'll be warm for a minute, but set him on fire, and he'll be warm for the rest of his life.
Game-theory 'parasitism' is more subtle than that (5.00 / 2) (#6)
by grout on Fri Jun 15, 2001 at 11:04:30 PM EST

As I understand it -- though I am not a game theory expert -- the game theory term 'parasite' refers to players who reap benefits without contributing work. Accordingly, a company like, say, Apple, is a parasite because it reaps benefits from the work done on Mach and X without helping the community projects.

Nevertheless, even if you don't consider the term 'parasite' appropriate, the basic argument holds: A license that brings more contributions into its host project will help it succeed in competition with projects with less contribution.

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Genetic my ass, and more (3.42 / 7) (#4)
by treetops on Fri Jun 15, 2001 at 11:02:29 PM EST

Society is built on exchange.

Ok...

One particular form of exchange that we're genetically wired for is reciprocal altruism: speculative generosity with expectation of future payoff.

Puh-lease. "Genetically wired" for reciprocal exchange? Anytime I hear a its-human-nature article my BS-detector goes off. Why is it that everything has to be blamed on genes? Aren't people capable of learning, and passing on what they've learned through society?

Your "speculative generosity with expectation of future payoff" could just as easily apply to venture capital or stock investing as it does to Free Software. Are venture capitalists more highly evolved? Are they closer to nature?

Getting to the rest of your comments:

1. The GPL will eventually dominate Open Source (if it doesn't already).

The problem with this article is that it assumes that a GPL project will gather more developers and effort than a non-GPL project, because it cannot be forked into a proprietary version. This can easily be shown incorrect. Consider the case of Apache vs. any GPL web server. Apache is more succesful, and attracts more effort for several reasons:

    1. It is better (or percieved as better) than the competetion
    2. It can be used easily in commerical projects, and many commerical projects *do* give back code to the community. Just look at IBM and Sun's involvement with either Linux or Apache
    3. Because it can be used in commerical applications, it gathers more interest because people usually want to see their code get used, rather than remain stagnant for political reasons.

The same principles can be applied to any succesful non-GPL system (such as the BSD variants). Claiming that the GPL will crush it's competitors is simply wrong and short-sighted.

2. We must preserve the GPL, for the sake of the community.

The GPL is hardly a necessity for open source. Perhaps it comes closest to *your* view of what open source is, and is a necessity for *your* opinion of open source. Many see Open Source as being good for business because it allows for businesses to easily make use of open source code, something which the GPL cannot do within the bounds of most business's needs (ie, devloping proprietary software). Look at Netscape for example. They rejected the GPL because it was too restrictive. The the NPL, everyone is satisfied: Netscape and the community. Hell, some projects don't have licenses and are explicitly public domain.

3. The GPL is good for business.

The GPL is only good for businesses that create new software. Other businesses that build on or integrate new systems are usually wary of using GPL'ed code for the very real reason that it would "infect" all their other code. For that sort of business, the BSD license or similar is much better.

In summary: We in the Open Source community need to stand with the FSF and defend the GPL against all comers

While I agree that the GPL should be defended (as a part of the Open Souce Community, but not the main part), the FSF and Stallman are not my ideal leaders or representatives of Open Source. For that matter, neither is ESR.
--tt

Apache, etc., and please read 'Nonzero' (3.50 / 2) (#10)
by grout on Fri Jun 15, 2001 at 11:41:32 PM EST

I understand your discomfort at the "genetic" comment. But it's factually correct. I suggest you read Robert Wright's book "Nonzero" for a long (but entertaining) explanation. But here is a key quote: "We tend to feel sympathy for those who are reliable reciprocators. We call them 'friends'." And consider that the emotional reaction we call "indignation", triggered by being cheated or deceived, is built in to the brain. Granted, culture affects our definitions of "cheat" and "deceive"; but the basic orientation toward favoring reciprocation is definitely part of our basic wiring.

Also, I never addressed issues anywhere in the neighborhood of 'virtuous' or 'closer to nature'. Understanding what society is and how it works is helpful in competing with other cultures for the long term. Your example of the free market and capitalism makes part of my point very well: While the Soviet Union was a huge and militarily strong nation, its economy just wasn't efficient enough to compete with the West. In the short term the outcome seemed to be in doubt; but the West outcompeted the USSR so thoroughly that now Russia is in some ways a third-world country by comparison. So while the GPL is "merely" a tool of cultural efficiency, efficiency can be an overwhelming weapon.

Oddly enough, Apache's success actually makes my point in an indirect way. I've claimed that the GPL is better because it prevents proprietary forks, which sap effort away from the project. And while Apache hasn't had a proprietary fork yet, it could have one at any time. ("It's never happened to us" is a foolish attitude toward possible disaster.) If that were to happen, Apache would then be at a competitive disadvantage ... especially against a hypothetical new "Gnapache" project that would take the Apache code base forward but under the GPL. (And if you think that wouldn't happen, just look at the origins of Gnome.)

X is a clearer example. Just imagine how much better X could be today if so much effort hadn't been spent and wasted over decades on proprietary code forks. And proprietary forks of X still existed as recently as a few months ago; they may still be around for all I know.

