Putting the Linux Desktop Closer to the User

by Tom Adelstein
October 18, 2004

The use of LiveCD's has helped many curious people discover Linux. One can simply place a LiveCD into a CD Rom drive, boot up their computers and watch the flexible Linux kernel run a complete operating system within memory. But, while this makes Linux available without having to install the distribution on a hard drive, it doesn't necessarily help a new user navigate around the system and/or understand Linux any better. New users could use a blueprint to go along with the LiveCD. Our product team has done that and in this article we tell you how and why.

The vast majority of Linux users once used Windows as their primary platform. In fact, I still have several old MCP certificates in a folder somewhere. And though I learned to use Linux quickly, it's easy to forget how difficult and challenging the changeover seemed in the beginning. Today, Linux is much easier to use. I'm so comfortable with it, I find it somewhat strange that people just don't change to it in droves.

When we wrote Exploring the JDS Linux Desktop my co-author and I wanted to create a portable Linux computing facility that lets a new user learn to navigate Linux in a easy way. Our vision included writing a book unique to any that had come before it and to provide a complete live operating system on a CD Rom. In other words, we wanted to create something you could take home, boot up and on which you could learn Linux, and/or Star Office, Open Office or any number of applications. We wanted it so portable one could use it without having to install it on his or her hard drive. If we could avoid the risk of disturbing an existing system that would give us an advantage over anything Microsoft ever offered.

We wanted to create a product that would compel people to give Linux a try in large numbers. If we put the book in the right hands, we knew we could attract people who had not stepped over the line and given Linux a try before.

For the average computer user, changing over to Linux presents a challenge. In a world where 93% of computers come with Windows built in, Linux might need some professional tweaking. Then again, it might install flawlessly. If a LiveCD works, then you know Linux will work on a machine.

Choosing a Linux Distribution

In December 2003, a colleague and I bought distributions of JDS 2003. In our separate computer labs we independently test Linux distributions. After numerous usability tests, we settled on JDS as a distribution we wanted to recommend for desktop clients. At LinuxWorld in New York, we received copies of the JDS LiveCD that used the Morphix Project. At that point we got the idea to create curriculum that would include a user manual with Sun's JDS LiveCD.

As a curriculum offering, we regarded the availability of a LiveCD as important as the distribution itself. Morphix allows anyone to build their own distribution using their kernel and components from the Debian project. At first, we thought of creating courses primarily for Openoffice.org's productivity suite. But, Sun's LiveCD gave us more options.

We wanted to introduce Linux, OpenOffice.org/Star Office, Mozilla, Evolution, Gaim and GIMP to users of Microsoft Windows. We felt if such users experienced these products, they would at least use the Windows versions and gain an appreciation of the added security and stability of open-source software. We also wanted to help open-source advocates by providing them with an inexpensive, tangible tool they could give to their friends and family.

We had some other considerations worth mentioning. Sun provided people who helped us get our jobs done. Other distributions provided little cooperation. We liked the JDS product and thought it provided the best face to users. We also felt the Sun brand gave Linux its most credible vendor with regard to new users.

Some people believe that Sun is tied to its own Solaris UNIX operating system and doesn't have a commitment to Linux. Sun's Java Desktop System runs on both Solaris and Linux. Sun has demonstrated its commitment by using Linux as its own corporate desktop and by continuing to develop JDS on the Linux platform. From our standpoint, JDS served our needs and at the end of the day it's still Linux. Coincidentally, JDS happens to have some superior engineering and that makes it a serious competitor to Microsoft.

Writing a Different Linux Book

The Linux community has many books from which to choose. In our analysis of the market, we did not find books that focused on the desktop. We saw the need for a Linux book that simply dealt with a user's core desktop requirements. Additionally, we began building a companion web site to provide support if people wanted to go deeper into the Linux experience.

My colleague, Sam Hiser specialized in Open Office and Star Office having worked on that project for two years. I specialized in Linux distributions, administration, usability and the core operating system. JDS included Star Office 7, Mozilla, Evolution and Instant Messaging. These were tools to which Sam related. Among other distributions, I had worked on the SuSE Linux Enterprise Server for two years on Intel, Power PC and Mainframe Platforms. We split up the material to meet our specialties.

To write a different Linux book, we needed a publisher that knew the Linux and open source market. A publisher whose very presence added credibility to the project. We also needed someone who could embrace our vision of reaching above the technically savy market line. We wanted to remain professional yet reach the market to which "Dummies" books appealed. But, neither of us wanted to become "Dummies" authors.

We needed a novice who could represent our audience and would say, "I don't understand what you just wrote. Or, what you just wrote didn't help me". We discovered such a person living in my house. My wife, Yvonne, helped us by taking the point of view of a nontechnical user who would interact with JDS.

She went from a computer newbie to a capable Linux user over the course of the project. In fact, she went from someone who had never seen Linux and could only use a browser, some word processing and email to a hands-on editor. She helped us better address the needs of readers coming fresh to Linux and kept us from straying into technical jargon. She was our proof of concept. Watching her progress with each lesson gave us confidence in our work.

Aside from reaching the right publisher, we knew we needed to work with an editor who could understand our vision. As it turned out, we found that editor in Andy Oram who not only embraced the project but worked with us like a hands on producer/director.

