Getting Linux Onto the Desktop:
A Small Business Guide

By Mark Rais, author of the new beginner Linux book Linux for the Rest of Us


Organizations can beneficially and successfully make the switch to Linux on the Desktop by keeping in mind two core factors.  By applying the foundational principals behind two factors, even the most techno-phobic and proprietary software entrenched small sized companies can painlessly move to Linux on the Desktop.
First is the Human Factor, which includes the psychological issues with change, any kind of change.  The human factor also takes into account the challenges of individual biases, whether they are against or for Linux conversion.

The second essential factor for a positive and efficient conversion to Linux is directly related to the Hardware Factor.  As obvious as it is, it can’t be understated that hardware and the planning in relation to hardware plays a substantial role in smooth Linux conversion.

Addressing both the Human and Hardware factors in relation to your small business will ensure a far easier and more efficient Linux conversion.  This article tries to summarize principles and examples of how to realistically apply these two core factors.

Human Factor is Number One
For some small business leaders I’ve consulted, the human factor is either a notion to be rejected or ignored.  After all, in a small sized operation, personnel need to “go with the flow” and often concretely follow their business leader.  Very small company leadership, with less than ten personnel, often incorrectly assume that interpersonal issues and technical decisions are not correlated at all.

Yet the human factor is probably one of the most significant in not only how smoothly the actual Linux conversion goes but also how well it is adopted and beneficial to your organization.

One very effective method to encourage even the most recalcitrant personnel to adopt Linux is to take the four-step approach listed below:
Step 1. Introduce the Idea
Step 2. Introduce the Product
Step 3. Introduce the First Result
Step 4. Introduce the Personal Plan

1.  Introducing the Idea is as simple as presenting to your staff a brief review of what the Linux Operating System is (leave out the tech talk), how it currently benefits other organizations (there are plenty of examples on the web right now), and that you are planning to explore the idea of using Linux as well.

Depending on your organization, the time between Step 1. and Step 2. could be as short as a few days or as long as a few weeks.  Wait any longer and you’re going to need to redo Step 1!
“Real encouragement comes when people see with their own eyes a Linux Desktop interface... ”

2.  Introducing the Product should be a simple demonstration of Linux, whether this is a brief tutorial presented to the staff or a more substantial hands-on experience depends on your budget and time.  Real encouragement comes when people see with their own eyes a Linux Desktop interface and how applications are all very similar to their existing desktop and experiences.  When it comes down to the nuts and bolts of daily use, such as opening and using a file, it will be the same basic function whether you’re in Windows or Linux.  People who have only heard about Linux, but who have never seen Linux, need to know this truth.

What doesn’t tend to work out well is either handing out free Linux CDs to your employees and telling them to “go home and try it out for yourself” or giving out cool Linux t-shirts.  In the latter case, it just looks to employees who don’t know enough about Linux like a sale job and basically baloney.  In the former case, those silly “just go home and try it out” lines are bound to get someone who’s already techno-phobic seriously in trouble – like the guy who came to me at the point of desperation having completely cleansed his PC of all of his files and programs and wondering “where they are now that I’ve installed Linux?”

I’ve had to help resolve this latter issue so frequently that I ended up publishing a book for new Linux users (Linux for the Rest of Us, in case you know someone in need), just to save them from the angst and anguish.  Linux conversion can be beneficial and fun, and does not ever have to be frustrating.

3.  Introducing the First Result may be one of the most encouraging steps, not only to your personnel, but also to you.  This first step basically proves the point loud and clear that Linux is the real thing and does promote cost savings or performance gains.  The first result comes from taking the initial step into the Linux world.  Conversion to Linux, even if it begins with a simple network server, will help encourage staff to see Linux as something viable and beneficial. 

4.  Introducing the Personal Plan is in essence the introduction of the plan to convert each person’s computer to Linux.  If you’ve done the previous three steps, this fourth should not only be already assumed and understood by each staff member, but also accepted.  Passive resistance, phobias, and even very real concerns such as “what happens to all my existing files?” need to be addressed before you send your IT tech to place a Linux CD into the drive of your staff’s computers.  Planning and then conveying the planning for each individual staff member’s transition to Linux on his/her PC is essential and must obviously be addressed in conjunction to your hardware plan.

Hardware Factor
You’ve done everything you can to help address the psychological fears and concerns with switching your business computers to Linux.  Yet, there may still be unresolved fears and phobias that could potentially make what should be a smooth Linux conversion into a volcanic eruption. 

You think I’m exaggerating?  In one business setting an employee stood in front of his computer system, wrapped his arms around the monitor and began shouting at the IT tech “you are not converting me to Linux.”  In most cases, such strange and often serious eruptions don’t occur.  Instead, a subtle, discrete negation of your Linux conversion plan begins to brew.  In the copy room, or at the company kitchenette people start talking about how it is a bad plan.
“The Human and Hardware factors must be addressed and implemented in parallel as part of a single plan.”

Why would this exist if you’ve already taken all of the Human Factor steps above?  This results only if your personnel communication plan and your hardware plan don’t coincide.  The Human and Hardware factors must be addressed and implemented in parallel as part of a single plan.  The two factors must be coupled in order to ensure a beneficial and productive conversion to Linux on the desktop.

There are of course many reasonable ways to actually convert Windows machines to Linux for desktop use.  Undeniably, one of the easiest methods is to simply do a complete backup of each PC and then place Linux on the hard drive as a stand alone OS.

This method actually works quite well as long as it includes:
1. a thorough back-up of data;
2. the installation of like or compatible applications (such as SUN’s StarOffice suite or OpenOffice);
3. the restoring of user data and files in an organized and systematic way;
4. and basic training for employees to traverse their new system and get back to work.

