The Start of Open Source

I have been researching the origins of open source recently and realized that I had missed an important anniversary last year.

On a very rainy day in early February 1998, a group of people very familiar with free software met at the Palo Alto home of Christine Peterson of the Foresight Institute. Many in the free software movement felt that they were on the verge of something very big. Netscape had just announced that it would make its source code freely available. Influenced by an article by Eric Raymond called the Cathedral and the Bazaar, the management at Netscape came to the conclusion that this was the way to build software. Chris invited Eric, Michael Tiemann, Larry Augustin, John Hall, Todd Anderson and Sam Ockman to session to discuss the unique opportunity of publicity this would create and how best to present the free software movement to business as a whole. Chris's living room provided a venue to brainstorm on new ways to brand free software.

The heavy presence not in that room was Richard Stallman - RMS. Richard Stallman provided a Patrick Henry-like defense of free software in a "Give Me Liberty or Give Me Death" sort of way. Although the group agreed and aligned with the principles of free software - free to share, free to choose, free to reuse, free to distribute, RMS's uncompromising stance on the term "free software" inhibited business users from taking up free software. Although business users of free software, particularly younger, early adopters, could agree and sympathize with these principles, they were suspicious of anything free. The term was too closely related to freeware or shareware that was usually a one-man outfit that relied on the contributions of those who liked the software. Freeware did not mean that the source code was freely available, so meant that there was generally no one else to work on the product to improve or fix it. RMS felt that this called for education, not stepping away from the term free that emphasized the principles of freedom.

The concept of open and free software has actually been around as long as their have been computers. Universities, in particular, had freely shared software and collaborated between each other to create new programs and new software systems. The Unix operating system and its successor Linux owe much to this early open, collaborative and free development of software. However, up until this date, the closest thing to describe this concept, process and set of values was the Free Software Foundation and the principles listed on its web site. However, the confusion of the word "free" and stridency of the Free Software Foundation's founder, RMS, was putting off business people. All the people in the room had experienced the frustration of pushing free software. As Sam Ockman pointed out, "People are cynical; they expect higher costs of ownership with anything that is labeled as 'free.' 'If I don’t pay now, I’ll pay later,' was a common mindset I encountered from IT buyers." Even worse was "Free? That sounds like communism!" I have to admit that I fell into that camp.

The group in Palo Alto felt that presented with an opportunity as big and important as Netscape adopting an approach of free software development, they must make the most of it. They had two objectives, to help make Netscape successful in its venture into free software and how could they take advantage of the publicity surrounding this release. With the second goal in mind, they took the approach of essentially rebranding free software. Eric Raymond, who had authored an influential essay on free software, The Cathedral and the Bazaar, felt the traditional term, "free software," had been a millstone around all of their necks, and was a nonstarter as rhetoric to convince any but the hard-core believers. Michael Tiemann had been running the oldest company selling free software, Cygnus Software, and was equally frustrated. As Michael put it, "We wanted a term that was more resonant with the business benefits rather than the moral arguments."

The session was not very long, perhaps about two or three hours. Michael Tiemann advocated the "source ware". Christine Peterson, a futurist from the Foresight Institute, liked the term "Open Source". Eric Raymond carried a lot of influence in this discussion. He was the one that had helped to persuade Netscape to go with free software. He was also an articulate spokesperson for the development methodology and a self-professed extrovert. Eric liked the term "Open Source" as well and open source carried the day.

A few days later, Eric raised the call to dump "free software" and start to use "open source".  The divide between Free Software and Open Source could not be clearer. Later that month, the Open Source Initiative was formed by Eric and Bruce Perens. Bruce's definition of principles of Debian Linux were used as the guiding principles of Open Source. In April, Tim O'Reilly brought together all the influencers and thinkers of the free software movement including Linus Torvalds of Linux fame and Brian Behlendorf of Apache. As is typical of open source and legalese, the passive statement was issued that "a vote was taken" to call the movement "Open Source". Tim called his conference the Open Source Conference from there on. RMS still hung on to the importance of freedom and the term "Free Software". He also complained about being "written out of history." However, the principles had generally not changed, only the tone, just as a new ruling government would take after a revolution.

With Netscape now open source and everyone other than the Free Software Foundation using the term, open source really took off. By coinciding with the massive explosion of internet and the most widely used software of the web being the open source Apache web server, open source could only accelerate. It also didn't hurt that on May 14, 1998 that Janet Reno caught Microsoft completely off guard by filing anti-monopoly charges against Microsoft. This set a white hat / black hat positioning that continues to galvanize the open source community. Although it would take some years for Mozilla, the reformed Netscape browser, to take off with Firefox, to take off, the launch of open source certainly caught IBM's attention. A little over a year later, IBM would commission the Bowen report and decide to move their tools and Unix businesses toward open source and Linux. Red Hat launched its enterprise business after acquiring Cygnus and really displace both existing proprietary Unix system and Windows systems with an enterprise grade Linux.

The business momentum stalled a bit after the Dot Com Crash, but open source did not miss a beat. The value of open source became crystal clear to a lot of people in the constraints of a recession, just as they are doing now. Open Source didn't really start with the name open source, but it certainly accelerated from that point and the timing couldn't have been better.

Comments

[TechUpdates]

Cloud Computing and Grid Computing are certainly making prominent entry into the low level server segment. Nice post got to know a lot of details.

[Thomas Lord]

I'm happy to try to help, John.