Finally, please note carefully that I have neither predicted nor requested the conversion of all Open Source projects to the GPL. Rather, I predicted that the GPL would in the long run help its host projects outcompete their peers, and therefore help the Open Source community outcompete other cultures. If you're going to criticize me for my conclusions, please understand them. first :-)

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Apache and other code forks (3.00 / 2) (#14)
by treetops on Sat Jun 16, 2001 at 12:04:24 AM EST

And while Apache hasn't had a proprietary fork yet, it could have one at any time. ("It's never happened to us" is a foolish attitude toward possible disaster.)

No forks? How about IBM HTTP Server?

If that were to happen, Apache would then be at a competitive disadvantage ...

It has happened. Now explain how this disadvantage has affected anyone.

You also make the assumption that forks of a project (whether proprietary or not) won't contribute back to the main source tree. This is demonstrably false. Look at the case of BSD: NetBSD uses code from OpenBSD, which results in NetBSD developers spending less time in security-related issues because the OpenBSD people do most of the work. Then look at RTMX, a fork of OpenBSD which was proprietary for a long time. Now it's being integrated with the main sources. None of these projects are GPL, yet code re-use is used to an amazing degree.

especially against a hypothetical new "Gnapache" project that would take the Apache code base forward but under the GPL.

This is absurd. By your reasoning, people should be dropping non-Linux versions of Unix by the boatload. Yet BSD continues to gain users and expand in commerical applications.

(And if you think that wouldn't happen, just look at the origins of Gnome.)

Errr, so if there isn't a Stallmaniclly-correct piece of software, we should ignore the non-GPL version and waste our time creating a clone from scratch? Talk about competitive disadvantage.

X is a clearer example. Just imagine how much better X could be today if so much effort hadn't been spent and wasted over decades on proprietary code forks.

If there weren't proprietary versions, it never would have gained the acceptance it did. Commercial vendors only bundled X11 to spite Sun anyway, and now we're seemingly stuck with it.
--tt
[ Parent ]

Forks, proprietary and otherwise (none / 0) (#18)
by grout on Sat Jun 16, 2001 at 12:34:20 AM EST

You also make the assumption that forks of a project (whether proprietary or not) won't contribute back to the main source tree. This is demonstrably false.

To the contrary: I simply made the quite obvious observation that forking cuts down on efficiency. I confess that I focussed on proprietary forks in my followups because they are the worst kind, black holes for code. But even non-proprietary forks are bad news for efficiency because they increase the work required to produce a given product. It's a basic issue of communication overhead, hard to argue with.

Hm, I just had a thought: Some level of forking is necessary in every project, insofar as each developer makes a private fork every time he edits a source file. But forks must merge as soon as possible, or they become drags on productivity.

So the real problem isn't forking, but forks maintained in parallel. Perhaps we shouldn't talk about forking as the problem, but rather, parallel maintenance.

I appreciate the pointer to IBM's HTTP server; it's news to me. But even if IBM contributes code from that project back to the main Apache project, they're still putting in extra work maintaining their private fork. If they only contributed back everything (in a perfect world :-)), then they'd have practically no work required to maintain their product. And which license most strongly encourages such behavior? The GPL.

By your reasoning, people should be dropping non-Linux versions of Unix by the boatload.

They are! Haven't you been reading the trade press for the last two years? Linux's most vulnerable competitors have always been the proprietary Unix variants maintained by various vendors like SGI and SCO. And, no surprise, proprietary *nixes are dying left and right.

If you're a fan of the BSDs, you're probably frustrated at the great publicity and enthusiasm behind Linux, despite the various technical and social disadvantages of Linux development. But consider this an opportunity to steal your competitor's secret weapons. If the GPL is a significant reason for Linux being so popular, then fans of BSD might do well to try to learn from its lessons. I haven't taken much time to think about how the BSD projects might benefit from these lessons, but perhaps that would be a good subject for a new article....

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

GPL is good for WHICH businesses? (3.50 / 6) (#5)
by Carnage4Life on Fri Jun 15, 2001 at 11:03:42 PM EST

The GPL is good for business. Companies that use the GPL are neither foolish nor stupid. They simply want to trust that other companies won't be able to take unfair advantage of them, and the GPL gives them that immediate security while simultaneously allowing open cooperation.

This is a rather interesting comment. The GPL has proven to be good primarily for companies that sell hardware and need someway to cheapen their TCO (e.g. IBM, TiVo, etc) as well as companies that use software in their day to day existence (almost all of them). The GPL forces software companies to become support agencies while still spending massive amounts of money on development.

In the long run the GPL will probably be beneficial to people who use software while disenfranchising a majority of those who produce software which will lead to less professional development of software but will make the software that does exist a commodity.