When in Doubt, Go to the Source

When we began writing our book, I also started monitoring the official Sun Java Desktop System Forum to find out what kind of issues people had with JDS and address them. Within a day or so, I noticed a shortage of Linux veterans. People were asking questions on the Forum but their questions would go unanswered. I realized other people were having the same problems because I would notice hundreds of views on those questions with no replies. So, I began to reply.

I soon discovered that lots of users indeed took Sun up on the proposition to switch from Windows to JDS. Many of those users came to JDS "cold turkey" and needed some hand holding. So I began to hold some hands.

Within weeks, more Linux veterans began to show up on the Forum and soon we had about a dozen active proxy moderators. We quickly learned of each other and began working in concert. We began discussing among ourselves existing support tools that a JDS Community needed. Some of the those tools included a database driven composite of Frequently Asked Questions, a Request Tract or trouble ticket tracking system, a repository of RPMS and patches similar to those maintained by communities such as the Fedora legacy project. We also needed to consolidate the resources from Sun into a more navigable format. We started building a web site but got sidetracked by people's time constraints.

That changed when JDS Release 2 came out. Existing users could upgrade for free, but they had to download a minimum of three 500 megabyte ISO images. JDS users discovered that they could burn CD's with JDS tools, but they didn't have a GUI to create a CD from ISO images.

The JDS Forum users seemed to cry out for help in getting their upgrades downloaded and burned to recordable CD Rom Disks. I decided to share a RPM of xcdroast I had built for myself. I offered it to the xcdroast project and they added it to their list of binaries.

I posted how people could install xcdroast and use it to burn their updated CD's for Release 2. That led to a discussion among the JDS veterans to put together a web site, consolidate the RPMS, scripts, and drivers people had collected. We also started collecting frequently asked questions (FAQs) and writing Howtos.

The companion site Sam and I envisioned at the start of our book project became JDSHelp.org. It's become the community site with a slant toward home users. The site contains a repository of popular software not available on the JDS distribution. We create our RPMS on running JDS systems. We also maintain a knowledge base, a library of Howtos, a news site and entertainment software including games from GCCLinux.com Participants in the community site come from all corners of the globe such as Norway, Argentina, Indonesia, Germany, Canada and numerous points in the US.

Members of the JDS community have embraced our project. In fact, when we needed technical reviewers for the book, we found them through the JDS Forum. Since the reviewers needed to have familiarity with JDS we knew of no where else to go. In many ways "Exploring the JDS Linux Desktop" has come from community involvement.

A recent post from the JDS Forum illustrates how much the community has contributed to our project:

I'm browsing your O' Reilly book on JDS (received yesterday from Amazon.com) as I write this. Great work!...



I was amazed to find in page 368, in the "problems and solutions" section a reference to "Sony VAIO laptops using an ATI Radeon IGP-345M". That's surely a result of my initial feedback on Sun JDS 2003, which I emailed to several Sun people! A company that listens to the customer... that's amazing. :)

I found it interesting that the information used on page 368 mentioned above came from the Sun JDS Forum. Several users had similar issues and we helped fix and document those while providing a solution.

So, we see the the fulfillment of our vision coming from the community. Not just the JDS Community, but the Linux community. Whether this occurs because we wrote a book or because someone embraces the vision becomes irrelevant in the fullness of time. Eventually, as a community we decide if people get the opportunity to experience the joy of Linux.

Some Final Thoughts

Nine months after we committed to create a portable Linux computing facility that let a new user learn to navigate Linux easily, our project arrived at books stores around the world. We have an active community and expanded the content of the book to include an on-line knowledge base, a long list of JDS Howtos and a significant repository of freely downloaded software. I don't believe any of the people who participated in writing our book envisioned the scope of what we have finally accomplished. But, now most of us believe we have only scratched the surface of what's possible.

On September 10, 2004, James Grisanzio blogged an interesting comment about something I wrote in a Linux Journal article. Jim wrote:

[Tom]Adelstein articulates the need for good community relations and that the big vendors(Red Hat, Novell SuSE, IBM, HP, Sun) sometimes neglect that important task. From his conclusion:

As Linux distributors march to the beat of their enterprise wins, they will continue to discover that the community is the source of their real success. As such, they owe the community better relationship managers than those they have furnished so far.



Government has turned more and more to off-the-shelf components for projects such as the Mars Rovers. It might be the efforts of a few Linux advocates who build a DVD encoder that make it possible to discover a cure for some deadly disease. So, let's not forget the people who brought Linux so far in the beginning. You need us.

To me, as the Community Manager for the forthcoming open source Solaris project, this last sentence is critical. Vendors need to invest in community relations people, true, but that's not nearly enough. Instead, companies need to realize that they are a part of the community on which their software systems are based. So, in my mind, everyone's involved in community relations -- engineers, marketeers, sales, PR pros, and executives. All of 'em. All our jobs, to one degree or another, involve community now. And our ability to learn the culture of community will determine who succeeds and who fails. Certainly, you don't have to lose your identity as a corporation, but it seems to me that you must adjust and contribute if you expect to have the support of the community. Any community.

So what do a couple of authors do to get their product known? Our conclusion became obvious. The open source community has provided places in which ideas are shared, discussed, argued and embraced. If we made correct assumptions, then we have accomplished our mission to have new users switch to the Linux Desktop.

Respectfully submitted