Edited by author 06-10-2003 01:36 PM
Yoz-- http://www.sweetcode.org is great-- thanks for the tip! Excellent point about how the really significant stuff usually comes out of academia, not the underground. I'd call that revolution from within, which is more "filtered" and probably has a better track record.

BTW, for a great portrayal (and implementation!) of a talented aspirant's struggle to achieve fame and acceptance by challenging authority, see the movie "Cremaster 3" -- I just saw it last night and totally loved it. It's more about industry and wealth (in the first part) and the art world (second part), but it's absolutely the same eternal tale as with OSS software, or any other proving ground. Highly recommended!

Concerning "Cremaster 3," which drew much of its imagery from Masonry, I would add that I see *huge* parallels between the OSS community today and the stonemasons of the early gothic period-- the ones who built the most impressive cathedrals and bridges, complex structures which everyone used and appreciated but which few people understood. The big change came after much of that knowledge was lost, and the buildings necessarily became less structurally impressive but more ornate and filled with symbolism (late gothic). The tradition that, centuries later, became capital-M Masonry continued as an elite, but it necessarily shifted its focus from actual talent and expertise in engineering to a system of arbitrary symbols and signs that worked primarily to retain power for insiders-- the "old boy" network. But that's another story...

WRT TSMO's commentary on what drives OSS development:

I don't altogether buy the "replace greed with vanity" argument. First of all I don't think that it is necessarily bad, as the tone of TSMO's comments seems to imply. People get motivated to do things for all sorts of reasons. The same arguments presumably apply to, for example, bloggers, but I don't think "vanity" or "attribution" is the *only* reason they do this, which is my second point. From time to time I have made very minor contribution to OSS (mostly bug fixes). I don't *have* to waste time packaging them up and sending them in, but I do it because I feel like giving something, however small, back to the "community". I neither know nor care whether my name gets attached to any of this. Having said that, when I make a bigger contribution to some community (not necessarily OSS) I feel that it is polite and proper that my efforts are recognised. Call that "vanity" if you will, but the absence of recognition is generally referred to as "taking people for granted". Communities that fail to commune don't usually work very well.

oh christ, ye olde licensing argument ;)
PS: to do a Spartacus -- I am an idiot and/or a troll!

Edited by author 06-09-2003 01:22 PM
Sorry, Paul, I think I misread your post - you weren't claiming that all OSS contributors are from the underground. I agree that OSS is a big attractor for hackers wanting to make their name with "revolutionary" software, though most of this tends to fall under the categories of either "clone of something from the proprietary world" or "nifty P2P app". It may cause a revolution in the literal sense, in terms of end-users switching away from proprietary apps, but it's hardly a revolution in ideas. For that, you really do want academia: check out the archives at http://www.sweetcode.org/ for the kinds of stuff that rarely makes the front page at Slashdot but is utterly fascinating, and which mostly comes from research projects. (But, yes, I take your point about revolution in industry.)

Fangs is both an idiot and a troll. Way to go, fangs!

But I'll bite/feed anyway. Go on, fangs: Explain to me why offering the source of a derivative work is equivalent to an attribution to the original work. (I assume that's what you mean, anyway. If it isn't, then what the hell are you talking about?)

Regarding BSD: Yeah, okay, you got me there, but the vast majority of BSD-ware in circulation still uses the original license. (And the "attribution" clause is much more in the spirit of the original BSD than the GPL)

Paul: I still disagree that it's anything that simple as "just fame". Firstly, you're doing a disservice to those who are working on OSS with genuinely altruistic motives. And then there's the other lot I mentioned, the corporations: they're self-serving, but fame alone doesn't put bread on the table. (Often, corps will release non-core-competence code that they use because they know/think that public contributions and maintenance of that code will ease their burden)

But the bits I take most issue with are:

For up-and-coming hackers, the most appealing projects are ones which outwit the people and institutions in charge, and bring power to fellow hackers outside the system.

and

The people who write it are smarter and more dedicated.

This basically claims that the vast majority of OSS contributors are part of the hacker underground. On the contrary: A huge amount of OSS comes from academia, from commercial ventures, and from people for whom programming proprietary software is their day job. (It's not like they suddenly shoot up a few IQ points when they go home.)