GPL is good in SOME way for all businesses (none / 0) (#8)
by grout on Fri Jun 15, 2001 at 11:12:59 PM EST

I never suggested that proprietary software will die out -- though many people have, especially GPL fans, so I can understand your confusion. Rather, I suggest that non-GPL Open Source licenses will gradually lose their "market share" (for lack of a better term) as the GPL makes its host projects more and more successful. The battle of Open Source vs. Proprietary is much too complex to address in the margin of this comment. :-)

As for your whether the GPL will "disenfranchise a majority of those who produce software", I have to ask ... In your view, what "franchise" exactly do programmers and/or software-vending corporations have, that you think the GPL is threatening?

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Misleading title. (4.00 / 1) (#9)
by Carnage4Life on Fri Jun 15, 2001 at 11:26:13 PM EST

The content of your post had little to do with its title.

I never suggested that proprietary software will die out -- though many people have, especially GPL fans, so I can understand your confusion.

GPL vs. proprietary software is the only battle of interest. The only other major Open Source licenses are the MPL & BSDL which are intrinsically tied to major projects and will probably exist as long as these projects exist. Whether or not all other projects are GPLed or not seems immaterial.

As for your whether the GPL will "disenfranchise a majority of those who produce software", I have to ask ... In your view, what "franchise" exactly do programmers and/or software-vending corporations have, that you think the GPL is threatening?

Currently, programmers can make money from writing and selling software. A GPLed world cannot support as many programmers as currently exist making a living off of software indirectly (either via support or hardware sales) which will mean that lots of people that would have been programmers would not be able to practice their craft.

[ Parent ]
Franchise Lost (none / 0) (#15)
by grout on Sat Jun 16, 2001 at 12:09:03 AM EST

I agree the title of my comment could have been a lot better. Sorry.

When you write: Whether or not all other projects are GPLed or not seems immaterial, you basically disagree with my basic premise that the license does matter.

And on the "franchise" issue: There is simply no practical chance that shrink-wrapped software will disappear from the world in my lifetime, at least. So scaremongering about the 'death of proprietary software' is nothing more than a classic straw man attack.

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Not clear what you are talking about (none / 0) (#28)
by Carnage4Life on Sat Jun 16, 2001 at 07:21:26 AM EST

When you write: Whether or not all other projects are GPLed or not seems immaterial, you basically disagree with my basic premise that the license does matter.

I am not an Open Source fanatic. If all Open Source code is under the GPL or the BSDL doesn't mean much to me even though I do prefer the BSDL to the GPL.

And on the "franchise" issue: There is simply no practical chance that shrink-wrapped software will disappear from the world in my lifetime, at least. So scaremongering about the 'death of proprietary software' is nothing more than a classic straw man attack.

You are confusing two things. Shrinkwrapped software is not necessarily proprietary software. Linux distros can be considered shrinkwrapped software (off the shelf, not built specifically for a custon purpose or user, etc) yet they are not proprietary.

[ Parent ]
Even if it doesn't matter, it might matter (none / 0) (#29)
by grout on Sat Jun 16, 2001 at 07:31:12 AM EST

I understand now what you meant: You don't personally care which Open Source license is used; it doesn't "matter" to you. But what I mean by "matters" is that, in my opinion, the GPL's use will have an effect on the projects it's used for and the community as a whole. It's a different kind of "mattering", and it doesn't conflict with yours per se.

And you're right about my sloppy use of the term "shrinkwrap". Substitute "proprietary distributed software" wherever I said "shrinkwrap", and I think the result will be more precise.

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

good for businesses who release code (none / 0) (#23)
by Delirium on Sat Jun 16, 2001 at 02:38:09 AM EST

I think perhaps a better way to phrase it would be "the GPL is good for businesses who are releasing their code as open source." If they release it under a BSD license or some other such license, their competitors can take it and use it in their own products. If they release it under the GPL, their competitors would have to release any changes back, allowing the original company to keep up on any new developments.

[ Parent ]
Yes, that's almost it ... (none / 0) (#24)
by grout on Sat Jun 16, 2001 at 03:05:00 AM EST

But the GPL is good for building a shared body off software usable by anybody. So I'd suggest a this revision: "The GPL is good for all businesses except those who depend on keeping source code secret."

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Really, How? (none / 0) (#27)
by Carnage4Life on Sat Jun 16, 2001 at 07:18:23 AM EST

I think perhaps a better way to phrase it would be "the GPL is good for businesses who are releasing their code as open source." If they release it under a BSD license or some other such license, their competitors can take it and use it in their own products. If they release it under the GPL, their competitors would have to release any changes back, allowing the original company to keep up on any new developments.

This is the naive view that people keep pointing out that is still yet to be proven. If a company GPLs its software then it has done free development for its competitors and its users. The fact that its competitors may steal its code and embrace & extend it is not as much of an issue as the fact that they have no way to recoup their development costs unless they have some other product or service that is a cash cow. Throwing in the comment about competitors being unable to steal code is and always has been a red herring.

The fact of the matter is if an company GPLs their code then they have mostly given up on making money from shrinkwrapped copies and instead plan to make money from other means. Unless these other means are very lucrative then the company is doomed to be unsuccessful or at best marginally successful.

Companies that may benefit from GPLing code are those that GPL code that isn't their major product or is a product that they make lots of money supporting so GPLing it may just be a way to get it to as many users as possible.

