Ian Murdock In His Own Words: What Made Debian Such A Community Project

from the looking-into-the-details dept

As you may have heard, there was some tragic news a few weeks back, when the founder of Debian Linux, Ian Murdock, passed away under somewhat suspicious circumstances. Without more details, we didn't have much to report on concerning his passing, but Gabriella Coleman put together this wonderful look at how Murdock shaped the Debian community, and why it became such a strong and lasting group and product.

Ian Murdock in his Own Words: "The package system was not designed to manage software. It was designed to facilitate collaboration" Ian Murdock (1973-2015).

Peering in from the outside, the Debian operating system -- founded in 1993 by Ian Murdock, then a twenty-two-year-old college student -- might appear to have been created with hardcore, technologically-capable power users in mind. After all, it is one of the most respected distributions of Linux: as of this writing, the current Debian stable distribution, Jessie, has 56,865 individual open source projects packaged (in native Debian parlance software is referred to as packages), and Debian itself has functioned as the basis for over 350 derivative distributions. Debian developers are so dedicated to the pursuit of technical excellence that the project is simultaneously revered and criticized for its infrequent release cycle -- the project only releases a new version roughly every two years or so, when its Release Team deems it fit for public use. As its developers are fond of saying, "it will be released when it's ready."

But if you take a closer look, what is even more striking about Debian is that its vibrant community of developers are as committed to an array of ethical and legal principles as they are to technical excellence. These principles are enshrined in a bevy of documents -- a manifesto, a constitution, a social contract, and a set of legal principles -- which guide what can (and cannot) be done in the project. Its Social Contract, for instance, stipulates a set of crystal clear promises to the broader free software public, including a commitment to their users and transparency.

In 2001, I began anthropological fieldwork on free software in pursuit of my Ph.D. Debian's institutional model of software development and rich ethical density attracted me to it immediately. The ethical life of Debian is not only inscribed in its discursive charters, but manifests also in the lively spirit of deliberation and debate found in its mailing lists. Ian Murdock, who passed away tragically last week, had already left the endeavor when my research began, but his influence was clear. He had carefully nursed the project from inception to maturity during its first three years. As my research wrapped up in 2004, I was fortunate enough to meet Ian at that year's Debconf. Held annually, that year's conference was hosted in Porto Alegre, Brazil, and it was the first year he had ever attended. Given his fortuitous presence, I took the opportunity to organize a roundtable. Alongside a couple of long-time Debian developers, Ian reflected on the project's early history and significance.  

By this time, many developers had already spoken to me in great (and fond) detail about Ian's early contributions to Debian: they were essential, many insisted, in creating the fertile soil that allowed the project to grow its deepest roots and sprout into the stalwart community that it is today. In the fast-paced world of the Internet, where a corporate giant like AOL can spectacularly rise and fall in a decade, Debian is strikingly unique for its staying power: it has thrived for a remarkable twenty-three years (and though I am not fond of predictions, I expect it will be around throughout the next twenty as well).

It was well-known that Ian established the project's moral compass, and also provided an early vision and guidance that underwrote many of the processes responsible for Debian's longevity. But witnessing Ian, and other early contributors, such as Bdale Garbee, articulate and reflect on that early period was a lot more potent and powerful than hearing it second hand. In honor of his life and legacy, I am publishing the interview here (it has been slightly edited for readability). Below, I want to make two points about Ian's contributions and do so by highlighting a selection of his most insightful remarks drawn from the roundtable discussion and his blog -- comments that demonstrate how he helped sculpt Debian into the dynamic project it is today.

1. Ian Murdock instilled a culture of reciprocity in Debian

Technologists and hackers contribute to free and open source software (F/OSS) development for a variety of reasons, many of which have little to do with altruism. But when Ian first cut his teeth on free software he was moved by the fact that other developers gave their code away freely.

What really grabbed me off the bat was the community. That was what really grabbed me and you have to understand at the time, it was a completely foreign notion that I had somehow stumbled upon this group of people that were interested in the same things that I was interested in, [and] who had basically for no particular reason built this thing, this operating system, and it had actually worked, and I could do my work on it, and I had not paid a dime for it; they did not ask anything of me when I download it or used it. And whenever you are in a situation like that, when people have given so much to you, one of the first instincts is like: "What can I do for you, what can I give back?"

