In open source software communities, few events are as exciting as the release of a new operating system. Community members may wait for months—even years—as fresh versions of their favorite Linux distributions are collectively and meticulously prepared, debugged, and packaged for the world.
Next month, Todd Robinson will release a Linux-based desktop operating system in a single day. Thirty-one times in a row.
Robinson, co-founder of Webpath Technologies and systems development engineer at On-Disk.com, is about to launch a project that he hopes will powerfully demonstrate the value and versatility of open source software. Dubbed "31 Flavors of Fun"—a nod to the retired Baskin Robbins slogan—the effort will involve assembling a different operating system every day during the month of August.
Robinson is no stranger to the art of customizing Linux distributions. For On-Disk, he produces, remixes, and personalizes Linux variants on numerous media—products he sells to home users, educators, and IT professionals alike. (Robinson assures us that operations at On-Disk will continue uninterrupted during the "31 Flavors" experiment.)
Expect the first release on Wednesday. And when the project concludes, Robinson will report on his experience in a presentation entitled "The Road to 31 Flavors" at this year's Ohio LinuxFest.
As Robinson prepares to launch his ambitious initiative, he graciously answered some of our questions about his goals, his concerns, and his love of open source.
How did you develop the idea for this project?
Initially the idea came in response to comments from many On-Disk.com customers and to negative articles floating around concerning Linux not having a "standard" desktop since Gnome 2. For me, it's the flip side of the coin that has always made things interesting: the flexibility of almost endless combinations of applications and desktop setups that can be used to accomplish incredible things. Over the years I've had many different desktop setups designed around the specific tasks I needed to perform. In fact, at the moment, it would be impossible for me to perform my daily functions using any existing proprietary solutions. Even the desktop I'll use to create the distributions in August will be different from what I'm using now. When the idea first came to me it was, "I should create a bunch of desktop releases to show off the variety of Linux desktop options," quickly followed by, "but that sounds like way too much work," and finally, "but I can't think of a better way to actually demonstrate the variety in Linux and advantages of open source development at the same time." And that last one got me hooked.
You see, at all the Linux and open source events we (my wife and I) attend, people boast about how superior open source development is, and even give personal reasons for their opinions. On the other side of the fence there are those who feel differently, and who also argue to present their case. But both sides, logical in their own reasoning, lack any sort of "real world" experiment as proof of concept.
I was going to try 30 days of releases, but August has 31 days and the Baskin Robins slogan, "31 Flavors of Fun", popped into my mind. I decided to go with it. Besides, I think I'll have an easier time of it if I remember that it's supposed to be fun.
What benefits of open source software and the open source development model is the project designed to highlight and demonstrate?
When all is said and done, and I run the numbers, I suspect it'll show developing under open source costs less and requires less manpower. But a huge benefit, which I think is often overlooked, is the much higher success rate. Lets face it: a lot of development attempts outright fail. By freely sharing our knowledge we're not dependent upon just what we ourselves, or a small development team, can come up with. Through online forums, blogs, and source listings everywhere, anything we need to know is literally at our fingertips. Open source is more about shared knowledge than just the source code itself. But it's precisely because the code is open and available to all that we can ask others about it, get opinions on the best way(s) to proceed, and quite often get free assistance along the way. Yes, I am making these 31 releases myself, but just like any other developer using open source, I do so standing on the giant shoulders of the open source community. If nothing else, I'd like to do my part to help stomp out the myth that open source solutions are unreliable. What better sense of security can a company have than knowing your techs and admins have the resources and expertise of the entire open source community to call upon should problems arise?
On a personal level, I'd like to help draw in as many developers as I can to participate in open source development. Only through openly sharing our knowledge and expertise does mankind have any chance of solving the problems we have created.
Producing an entire operating system in a single day will certainly be a challenge. What hurdles to success do you anticipate?
Time. Time is the biggest hurdle. An average ISO build can take anywhere from a 1/2 hour to three hours depending on the approach. It's when something in the build fails and it needs to be debugged and re-built that will seriously eat up the clock. Just like any system administrator or software developer out there with a deadline approaching, I simply won't have the luxury of time to redo something so I need to make sure every build has the best possible chance of success. And obviously there is only so much testing I can get done before needing to commit to the build, so I'm sure some releases are going to be a lot more functional than others.
The most challenging of the proposed releases will be for the Raspberry Pi. Chances are good that the release for it may not come until after August. It's going to take a lot of trial and error to come up with something truly usable while only having a maximum of 233MB of RAM to work with. My plan is to start working on it throughout August, and if it gets done during the month and can be included as one of the 31 flavors then that's great. If not, that'll be OK as well. Even if it's released during August, I'll most likely do some more work on it later because it's a custom release I committed to some time ago that we're planning to maintain.
What do you hope will happen to your operating systems after their initial releases?
I don't really have any hope for them one way or the other. It would be great if they can remain available on the mirrors for a few years to allow people to experience some of the variety that Gnu/Linux has to offer. I'm sure that at least a couple will be maintained and made available through On-Disk.com, like I do with Unite!, and through occasional limited releases.
How can others get involved in the project?
The biggest way is just to spread the word and tune in as we move forward. The more people watching as I do the impossible using open source solutions, the louder the message will be. And of course try out some of the Releases as they become available. Have fun with it. Think of it as a month long Linux/open source fest. I'm still taking requests, and the feedback has been very helpful. So, please keep them coming.
- A Twitter feed
- A daily activity log at my 31 Flavors blog
- A "31 Flavors of Fun" forum
- An "Anticipated Releases" page, which will turn into the release announcement page as each release becomes available.
- Download mirror(s)
Sponsors are needed so that publicly available downloads of the 31 releases can be offered. As much as 60GB of server space may be required, so please contact me if you are able to help. Your sponsorship will be public and mentioned at all talks and publications in regards to this project.