[ Parent ]
yeah (none / 0) (#36)
by Delirium on Sat Jun 16, 2001 at 02:56:40 PM EST

Yeah I can see that point, and I wasn't really trying to argue that the GPL is better for companies than a proprietary development model. I was just pointing out that a company is better off releasing code under the GPL than under a BSD or similar license, since under a BSD-style license the company still has all the disadvantages you mentioned, plus the additional one that their competitors can now use the code freely. So the GPL is the best free software license, in my opinion, for a business to release its software under, but this doesn't mean that it's not still a pretty bad choice.

I do think there are a few cases where it makes sense though, such as releasing old code that is no longer a cash cow anyway.

[ Parent ]

But GPL software is already better! (3.00 / 2) (#59)
by hotcurry on Sun Jun 24, 2001 at 01:03:40 PM EST

Your allegation about software quality being threatened is pure, unadulterated FUD. Even you should know better.

[ Parent ]
What post did you read? (2.00 / 1) (#61)
by Carnage4Life on Sun Jun 24, 2001 at 05:38:45 PM EST

Your allegation about software quality being threatened is pure, unadulterated FUD. Even you should know better.

Hmmm, interesting. Where does my post claim that software quality is threatened by the GPL? Open Source zealots like imagining attacks when there are none.

Less professional development of software does not automatically translate into "GPL software is of poorer quality than proprietary software". I won't bother explaining the distinction to you since the words would be wasted on you and you'd rate the comment to a 1 and move along. I hope you feel empowered modding down the posts of a M$ employee, that's really striking a blow for Open Source ™ :)

PS: I probably wouldn't have found your post if I hadn't wondered who was the Open Source zealot who was modding all my posts down to 1. Keep the zealot fire hot, your comments and ratings are rather amusing.

[ Parent ]
So what was your point, then? (none / 0) (#62)
by marlowe on Sun Jun 24, 2001 at 05:50:07 PM EST

If you don't mind my asking.

If less professional development isn't a bad thing in your view, then why did you bother to bring it up?

Please don't try to tell us you were just making an observation. The total pattern of your posts shows your hatred of open source and its advocates. The only mystery is the motivation behind it. You've got some explaining to do.

-- The Americans are the Jews of the 21st century. Only we won't go as quietly to the gas chambers. --
[ Parent ]
Is that your other account? (none / 0) (#64)
by Carnage4Life on Sun Jun 24, 2001 at 06:23:30 PM EST

That was a really quick response to a post a days old thread. Is hotcurry one of your alternate accounts, that would make sense since your opinions and rating style seem similar.

If you don't mind my asking.

If less professional development isn't a bad thing in your view, then why did you bother to bring it up?


I didn't say less professional development was not bad, I said that it did not equate to poorer quality software.

A professional is someone who is trained and skilled at a particular task and does it for a living. A professional typically has more skill and experience than an amateur or hobbyist.

Professional software development is typically the boring crap that everyone knows should be done but rarely does. Requirements elicitation, blackbox and whitebox testing, clear documentation but for users and fellow developers, modular and extensible design, etc are all aspects of professional software development that are typically against the hacker grain. The fact of the matter is that most successful Open Source projects are primarily staffed by professional software developers and not the amateurs and hobbyists that Slashdot likes to imagine are in control of Open Source projects.

The less software houses doing professional development means that there will be less professionals who will bring that experience to Open Source in their spare time. Also the fact of the matter is that Open Source software indirectly benefits from work financed by proprietary shops either via copying User Interfaces which are the results of focus groups and usability research or by picking the brains of professional developers who have seen expensive mistakes first hand and have learned how to avoid them.

Please don't try to tell us you were just making an observation. The total pattern of your posts shows your hatred of open source and its advocates. The only mystery is the motivation behind it. You've got some explaining to do.

I don't hate Open Source. I think Open Source Software is a good idea, I just don't see it is pragmatic choice for companies that plan to primarily be in the software industry neither do I see it as a Silver Bullet which cannot be criticized and is the best thing since the coming of Christ.

As for Open Source advocates, I have nothing but disdain for the vocal, constantly irate and virulent characters that fill slashdot and have slowly begun to take over K5. I also have nothing but respect for people like Linus Torvalds, Larry Wall, Miguel De Icaza and Theo De Raalt who let their code do all the talking and don't let the adulation of teenage script kiddies on Slashdot make them believe they are more important than they are thus becoming arrogant gas bags like some other Open Source "advocates".

[ Parent ]
False dichotomy, bogus points. (none / 0) (#65)
by marlowe on Sun Jun 24, 2001 at 09:55:57 PM EST

I've done both profressional software and open source. And I'm here to tell you that not one of the things you say as peculiar to "professional" (for pay) development are either the sole purview of such or even universal to such.

You're really not fair to open source at all. And you give closed source way more credit than it deserves. I've seen how shrinkwrapped closed source PC software gets put together. It's not pretty. Inhouse development is much more disciplined, but it's also very enterprise specific. It's not of much use outside the company, so the open v closed source distinction doesn't apply.