Compelled to return the gift, he created Debian and fostered within its community a generalized culture of sharing and responsibility -- a culture even more remarkable when one considers the fierce commitments to individualism held by many of the technologists involved. When the news of Ian's passing circulated online, Branden Robinson, a long time Debian developer who had worked closely with Ian, paid tribute to his friend by praising Debian on these very terms:

Debian is about individual empowerment in a cooperative social environment -- about taking responsibility while sharing responsibility. If that sounds like a tension or a dialectic, you're on the right track. One can learn a great deal from working out such demanding principles; I have, and continue to do so.

So how did Ian establish a project whose participants, already so busy with technical labor, were willing to expend even more time to work through such "demanding principles"? Obviously, his leadership style played a vital role. But equally important was his decision to articulate a vision for Debian and publish it as a manifesto. Released shortly after the founding of Debian in early January 1994, Murdock stated in no uncertain terms that Debian would be a "non-commercial" distribution, developed by the community of developers for the community of developers. What is especially notable about this document is its prescient ending:

The time has come to concentrate on the future of Linux rather than on the destructive goal of enriching oneself at the expense of the entire Linux community and its future. The development and distribution of Debian may not be the answer to the problems that I have outlined in the Manifesto, but I hope that it will at least attract enough attention to these problems to allow them to be solved.

Let's pause for a moment to contemplate the forward thinking nature of his statement. This was 1994: a good six years before Silicon Valley entrepreneurs began to swoop in and avidly embrace free software, rebranding it as open source in the doing. While the F/OSS development community largely benefited from this attention both with increased monetary resources and media visibility, projects ranging from Ubuntu to Tails undoubtedly benefited in equal step from Ian Murdock's perceptive decision to charter Debian as an autonomous community of developers and a commons.

There was a visionary substance behind his manifesto. There is perhaps no greater testament to this substance than the Debian developer community's continued commitment to his early vision year after year. In 1997 they hammered out a set of legal principles for admissible licenses -- what came to be known as the Debian Free Software Guidelines -- and made a series of promises to the broader F/OSS community in their Social Contract. The following year they articulated the project's goals and governance structure in their Constitution. During the Debconf roundtable, Bdale Garbee emphasized how important it is to Debian that the project's commitments be voiced explicitly -- a tradition that began with Ian's Manifesto.

So the creation of the social contract and the DFSG [Debian Free Software Guidelines]: that was one of the points where the group had grown enough that it was important to articulate what the group's fundamental values were, because in order to build a community you have to have some way of articulating and connecting some shared values..."

2. Quality work requires a concrete and often institutional framework for collaboration.

One of the most frustrating misconceptions I routinely encounter regarding F/OSS development is that it occurs in some ad-hoc, laissez faire fashion. The wrong-headed perception goes something like this: a bunch of developers have a good idea about a piece of software they want to write, so the coders get together on Internet Relay Chat and mailing lists and start madly coding away, hosting their code on some super awesome versioning software like Git, and then bam!, like magic, they release a piece of functioning software a few months later. Sure, a loosey goosey style of development sometimes works. With smaller projects it is probably even the norm; most run -- and run well -- on loose consensus and the energy of excitement alone. Nevertheless, any online project that scales and grows -- or at least any project that wants to create quality material as it scales and grows -- usually has to formally organize itself, or else face the threat of extinction (or, at the very least, face the threat that the developers will have to live with themselves after releasing shitty software).

Like a wise master craftsman who knows norms and policies matter to the crafting of excellence as much as individual skill and capacity, Ian Murdock decided early on to implement frameworks for quality collaboration. One of his most famous contributions to F/OSS development techniques is Debian's packaging system. Prior to this system, it used to be a long and complicated process to install any software. You would need to find the source code, figure out how to compile it, resolve all the dependencies it had by hand, and then install it. Every piece of source code did it differently, and once it was installed, you couldn't easily remove it. It could take hours to get something installed. The package management system changed that completely, everything was bundled together enabling you to simply type one command, or click a button, to have it installed or removed. As described on Ian's blog in the post titled, "What's the single biggest advancement Linux has brought to the industry?"

It's an interesting question, and one that in my opinion has a very simple answer: Package management -- or, more specifically, the ability to install and upgrade software over the network in a seamlessly integrated fashion -- along with the distributed development model package management enabled.

And yet, crucially, in the words of Ian himself: "The package system was not designed to manage software. It was designed to facilitate collaboration." Eventually, this managerial approach was expanded to encompass more than just the packaging system, enshrined into the policy system still alive in Debian today:

It was not until late 94 when we had a working package system. We started to realize that just having packages was not enough. We had to have policies. Here is the mechanism, what is the policy?  If you want something, if you want a service to start at boot time, what do you do? And at that point we had a bunch of really horrible things, like, you know, we had post installation scripts and so people were editing/etcrc and you could imagine what happens after awhile....