Actually, Manuel Castells identifies this (donate to the common software base for the sake of attribution or "fame") as one of the four founding cultures of the Internet. BTW, his four founding cultures are: Academic (currency = peer recognition and attribution), Hacker (currency = elegant solutions to problems, peer recognition, attribution), Communities (currency = mutual support, continuity), Entrepreneur (currency = make money fast - more or less). Naturally, I simplify his complete argument and derivation. However, we see that three of the four are thriving on the Internet today; moreover, those that are using the Internet for commercial advantage are, by and large, not "Internet businesses" per se, but rather companies that use the 'net to enhance their ability to do their conventional businesses - and this applies to the entire spectrum that runs from huge to tiny enterprises.

Yoz is either an idiot or a troll.

GPL'ed software requires no attribution. All it requires is that derivative products are licensed against the GPL as well.

No. Section 3 of the GPL requries that if you distrubute GPL software in binary form, you accompany the software with either the source, or a written offer for the source. All GPL'ed binaries, modified or not, must be attributed as such. http://www.gnu.org/licenses/gpl.txt

The BSD license effectively comes down to "Do what you like, but attribute".

No. The BSD advertising clause was removed on July 22, 1999.

You can read the message sent from the UC Berkeley Office of Technology Licenscing retracting the advertising clause here: ftp://ftp.cs.berkeley.edu/pub/4bsd/README.Impt.License.Change

Fame all the way. First corollary: the OSS movement is a meritocracy-- as much of one as, say, competitive sports or chess. It's also about revolution and the succession of generations. Here's a paragraph from a book proposal I wrote about a year ago (which was never picked up, maybe because it was too overblown):

> As with other fields, the top talents in software enjoy showing off how much smarter they are than everyone else. Open-source software, by its very nature, is a magnet for such personalities. The cream of the crop will always do its best work for the competitive challenge and the glory, not for a regular paycheck. For up-and-coming hackers, the most appealing projects are ones which outwit the people and institutions in charge, and bring power to fellow hackers outside the system. This dynamic is the archetype of heroism and revolution, and it will never change, and the most potentially subversive, revolutionary open-source projects will always be considered the Big Leagues. That’s why, despite the money behind commercial software, the free underground stuff will triumph. The people who write it are smarter and more dedicated.

"fangs" has it exactly wrong:

GPL'ed software requires no attribution. All it requires is that derivative products are licensed against the GPL as well.

The BSD license effectively comes down to "Do what you like, but attribute".

However, the line from TMSO that fangs quoted was similarly nonsense. Most OSS products rely on hundreds of popular open-sourced libraries that are never credited anywhere in the documentation apart from the "Requirements" section. And while it's true that forking someone else's GPL code without credit is considered impolite, that's all it is - as opposed to a breach of copyright that'll get you sued. (This is why Microsoft's continued lecturing on the "infectious" evils of the GPL is so funny: Try creating a derivative work of MS Word and redistributing it under an MS license, and see how long it takes for the lawyers to show up.)

Similarly, the idea that OSS has nothing to do with community is also nonsense. The antagonism that TMSO describes would surely happen in any community environment where members felt that their trust and contributions had been abused. It's true that many developer communities have an implicit hierarchy of ability and achievement, but if it were just about ego-driven self-interested ladder-climbing then no one would ever lower a hand to help someone up, for fear of competition. On the contrary, thousands of OSS developers help each other like this all the time, in situations where the only lasting monument to an ego is a post on a mailing list or a transient few lines in an IRC chat.

Fact is, everyone has a different reason for contributing, which is why you get communists and objectivists working on the same projects. While neat little packaged ideas such as "egoboo" provide tempting bait for those trying to encapsulate the OSS movement (and provide easier targets for criticism), it's just not that simple.

(And we haven't even touched on the reasons why corporations release OSS)

Edited by author 06-08-2003 07:14 PM
Does anyone but me think that the word "troll" is so overused as to be ineffective as a supposed insult? And it's also interesting to see someone make an argument starting with an ad hominem attack. Usually you slip those in later. Only an idiot or a moron would start with one. (See how I did that!)

Chris... it isn't just selfishly taking from the public domain. It's also selfishly (but somewhat altruistically) giving to the public domain so that you can get the features and fixes that you want. The nicer phrase to use is 'enlightenned self-interest'.

We can compare the motivations for contributing to open source to a lot of other things, but ultimately the way things work are unique --similar, yet different. Like it or not, self interest is, has, and probably always will be the biggest motivator of humanity.

As for fame, if we really all wanted to be popular or superstars, there are other more obvious, far more direct ways to do so that don't involve staring at a monitor 12+ hours a day.

Is open source about altruism or fame?

It's about selfishness. I don't have to re-invent the wheel if there
exists a public domain where I can borrow libraries of code to
scratch my own itch with.

I can get the job done quicker by standing on the shoulder of
giants.