(I'm a friendly "stalker" of hotcurry via the search page, and, I suspect, he returns the favor. Us open source zealots run in packs, and are capable of far more coordinated activity than your bigotry will allow you to admit. So watch out. You're starting to piss several of us off.)

-- The Americans are the Jews of the 21st century. Only we won't go as quietly to the gas chambers. --
[ Parent ]
My two bits (4.00 / 1) (#7)
by regeya on Fri Jun 15, 2001 at 11:05:21 PM EST

The more I think about it, the more I think a writing analogy applies.

It a bit like, I think, having one's work published. It's done with the faith that people will be able to share your work, and that people will largely respect the copyright. Sure, there will be violators, and of course, if you want to protect your work (hey, some people don't) you'll chase down blatant violations.

The closed-source method seems like distributing Cliff's Notes, for a fee, then stating that the full work is available for a much larger fee and that the recipient mustn't reveal any extra information gleaned from the full work. Oh, and it's illegal to extrapolate from the Cliff's Notes any part of the larger work not contained in the Cliff's Notes version.

And really, I think it's fair to point out other licensing agreements when someone equates open/free software with the GPL. There are other licenses out there, people choose them rather than the GPL for their own reasons, and to not point out the (in my opinion, deliberate in the case of MS) errors of the vocal minority is to dishonor a larger community of sharing individuals.

[ yokelpunk | kuro5hin diary ]

OS *is* more than the GPL; but I'm talking *tone* (none / 0) (#12)
by grout on Fri Jun 15, 2001 at 11:44:10 PM EST

My suggestion was that we avoid making "open source is more than the GPL" our only, or even primary, answer to criticism of the GPL itself. Such an answer is basically a tactical retreat, and we shouldn't give up that ground without a fight.

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

writing analogy of the GPL (none / 0) (#30)
by garlic on Sat Jun 16, 2001 at 09:45:58 AM EST

You forgot the other half of the analogy. What about the GPL side of the analogy? What about MS's legitimate fears that if they use GPL code anywhere, the entire project/program will have to be GPL'ed?

HUSI challenge: post 4 troll diaries on husi without being outed as a Kuron, or having the diaries deleted or moved by admins.
[ Parent ]

open source is not a religion (4.50 / 4) (#17)
by heighting on Sat Jun 16, 2001 at 12:20:58 AM EST

People write free software do it because they estimate that the potential prestige/goodwill that they derive from a successful project would outweigh the financial opportunity costs of releasing it for free.

Little opportunity cost

The fact is, most successful open source software are commodities and would probably not find a paying market, eg. the Linux kernel (essentially Tenenbaum's Minix), Apache web server, and almost all the projects on sourceforge.

The browser is a good negative example. It is non trivial to write a standards compliant browser, and not many people even attempt to write one. Mozilla was a mess until very recently, and still doesn't measure up to IE or Opera.

Another source of open source software is from the academic community, eg. Tex and other languages, where the developers (students, researchers) often have neither the time nor inclination to market the software.

Prestige

The amount of prestige to be gained from being associated with a successful project is huge. The lead programmers for MS Office laboured in obscurity, but Miguel Icaza is almost a hero for leading the Gnome project, which is not exactly cutting edge software. Similarly, Linus' prominence probably got him his job at Transmeta.

My two cents: write whatever software you feel like writing, sell it if you can, give it away if you can't. Just stop expending so much energy over meaningless licensing issues or pretend to have some sort of moral high ground.

Preach it, Brother! (3.00 / 2) (#19)
by grout on Sat Jun 16, 2001 at 12:40:03 AM EST

Prestige from working on a project goes up with several factors, including particularly the number of users it gathers. So if a license change can increase the chance that a given project will achieve dominance in its niche, well, then, the license really does make a difference, even if you're just out to become a Big Man On Campus and impress the chicks. :-)

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Profit is not a religion either (4.00 / 1) (#33)
by MrMikey on Sat Jun 16, 2001 at 11:42:39 AM EST

People write free software do it because they estimate that the potential prestige/goodwill that they derive from a successful project would outweigh the financial opportunity costs of releasing it for free.
Unless you've got Cerebro in your basement, and are able to scan the minds of each and every person who writes software for free, isn't it just a little bit pretentious of you to assume you know the motivations of every, or even most, of those who write free software?

I've got a project I'm considering. If and when I finish it, I'm releasing it for free. Could I generate profit by selling it? Maybe. Will I? No. Will it earn me prestige? I very much doubt it. Will it add to the body of useful free software? I certainly hope so... I'm sure someone will have a use for it. That's good enough for me.

The Apache web server "would probably not find a paying market"? Care to support that assertion?

[ Parent ]

I wasn't making a value judgment, (4.50 / 2) (#34)
by heighting on Sat Jun 16, 2001 at 12:35:18 PM EST

so don't feel pressured to go on the defensive.

It's pretty well documented that prestige is a very large factor in releasing free software, just type 'free software prestige' into google.

I don't know what your project is, care to talk about it? If your software can potentially make millions, and you are not opposed to putting in the effort to make it succeed, do you honestly think you would release it for free?