Over the years, developers followed in his footsteps: they continued to implement techniques and protocols -- at times through technical means and in other instances by devising distinctly social mechanisms -- to ensure the persistence of quality even as Debian ballooned in size (Debian is today one of the largest free software projects, with nearly 1000 official developers and thousands of other contributors). It would be no exaggeration to describe some of the technical engineers within the Debian project as shrewd social and political architects. Ample justification for such an appraisal can be found in Debian's impressive "New Maintainer Process" -- an entrance exam designed to test new developers' knowledge of technical, legal, and ethical matters prior to their enrollment in the project. In the panel, Bdale emphasizes the manifesto's importance in attracting a cadre of geeks and in connection to the subsequent documents and procedures.

[The Manifesto] acted like a geek magnet. It was the attractor that caused people to come here and say: "yes I want to participate in it." … So the manifesto acted as an initial attractor, and when the group got big enough it was important to make sure that they understood what the expectations were, and the Social Contract, and the keysigning process, and the NM [New Maintainer] process: We sort of require that you have said that you have read this, and you agree and abide by it. And there are of course people who have been around longer than those documents, and most of them would not have stayed if they did not agree with it.

I would argue that with these social and ethical protocols Debian transcends a mere technical project that produces an excellent product, and instead also qualifies as a sort of miniature society, and given its social contract and constitution, it also has a very 19th century, Enlightenment feel to it.

Ian Murdock left the project after a rather short period of just three years. But this was by no means an abandonment. The fact that he could step away represents just how important he was to the early project: He built an institution, an organization, that could function without its founder. He said as much himself during the roundtable:

Yea and I was in school and basically at that point we had a thriving community. We had infrastructure, we had lots of users. And my final test as to whether or not Debian succeeded was: could the founder step away from the project and could the project keep going because that is the only point at which you know that the project has basically taken a life of its own.

This league of technologists and programmers who proudly call themselves Debian developers are mourning the loss of one of their own. Ian's life was cut way, way too short. But he will be fondly remembered for building a project that teems with life. Which, to use his own words, "took a life of its own" thanks to what he did as a bright-eyed, clear-sighted twenty-two-year-old. He will be dearly missed.

Reader Comments

Subscribe: RSS

View by: Time | Thread

  • identicon
    Anonymous Coward, 14 Jan 2016 @ 3:16pm

    Ian was the student of my students...

    ...and I was delighted to see that he excelled, in many ways surpassing their achievements and mine. We call this "progress", and he created a great deal of it. I am sorry he left so soon; but I hope that those who are his students, and the students of his students, will carry on the work -- always mindful of the debt of gratitude they owe him.

    reply to this | link to this | view in chronology ]

  • identicon
    Lawrence D’Oliveiro, 14 Jan 2016 @ 3:28pm

    The Stroke Of Genius That is Package Management

    Let me amplify on some aspects of the package-management issue:

    • Package management is scalable and maintainable. As mentioned, the standard Debian distribution includes over 50,000 packages. And you can actually install most of them and 1) still have a working, stable system while 2) being able to keep everything automatically up-to-date. Imagine trying to do the same with, say standard vendor-provided installers on your choice of favourite proprietary platform--it would be a nightmare. And proprietary installers have no standardized update mechanism--either it is up to you to go back to the websites and check for updates, or the installers will insinuate their own proprietary updaters into your system, bogging it down with extra background processes.

    • All the tools for building and maintaining packages are part of Debian itself. I still sometimes want to build my own versions of packages that Debian may provide (e.g. Blender). This may be because the upstream project is much faster moving than the Debian package maintainers can keep up with. This means I have to make sure I have all the necessary dependencies installed on my system. The command
    apt-get build-dep package

    will ensure that I have all the necessary dependencies to build the standard Debian package package from source. And that also gets most of the dependencies I need for my custom non-package version--saving a lot of work!

    Consider when the Raspberry π first came out. There was a version of Debian available for ARM CPUs, but the particular combination of ARM architecture features that the π used was not well supported. So a couple of volunteers decided to take on the task of rebuilding the whole of Debian from source, optimized for the π.

    They got most of the job done in six weeks.

    By contrast, look at how much effort, for so little result, Microsoft has expended coming up with its cut-down, feature-limited yet still resource-hungry “Windows 10 IoT Edition” for the Raspberry π 2.