However, I’ve found that in larger scope settings, where more than a dozen employees are affected and limited down time can be accommodated, this method leaves too many holes and questions.  For example, what about a back out plan in case something goes terribly wrong with the desktop conversion?  Often, even with data backups and plenty of time, a particular PC will give the installation software a real hassle and consume far more time than it should. 

In one installation scenario an IT Tech took an average of 45 minutes to convert each machine, until he encountered a very differently configured system that had an unrecognizable SCSI hard drive, video card, modem, etc.  It took the tech over four hours to address this single machine’s idiosyncrasies.  Few of the issues had to do with Linux so much as the ancient and unsupported hardware.  There are also times when Linux simply does not include certain core functionality or device drivers that your hardware needs.  In both cases the installation and conversion process will take far longer.  With a reasonably thorough review of potential hardware issues as part of your installation planning, and prior to any installation, delay points can be avoided or at least worked around.

A more satisfactory answer to addressing potential hardware pitfalls is to take the time and add a little extra money to the conversion budget to ensure that redundancy and back-out planning is included.  This will more than compensate in returns including keeping the time to convert individual machines to Linux down, in most cases to less than one hour per each PC.  I’ve had a lot of folks ask me how this is possible, only to discover for themselves that it takes simply a little extra money per PC and some additional preplanning.

The Dual Boot Option
One way to reduce the time of individual conversions and to further resolve the phobias and fears that may exist is by installing Linux onto a second hard disk drive in the PC.  The boot loader would reside on the first (master where Windows is installed) hard drive and allow a dual boot machine.
“I know this may sound like
a terrible blow to Linux purists,
but it also helps hesitant companies
to see a viable plan to move
Linux to the desktop.”

By leaving the existing OS and the existing hard disk in tact, but adding a second drive, you gain an inexpensive means to decrease down time, conversion issues, and potential pitfalls.  I know this may sound like a terrible blow to Linux purists, but it also helps hesitant companies to see a viable plan to move Linux to the desktop.

Specifically, several benefits from this type of hardware installation plan include:
· Far less psychological barriers to desktop conversion since it is a true dual boot system and nothing is “lost.”
· Faster implementation and installation time, although data-backup must be done regardless.
· Compatibility and available application issues are resolved since both Operating Systems coexist.

There are of course drawbacks to this plan:
· Cost per machine for Linux conversion will go up by around $60, which is the cost of a new hard disk drive kit (in my specific conversion case it was with Western Digital 40GB Ultra ATA/100 EIDE at $55 each).
· Personnel have the ability to choose and use either OS, which may not be optimal in fostering a change in mindset and resolving longer term licensing costs.
· The boot loader information will reside on the Windows hard disk drive and if an issue arises with this drive it will make both Operating Systems non-functional.

The good part is that such a dual boot, gradual conversion plan offers reasonable solutions to each of the issues noted above.  For instance, the cost per machine for new hard disk drives is more than compensated in the long term by placing a quality new disk into each PC, which will likely be larger and offer better performance with Linux.  The boot loader concern is addressable so long as emergency boot disks exist for the PCs, to ensure Linux can be recovered even if the boot loader on the master drive fails.

But the human factor of personnel sticking to their old operating system rather than using Linux in the dual boot is also addressable.  To make sure that even in a dual boot scenario office staff do not cling to their past desktop but acclimate quickly to the new desktop simply requires a “conversion plan.”  All of the employees know they will have a dual boot system that is fully functional for the first few weeks (a set amount of time defined by you in the plan) to ensure all of the kinks are worked out and that they will be able to keep working and operational during the conversion.  At the end of the first few weeks, most of the data and programs used on the Windows drive will then either be deleted or moved to the Linux drive by the same IT person doing the initial conversion.

This pragmatic approach forces personnel to focus on learning and using Linux within set and reasonable timeframes, rather than reverting back to booting into their old OS for daily tasks.  It also reduces anxiety about the conversion to a Linux Desktop because you’ve basically kept the entire old world in tact for a time as a fail safe.  Those programs or issues that have not yet been addressed or resolved for Linux can still reside on the original drive and remain operational as long as they are needed.  This is especially important if a particularly vital application does not yet exist for Linux.  It also addresses the issue of forcing the conversion into tight deadlines.  Employees enjoying the new Linux desktop, getting acclimated, while still having the old OS and data around makes a world of difference for the IT tech dealing with hardware issues and installations.

In Summary
A smooth, beneficial and productive conversion comes as a result of addressing both the fundamental Human Factor and the Hardware Factor in your Linux Desktop plan.  Such a plan also resolves many small business concerns and phobias of having Linux on the desktop.  Concerns can easily and systematically be addressed to gain the substantive and very tangible benefits of licensing and maintenance cost reduction.  Linux on the desktop, as a result of a smooth conversion, is also likely to address a number of performance issues as well.

As more and more small businesses are realizing today, Linux belongs not only in the server room, but also on the desktop, offering cost reduction and future freedoms from licensing and proprietary software.  And most importantly, a conversion to Linux on the desktop is a reasonable and attainable goal.

Mark Rais authored the recently released beginner Linux book Linux for the Rest of Us, and has written articles on various small business Linux topics including Moving to Linux: A Small Business Guide,  as well as serving as a senior manager in the IT industry for six years.

More articles on switching to Linux are available from

Copyright © 2003 Mark Rais.  Published to the web by REALLYLINUX.COM xv31503
Legal Information.

Linux is a registered trademark of Linus Torvalds. All trademarks and registered trademarks on this entire web site are owned by their respective companies. The contents presented are computer consultant opinions and not to be assumed as fact. For factual information refer to more than one source and judge for yourself.