Apache

Regarding Apache, I'm not saying it's not a good piece of software, just that the economics of the niche makes it a hard sell. First, there is a huge demand for web servers. Second, web servers, even a relatively full featured one, are not really hard to write. This means that any commercial web servers without some proprietary edge will be swamped by lots of free substitutes.

The commercial web servers on the market have special advantages: IIS is backed by MS and is well integrated with W2K, Zeus is highly optimized and very much faster than Apache.

[ Parent ]

OT: genetically wired for reciprocal altruism (3.00 / 1) (#21)
by Justinfinity on Sat Jun 16, 2001 at 01:26:46 AM EST

i think the reciprocal part is done by society. if left to nature alone, as it were, we would be wired for truism alone. of course that truism is strongest for self, but can sometimes be overridden by things such as love, fear, mourning, etc.

we've all heard the stories of people willing to give there own life in an attempt to save another life. instincts take over. the reciprocality programmed by society goes to the wayside. they don't expect anything back from the person they're trying to save. nature takes over and pure truism takes over, directed at the person whose life is on danger.

perhaps we'll never know why people do this. perhaps we'll find out that it too is a by product of society. who knows? let's find out.

-Justin
If this is all a dream, please, don't wake me.
Got Water? Money sucks. Groove.
:wq


"Altruism" is in quotes on purpose (none / 0) (#22)
by grout on Sat Jun 16, 2001 at 02:07:01 AM EST

i think the reciprocal part is done by society. if left to nature alone, as it were, we would be wired for truism alone. of course that truism is strongest for self, but can sometimes be overridden by things such as love, fear, mourning, etc.

Ah, but how did the emotion of love evolve? What survival purpose does it serve? One reasonable answer: It cements relationships of mutual reciprocation.

perhaps we'll never know why people [sacrifice themselves for others]

I really think you'd enjoy reading the book "Nonzero". It sets forth a plausible answer for this and other mysteries. (Not that Mr. Wright did the original research. He pulls the work of many into the book's far-ranging narrative.)

BTW, kin selection is also at play in some sacrifices. If you give your life to save a close relative, you may be enabling enough of your genes to go to the next generation to make your sacrifice "worthwhile" in a genetic sense; and that may be enough of a reason for kin selection in its current form. But I'm not a biologist, so I'll leave further explanations to the experts.

--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Missing a key point (3.33 / 3) (#35)
by B'Trey on Sat Jun 16, 2001 at 01:24:59 PM EST

You, along with those who claim that Open Source is socialism, are missing a key point. Open source is NOT driven by recipricol altruism. It's driven by pure selfishness.

Linus did not sit down and write the Linux kernel because he thought it might help someone else, who would later help him. He didn't write it to become famous or out of the goodness of his heart. He wrote it because he was in college, he needed an OS for his personal use and he couldn't afford to purchase one.

Sure, people groove on being known in the Open Source community. They share their source in the expectation that other people will do the same. But, at best, reciprocal altruism drives the distribution of open software. It doesn't drive it's creation.

Software is also about the only area of endeavor where these metrics work. If I go out and sweep the street in front of my house, it does nothing to clean the street in front of your house. And chances are very slim that I'm going to go sweep your street in hopes that you'll later come mow my lawn. OTOH, if I write a program for my own use, it cost me nothing to make it avialable for your use as well. I don't have to put forth extra effort. I can distribute my program, either from pure altruism, reciprocal altruism, in search of prestige, or for any other reason, with no additional effort on my part.

_Impure_ selfishness works better (4.33 / 3) (#38)
by grout on Sat Jun 16, 2001 at 06:35:15 PM EST

I think we agree, basically. Pure selfishness works well, but impure selfishness works better from an operational perspective; that's the point of the "non-zero-sum games" that I referred to. The term reciprocal altruism means giving to someone else with the expectation that it will be worth your while, i.e. that you will receive an adequate reward in the future. That's the reciprocal part. It's sometimes written as reciprocal "altruism" to emphasize that the "altruism" isn't really altruism in the long run.
--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]
Oh, and software isn't unique.... (3.50 / 2) (#40)
by grout on Sat Jun 16, 2001 at 07:26:52 PM EST

There are many goods people can distribute at no incremental cost, now that we have computers, starting with text and pictures. Anything that is made of bits can and will be copied. ("Trying to make bits uncopyable is like trying to make water not wet." -- Bruce Schneier)

Software is unique because is the first functional good with zero incremental distribution cost.
--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

parent comment has bad title (none / 0) (#41)
by grout on Sat Jun 16, 2001 at 07:29:19 PM EST

First I say software isn't unique in the title, then I say it is in the body. *sigh* Note to myself: "Preview is your friend."
--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]
Certainly... (3.00 / 1) (#44)
by B'Trey on Sun Jun 17, 2001 at 10:25:57 AM EST

...data can be distributed at essentially zero cost. But what forms of data other than software are generated largely for one's own use and are also useful to others? Text and pictures are normally either created for the express purpose of distribution or have little or no appeal to others.

[ Parent ]
Question is too limited (none / 0) (#54)
by grout on Wed Jun 20, 2001 at 06:06:48 AM EST

Do I work on Perl because I use it? Even if I were unable to keep using Perl for some reason, wouldn't I want to keep it up just because of my personal investment?