    All in all, you can see why the cutting edge of software development these days is based on Open Source, and why Linux is the premiere Open Source platform. The network effects from being able to do open, collaborative development are so massive, they leave proprietary projects in the dust.

    reply to this | link to this | view in chronology ]

  • identicon
    Shmerl, 14 Jan 2016 @ 6:03pm


    I'm very grateful to Ian for creating this great project. I'm using Debian on the constant basis (writing this from Debian with KDE Plasma 5).

    reply to this | link to this | view in chronology ]

  • icon
    Ninja (profile), 15 Jan 2016 @ 4:24am

    May he rest in peace. I am particularly fond of Debian and its derivatives and incidentally it was the very first Linux distro I got in touch many years ago. I wish the community can carry on what he developed.

    reply to this | link to this | view in chronology ]

  • icon
    Mason Wheeler (profile), 15 Jan 2016 @ 7:34am

    Let's pause for a moment to contemplate the forward thinking nature of his statement. This was 1994: a good six years before Silicon Valley entrepreneurs began to swoop in and avidly embrace free software, rebranding it as open source in the doing.

    Can we please back off on the rewriting of history? Those of us who lived through that period know better.

    The Open Source Initiative was founded in 1998 by several developers, including Eric S. Raymond, who wrote the seminal paper The Cathedral and the Bazaar in 1997. The term "open source" was already in wide enough use at that point that no one was confused by his use of it in the paper.

    The basic open-source concept of code being shared around freely for people to use as they wish, including incorporating it into proprietary software, is older still; it's been the natural state of software development ever since programmable computers first emerged from being a government-only project to something mere mortals could work on. It only had to be formally re-branded as "open source software" and presented by the OSI because of all of the damage that the FSF's extremism was doing to the image of non-proprietary software.

    Today, open-source software has found wide acceptance and is used everywhere, whereas Linux, locked away behind restrictive FSF licensing and a user-hostile community who are stuck so far behind the times that they still think a command line is a user interface, languished in obscurity for decades, never finding acceptance outside of the server world, until Google built a proper user-friendly OS out of it, (going well out of their way to avoid GPL entanglements as they did so,) and the rest was history; today it's well on the way to taking over the world.

    reply to this | link to this | view in chronology ]

    • identicon
      Lawrence D’Oliveiro, 15 Jan 2016 @ 12:21pm

      Re: The basic open-source concept of code being shared around freely ... is older still

      Maybe, but only to a point. Specifically, to the point that the author says “Copyright? What’s that?”.

      The modern concept of “open source” or “Free software”, on the other hand, explicitly recognizes that copyright still applies (if only because in many jurisdictions, you cannot simply put your creations into the public domain). While we freely pass stuff around, we are scrupulous about licensing, and in particular about adopting and promoting licences that respect the freedoms of the users.

      Yes, I lived through the period too.

      reply to this | link to this | view in chronology ]

  • identicon
    Matthew A. Sawtell, 17 Jan 2016 @ 10:15am

    The key to this... Faith

    When it comes down to it, what Ian was able to do is a testament to the faith that people had in him at the beginning, and then in the system he helped to create after he stepped down as the lead. As time moves on , as the system marks a lifespan longer than its creator,then the final testament to that faith can be truly be bestowed upon Ian, and the people that rallied around him.

    For those of certain age, and have worked in certain careers, we live in interesting times - where we will have the ability to observe if certain technologies/concepts/beliefs will or will not outlive their creators.

    reply to this | link to this | view in chronology ]

  • identicon
    Anonymous Coward, 24 Jan 2016 @ 5:14pm

    I'm grateful to Ian, because Debian is my favorite GNU/Linux distribution. Not many distributions separate free software repositories from non-free software repositories.

    Plus I simply love Debian Project's overall philosophy. I really relate to it's non-corporate, free software ideology.

    Not to mention there's so many software packages to choose from!

    reply to this | link to this | view in chronology ]

Add Your Comment

Have a Techdirt Account? Sign in now. Want one? Register here
Get Techdirt’s Daily Email
Use markdown for basic formatting. HTML is no longer supported.
  Save me a cookie
Follow Techdirt
Insider Shop - Show Your Support!

Report this ad  |  Hide Techdirt ads
Essential Reading
Techdirt Deals
Report this ad  |  Hide Techdirt ads
Techdirt Insider Chat
Report this ad  |  Hide Techdirt ads
Recent Stories
Report this ad  |  Hide Techdirt ads


Email This

This feature is only available to registered users. Register or sign in to use it.