Before we ask about the history of "open source" we ought to ask what "open source" stands for in the world and how that compares and contrasts to what "free software" stands for.

RMS' widely adopted working definition of "software freedom" is an expression of a set of values - some moral beliefs. It says:

"All users of software should have access to the source code of that software and the freedoms to run it as they see fit, study the source, modify the source, and share the source with and without modifications under these same conditions." -- free software in a nutshell.

"Open source" is also an expression of values. It says "*Some programs* should be available for anyone to run, copy, study, modify and share as part of a development method that solicits unpaid labor thus lowering costs while at the same time exploiting peer review to ensure higher quality results".

Notice how those are different: the free software movement aims to establish, protect, and defend the rights of all users. The open source branding campaign discards concern for users sets a very different goal of saving some money on software development.

You may wonder: "Yes, but if everyone does Open Source then don't all users have software freedom?"

The Open Source Initiative folks and Bruce Perens, the author of the Open Source Definition, and Eric Raymond can answer that for you: they *encourage* the use of Open Source development as a method for lowering the cost of developing non-open-source systems. They are often heard to argue that sometimes users should not enjoy their software freedom. Mr. Perens has said publicly that he sometimes consults with large firms to help them understand how they can legally develop proprietary systems using open source for parts of those systems, for example.

I see another example just in quite recent news: ESR arguing that the GNU Public License is a burden and that the Linux Kernel should embrace more proprietary software (software of which users are not free to study the source, etc.). See http://www.osnews.com/story/21192/ESR_GPL_No_Longer_Needed

Now, in that article you'll see ESR's argument that "the market will punish" people who develop proprietary software with Open Source parts but that is a wildly disingenuous argument: the industry is rife with counter-examples (think MySQL or of countless "network appliances" on the market). There are few examples in support of his position and they are on the margins. And he contradicts himself as in the interviews for this 2004 article in which, like Perens, he entertains a proper "role for proprietary software": http://news.zdnet.com/2100-9595_22-135953.html
though he confidently predicts that that will amount to "but 5% of the market". (On what basis this prediction is made with such confidence in the face of contradictory evidence is not explained.)

The difference in these two positions is critical in this way (among others):

If you believe in software freedom then you have nothing kind to say about mixed open/proprietary systems such as how MySQL was run for so long. You don't point to those guys and say "They are on our side. They are fighting for our cause." You instead say "It is good that they are doing some of that program in freedom but when they add the proprietary parts, just to make a buck at the expense of the freedom of their customers, they are behaving unethically."

If you believe in the open source method of cost savings you point to "MySQL" and say "See, those guys are doing the right thing!"

They are deeply different positions.

And so that brings us back to your history account. There are two points to consider:

First, as a matter of logic, "open source" is not simply an alternative brand for "free software". It is a deeply different position, as illustrated above.

Second, as to the matter of intent, we can only make reasonable inferences. The man who takes the unusual step of bringing a gun to a nightclub and then kills his estranged boyfriend there has a hard time arguing that the crime was committed in the heat of passion: it is by all reasonable standards a pre-meditated act.

Did the "open source" folks *intend* to suppress care for the freedom of users?

As circumstantial evidence, the days-later attack on RMS and the free software movement and the relentless continuation of such attacks ever since suggests that, yes, they did.

As further circumstantial evidence we can look at the various business models that were popular with investors at the time and increasingly over the next several years. Many-a-plan was funded with the model "open source part of this for development cost savings but also develop proprietary code for the benefit of the investors". In light of that circumstance, isn't it interesting that at that meeting were two darlings of the VC world, hosted by a board member of Foresight, with insider knowledge of the impending Netscape release, for the purpose of strategizing on the coordinated front that community would put up to marginalize RMS and the free software movement?

It was not a branding exercise, it was a political exercise using dirty-tricks tactics from an economically advantaged position in order to try to suppress a system of political beliefs that was growing (and that continues to grow) in popularity.

-t

[John Newton]

Gee Thomas - I didn't mean to make you so angry. I can't be lying, since I wasn't there and as I admitted, I thought all this stuff was communism back then.

I'd rather not call people liars since it is implying a state of mind that we can't possibly know (although it seemed to work wonders for Al Franken ;) ). However, if there were any distortions or misunderstandings of the truth, please point me to the right sources. I will in turn dig up more from RMS's writings and try to contact him. I'm really more interested in accuracy.

Thanks for your comments,
John

[Thomas Lord]

The story as you tell it doesn't add up. Either you believed the lies of others or you lie yourself.

Beginning with Eric Raymond's letter that you link to but going far, far beyond that the society of influencers around that meeting aggressively attacked RMS, attempting to marginalize him, the Free Software Foundation, and the GNU project. That was no exercise in "brand creation" that was a coordinated political action.

Were it the case that they merely sought to make a business-friendly brand for free software, there would have been no reason to attack Stallman.

For what political reason would it be important to some clique to marginalize Stallman? Perhaps the answer is plain by looking at the "Open Source Initiative" which defines its mission around saving dollar costs of development and ownership of software vs the "Free Software Foundation" which defines itself around the freedom of software users.

The folks in that room were representatives of big-C Capital brainstorming on an aspect of how to put down a freedom movement arising within labor but threatening to win over consumers to its cause.

-t

[James Dixon]

Nice summary John.

[Harold]

John,

Nice article, its nice to know this information

Cheers,
Harold