On the other side: There's a long history of distributing text and pictures at high cost. So whatever is new about software, the fact that it's valuable (per that history) and zero-incremental-cost copiable (like .text and .jpeg files) is nothing unique.
--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Early Linux and the GPL method (none / 0) (#52)
by simon farnz on Mon Jun 18, 2001 at 12:43:14 PM EST

Of course Linus wrote his kernel for his own benefit; when he released it under the GPL, it was largely unusable except as a CS curiosity (unless your machine matched Linus's).

The reason that Linux is now seen as a threat to MSFT is that the early GPL releases provided a base for a Unix-like OS; other people enhanced Linux under the GPL so that it matched their needs, and the project snowballed.
--
If guns are outlawed, only outlaws have guns
[ Parent ]

Pure selfishness? Nonsense. (none / 0) (#63)
by marlowe on Sun Jun 24, 2001 at 05:56:51 PM EST

If there were nothing more it than selfishness, Linux would never have bothered to make a relase. He'd have written Linuxand then kept it to himself. Or maybe tried to sell it.

In which case it wouldn't have been open source. So yes, altruism did create open source as such.


-- The Americans are the Jews of the 21st century. Only we won't go as quietly to the gas chambers. --
[ Parent ]
Okay (2.40 / 5) (#42)
by spacejack on Sat Jun 16, 2001 at 08:27:35 PM EST

For all the talk of "trust", "discouraging parasitism", "helping culture", I would really like to know why anyone should care about promoting the GPL -- a license that is only really beneficial to corporations and a few programmers -- above all other licenses.

I don't know anyone besides programmers and corporations who uses a single piece of GPL software. It just hasn't produced anything the public wants to use. How can it possibly displace proprietary software that bends over backwards to meet the public's wants?

It's not that I think the GPL has no place; for some developers it can be beneficial for certain types of software (eg, code feedback, improvements, commonly re-used tools like compilers or web servers, fame, closed-license sales). But as soon as you move away from things like software development tools, or, say, database applications that megacorps like the RIAA use to predict the public's purchasing habits and control the music industry, its value dwindles to zero. In other words, since the public is either incapable or uninterested in contributing back to the software they use, then I think we should be perfectly satisfied to have them contribute in the form of cash; i.e.: 1 copy == 1 sale. What could be more fair than that?

Bottom line is, it's just another license that has allowed certain types of academic/corporate/programmer-friendly software to flourish. It has made very little impact on traditional proprietary software for the public, which has always existed just fine with or without the GPL's existence. If the whole industry suddenly decided to go all GPL, the public would dump their computers overnight.

GPL software (4.00 / 2) (#45)
by FnordLord on Sun Jun 17, 2001 at 05:53:31 PM EST

I am neither a programmer nor a corporation, but right now I am using a variety of GPL programs: 1. Konqueror 2. WindowMaker 3. wterm 4. apt (maybe GPL, not sure) In addition, I frequently make use of GNU utilities such as BASH. Why would people throw away their computers if the license became less restrictive?

[ Parent ]
some mal-formed sentences (none / 0) (#48)
by spacejack on Mon Jun 18, 2001 at 08:42:25 AM EST

I should've said if everyone had to switch to GPL software, I didn't mean GPL'ing existing software (neither of which is going to happen anytime soon anyways, so that point is pretty hypothetical).

But thanks for the insight. I honestly haven't met anyone that uses those apps besides developers. I was also in a "western" mindset; perhaps Moscow residents (if that's where you're from) are using more GPL software than we are here.

[ Parent ]
Argument from ignorance. (5.00 / 1) (#60)
by hotcurry on Sun Jun 24, 2001 at 01:06:33 PM EST

It so happens that most internet web and mail servers run on open source code. If you don't happen to know that. it's your problem, not open source's problem.

[ Parent ]
us vs them analyses are faulty (2.50 / 2) (#46)
by passer on Sun Jun 17, 2001 at 10:09:22 PM EST

because the above analysis fails to see that there is a significant portion of the open source population that works on writing closed source software during the day and open source by the night.

Any analysis, that doesn't take this into account is fatally flawed, IMO.

One Person, Many Roles (none / 0) (#56)
by grout on Wed Jun 20, 2001 at 06:17:54 AM EST

My first take on this is that there's more than one "me". When I work, I have to look at things from the company's point of view as well as my own... to some degree that's what I'm paid for. But on my own time, I'm looking out for my own best interests, both short- and long-term, without corporate interests interfering.

Think of it as Dr. Jeckle and Mr. Hyde, but with Free and Proprietary as the personalities. My essay is about both personalities, but mostly Dr. Free. Mr. Proprietary is involved only in the discussions of corporations GPLing some of their open source work.

