Editor's Column

1 February 2001

Open Source and Its Culture

About this column

I’d like to clarify the purpose of the Editor’s Column. The articles that comprise "Editor’s Column" do not seek to represent Sun’s view, and I don’t know about OpenOffice.org’s views, as it’s not clear that one can easily synthesize the disparate views of an open-source community as such into something coherent enough to fit into a column. Rather, the purpose of the column is to focus on issues that the community has found interesting, as evidenced by discussions in the mailing lists, or might find interesting, because they relate to Open Source, its communities, and important spokespeople.

All this is to say that the articles I write under the rubric, "Editor’s Column" are meant to be discussed. So, if you disagree with the ideas, logic, or characterizations, please, feel free express your opinions. And, if you feel that I am not addressing some pressing issue, don’t hesitate to let me know. As I’ve stated before, I would be delighted to receive suggestions about topics, interviews, what have you. In fact, this is a serious call for ideas: Send in your suggestions either to the discuss list or to louis at collab.net. It would also be great if a community member were to submit for consideration an article, interview, or piece that he or she felt others in the community would find interesting.

Questions about Open Source

This article inaugurates a series of articles questioning the shape and force of Open Source, as a culture of work and as an organizational system. I plan on these articles being intermittent; the sequence will doubtless be interrupted by more immediate news or controversy related to OpenOffice.org.

There are several questions that motivate this inquiry; they include:

How are (other) large open-source projects organized?
What are the issues at stake? (I.e., what is hoped to be accomplished by open-sourcing the code?)
Under what license or licenses are these constituted?
What is the relation between the sponsoring corporation and the open-source project?
Who are the developers, and what do they have to gain from the project?
And, finally–and most intangibly–what is the "culture" of the project?

Clearly, this project is large in scope and will take some time to fully complete; in fact, I’m sure I’ll have to revisit the Darwin project later on. My plan so far is to examine not just Darwin, but also Mozilla.org, and other large, open-source projects. Why large projects? Size is a criterion here simply because I suspect a large open-source project differs importantly from smaller projects, and not just in the logistics of arranging the release of enormous blocks of code to an uncertainly defined community. The nature of the community, its "culture," how it works together, and the manner in which it contributes code also changes according to the size and scope of the project.

This project is evolving. If the community wants to contribute suggestions–such as different questions to ask, or particular projects that should be examined–please forward them to me, and I’ll see what I can do.

Darwin

Apple Computer’s open-source Darwin was announced on 16 March 1999, and, in the words of its statement to the press, Apple became "the first mainstream operating system provider to release its source code to the public and base its system software strategy on Open Source technologies." Initially, the focus was on Mac OS X Server; it quickly included the consumer-oriented Mac OS X, which is slated for release on 24 March 2001.

Darwin 1.2, the current release, is enormous. 135MB for a disk image of the binaries. Interested developers can also download source code; the project allows modifications to be submitted through mailing lists and CVS (Concurrent Versions System), which is recommended but not required.

Despite its size and ambition, however, Darwin as an open-source project seems to be often ignored by other open-source groups. OpenOffice.org, for instance, is usually compared to Mozilla.org, and is touted as being the largest open-source project. Darwin is never (or very seldom) mentioned–not by Sun press, nor by the media at large, when discussing OpenOffice.org

This lack of mention probably has something to do with the common notion that Apple exists in a universe of its own. But it also has to do, I think, with the controversy surrounding Apple’s open source credentials. Shortly after the announcement of the project’s inception, however, the license under which Darwin project operated and which defined the project’s status as "open source," became the focus of a debate between Open Source luminaries Eric Raymond, of OpenSource.org, Bruce Perens,who gave us "Open Source," and Richard Stallman, who can be said to have started Open Source (or, as he prefers, "Free Software").

To summarize the history, Eric Raymond gave, if not his blessing, his certification, to the Apple Public Source License under which Darwin was constituted (APSL Ver. 1 and shortly later, 1.1; it is now at Ver. 1.2). Raymond was publicly criticized for his actions by his colleague Bruce Perens, who along with Wichert Akkerman, Debian Project Leader, and Ian Jackson, President, Software in the Public Interest, argued that the APSL Ver. 1 failed to meet the necessary criteria for Open Source (Version 1.1, issued 19 April 1999, mostly satisfied the authors). Richard Stallman was unpersuaded by the changes Apple made to its license and condemned it as roundly as he had the first version (Ver. 1.1 fell "short of being a free software license"). The debate got a little savage, but that is because the issue at stake was so important to Open Source.

The issue had to do with the relation between corporate interests and Open Source. As Perens rather despondently concludes in his rebuttal to Raymond, "The needs of corporations are not necessarily those of the free software community, and it may even be the case that the twain will never meet. Open Source appears to be splitting into something I'd call "Corporate Source", semi-free programs with disclosed source but less than the full set of rights we are used to, and true Free Software as represented by the GPL, LGPL, X/BSD, and other licenses. Public discussion of this fact is essential. We may eventually have to accept that it will never be possible for corporate participation in the free software community to be as full as we would like. Contributions like the MacOS X source may end up being useless to the free software community as far as code reuse is concerned, but they may still be good documentation on the underlying hardware, and will be useful, with some caution, to authors of fully free software."

Perens wrote gloomier than warranted. Not only has OpenOffice.org evidenced that a company can move large, proprietary, software using true Open Source licenses, but Apple itself has recently (4 January 2001) revised its license for Darwin. The short blurb under the link for the new license states, "The Apple Public Source License has been updated to make it easier for people to contribute to and use the software," and after the most cursory glance, it seems to be so. Rather belatedly, Apple has entered Open Source (even Slashdot seems to agree).

What is left hanging, however, is the question of the culture of Darwin. Future articles will examine this aspect of the project. I will also begin a look at Mozilla.org.

Previous columns

23 January 2001 Community Action

16 January 2001 Quo Vadis OpenOffice.org?

9 January 2001 The 613 build: problems and opportunities

3 January 2001 Sun's open door

E-mail: Louis at collab.net