That's also an incentive to give back to the public domain -- to keep the symbiosis going.

Whether or not I become famous doesn't matter. Whether or not someone else benefits from my work doesn't matter. The fact is, I was able to selfishly TAKE from the public domain for my own benefit.

Either John and TSMO are idiots, or they are trolls.

Try to use someone's OSS code without attributing original authorship, and you will see how quickly your quaint "community" devolves into harsh campaigns of public remonstration towards the violator. It is of primary importance that original authorship always be identified.

This might be true for for GPL'ed code, but this isn't true for the hundreds of millions of lines of code under different OSS licences, like the BSD licence. Such OSS code can be used freely, without attribution.

mercer says: Wow, I'm glad someone finally got through to you about this John, we've all largely assumed it as a given!

I doubt you think this is true. And it's not that I never knew this. It's that I never thought about it in these exact terms. For some reason the way it was put triggered some interesting concepts that I'll be boring people stiff with for years to come!

Fame is overrated... and not the primary reason why most people are in open source.

For me, I think my biggest motivation was that I can be a bit of a perfectionist. I saw an app that I found useful and cared about, but it had these niggling little problems that drove me up the wall. A big part of my motivation was selfish --I wanted things to work the way I thought made sense-- and part of it was altruistic, in that it would benefit everyone who used the app... but reputation or fame was not even a factor in my initial decision.

I know that getting credit for your work is something that everyone appreciates, but people appreciate getting credit within corporations, too. Credit, altruism, and the ability to implement changes, however, are all important motivators in open source... or for that matter, any kind of "volunteer work", just because money usually isn't a factor.

When you are overseeing any open source application, this is abundantly clear, because actually getting people to follow through and help your project is so damn hard. It's like pushing a rope, instead of pulling it. As such, you will grasp onto anything that you can use to motivate others to assist the project.

Now, when you're a Linus Torvalds, I'm sure you probably don't have to humor, cajole, or beg quite as much for people to assist you, but you have to remember that only the biggest open source applications are in that position. The larger the app, the greater the motivation for people to help, I'm sure... and yes, that *IS* influenced more considerably by reputation. Then again, money tends to be more of an influence too.

Respect from one's peers is nice at times, and when you get it from some of the better known people out there, it's even better. For an awful lot of people, Linus Torvalds is a role model and a hero... and the same applies to a few others in the open source movement.

As for "fame", however, there is precious little of it out there and it's only fame in very small circles. As they said in "Sixteen Candles", it's a bit like being "king of the dipshits".

I am considered a "famous LiveJournaler", but that fame is only of any note to a handful of fellow volunteers and to perhaps to a few percent of the LiveJournal userbase. It doesn't resonate much outside of that. Most of the fame is concentrated on the initial founder of any application... and I can tell you that they invariably never expected to get that level of fame. It can be a burden, frankly, because all decisions (and complaints) regarding the application tend to bombard the founder, or at least flow to them.

In that sense, a hierarchy based on reputation can actually be a limiting factor in the growth of an application. Open source, for all its presumed benevolence, can be terribly undemocratic and even dictatorial at times. Sure, people can always fork the code, but more often than not, forked code indicates a fundamental systemic failure based on people feeling like their concerns are being ignored; it indicates a lack of flexibility and scalability on the project's organizational level.

For all of the emphasis on opening up the code, there is usually precious little emphasis on how groups of people should interact to make decisions. Linus Torvalds is clearly not always right, in the same way that Gates is not always right. It also sets you up as a bottleneck and a gatekeeper, who is more prone to accept contributions from the best known over the least. What that can do, however, is discourage new contributors to the project.

The ultimate goal of any open source project should be to harness the goodwill that others have for your project and to turn that into code. In that sense, I still think that most open source projects are realizing only a fraction of their potential... their decision making structures are usually simplistic and inappropriate for encouraging the widest degree of contributions, and do not scale well.

In my opinion, one of the most important things missing that would greatly benefit the open source movement are better applications for helping large groups of people work together online, share ideas, interact, and make decisions. The easier, clearer, and more transparent the operation of a project is, the easier it is for new contributors to help out.

This needs to be based upon some kind of common standard for how decisions should be made, of course. If open source could use anything, it is this: rules to make rules.

Which, incidentally, sounds like another open source project...

You neglect to mention that most larger open source projects (Linux, etc.) were created for personal use. Their release to the public was relatively altruistic; go look at early USENET posts on Linux. The continuation of their development may be done for fame, but they were created because the writer wanted something, and released to everyone else for altrustic purposes.