PS: This also points out that free time isn't free. My spending time on Open Source carries a heavy opportunity cost, as there are many things I don't have time to do. Right now I could be up in bed, getting rest for tomorrow, but instead I'm contributing to the Open Source culture....
--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Good piece but not too many details (2.00 / 1) (#47)
by mark2776 on Mon Jun 18, 2001 at 06:29:05 AM EST

A few comments: 1. To all of you that think that only society could be responsible for altruism - you are wrong. Altruism has evolved in nature (and is even present in the most vicious animal - the human). 2. I suggest you read the following book - "The evolution of co-operation" by Robert Axlerod if you want to understand how can altrusim evolve. 3. It is quite obvious to anyone in the software business by now that in 5-10 years ALL the software that will be on the commercial shelf will come on the first CD of the linux installation. Goodbye software companies!!! Its time to say night night... Mark.

Another reason to use the GPL... (4.66 / 3) (#49)
by simon farnz on Mon Jun 18, 2001 at 09:50:37 AM EST

... is if you have a utility that is useful to you, but needs a lot of work to become commercially saleable; you can release it, and gain the benefits of other people's improvements. Eventually, we will get to a point where it is not worth starting a software project under a different license. The gains from using the existing GPLed base will more than outweigh the cost of placing your code under the GPL.
--
If guns are outlawed, only outlaws have guns
Also, read up on the GNU Project (none / 0) (#50)
by simon farnz on Mon Jun 18, 2001 at 12:10:49 PM EST

Indeed, this is the point of RMS's GNU system (I happen to run a GNU/Linux system at home for all my work, mostly accounts and wordprocessing). He said that once the GNU system was complete enough, no one would start a non-free system, as it would be better value to enhance the free system. As a result, competition would move elsewhere in the system.

See the various papers linked from www.gnu.org for more ideas along this line
--
If guns are outlawed, only outlaws have guns
[ Parent ]

Faulty link in previous message (5.00 / 1) (#51)
by simon farnz on Mon Jun 18, 2001 at 12:15:19 PM EST

Link should be www.gnu.org
--
If guns are outlawed, only outlaws have guns
[ Parent ]
Trust not important in a 0-scarcity environment (5.00 / 1) (#53)
by Zimran on Mon Jun 18, 2001 at 05:18:12 PM EST

Good article.

Nevertheless, trust is not important in a zero scarcity environment. Parasitic community members that take but do not give back are problematic in alturistic societies where the good being traded is limited in some way (ie. food). So, if there is a limited amount of food available, it makes sense that it should be divided amongst those who created the food, and not to parasites.

But with software, one person using another's code does not deprive anyone of that code. Free-riding is *not* a problem with open-source, and so the GPL is not neccessary for open-source to continue.

However, those who create open-source *do* deal in a different limited good -- fame (see HtN). If the GPL somehow helps distribute fame appropriately, then it might be necessary.

Also, please note that just because the GPL may not be necessary does not mean it cannot be useful.

Zimran

Open Source is a High-Scarcity Environment! (3.00 / 1) (#55)
by grout on Wed Jun 20, 2001 at 06:10:47 AM EST

Sure the code is free, but programmer time is excruciatingly valuable ... especially talented programmer time.

So programmers don't want to contribute to free-rider-prone projects, all things being equal, if a free-rider-proof alternative exists or could be created.

Mind you, I'm talking about trends here. Individual programmers may well avoid GPL projects for philosophical reasons; and so the non-GPL projects serve a useful purpose if only because they capture those GPL-averse volunteers.
--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

Why Now?? (4.00 / 1) (#57)
by gbvb on Wed Jun 20, 2001 at 09:33:26 AM EST

I wonder what changed in the past 6 or 8 months that MS is trying to work so hard towards making GPL visible to the end users. Its not like an average user would care one way or the other..
GPL has been a Software developer's thing. Only they knew and cared about the licensing of the software. Why is it that MS is going gung-ho against GPL? I think there are a couple of possibilities.
1. Linux and other GPLed software is gaining too much importance and MS is uncomfortable with that.
2. MS wants to sell .Net web services as a way of providing services to the customers without them thinking about GPL/BSD/Mozilla licensing.
3. MS does not want to fight SUN or other unix server vendors. So, it is trying to pick a fight with Linux which seems to be the easiest target because of what it stands for.
4. MS is trying to justify the cost of Windows XP and .Net services.
I wonder which one this might be..??

GPL is a convenient place for FIRST divide&conquer (none / 0) (#58)
by grout on Fri Jun 22, 2001 at 04:28:41 PM EST

MS knows that their current business model has peaked. They want to move into a rental model, as everyone can see. For the first time, then, Open Source software is standing in the way of a Top-Level Microsoft Strategy. It's no wonder they're bringing out the FUD guns.

"Divide and conquer" is just as venerable as a war strategy as "embrace and extend." The community's gradual division into the "Open Source" and "Free Software" identities provided MS an obvious place to aim its first FUD attacks. Fortunately for us, they misjudged the defensive reaction they have provoked. It's too late for MS to play nice, anyway; we know too much about them. I'm reminded of a scene from The Princess Bride, which I here edit slightly: "I could give you my word as a Microsoft representative." "No good; I've known too many Microsoft representatives."
--
Chip Salzenberg, Free-Floating Agent of Chaos

[ Parent ]

The GPL: A Technology Of Trust | 65 comments (54 topical, 11 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!