The trolls who visit my blog and deny that there is any merit to using GNU/Linux on the desktop should save their typing.“W. McDonald Buck, retired CTO of World Bank, believes we need to take a more honest and frank look at the Cost Analyses it will take to put Linux on the corporate desktop. He thinks we may be fooling ourselves now. In this four-part essay he will address the real-world scenarios from small business to the the hard and soft costs that big-business CTOs look at when considering the corporate Linux desktop and what we need to know about making it a successful decision.” It’s all been typed before back in 2005 by the retired CTO of the world bank. He knows what he typed but just like all the trolls he gets it wrong…
Back in 2005, XP was in ascendance/supremacy/peak of its game. Folks with thousands of PCs found it worked for them except for the re-re-reboots, malware, slowing down, etc. It was the accepted standard for IT on the desktop and dozens of lock-ins that M$ conspired to put in place all were working well except where they didn’t.
Munich could have chosen to stick with M$ and adopt XP but they didn’t. The real costs of migrating to GNU/Linux proved less than the costs of staying more than one step on the Wintel treadmill. That’s where Buck got it wrong. While there are thresholds that have to be crossed to migrate a business of any size to GNU/Linux, they have to be crossed only once and then the rest is gravy. Do the maths: $latex N X some cost > some cost $ for $latex N > 1 $. Munich studied all the problems Buck outlined and reduced them to some minimum mix of applications and systems to minimize the overall cost and then just did it. It was a lot of work but it paid off.
In schools where I worked the cost/benefit analysis was trivial. Mostly we needed an OS, browser, office suite and networking, basic stuff which can be combined many ways to accomplish the majority of tasks in our system. Very little training was needed. It was a menu-based system largely with a mouse-pointer. In businesses which Buck was considering, the mix of applications of special purpose was almost certainly more diverse with each department in the business having some set of non-Free software applications designed for their purposes. Still, the cost of replacing/upgrading the hardware and software forever periodically is infinite. With GNU/Linux those costs still exist but after the switch to GNU/Linux the costs at every stage are less.
A business doesn’t need a fleet of GNU/Linux guys to run IT. A few will do because one person can easily manage thousands of PCs with FLOSS. There are no licences to count, no networking limitations, no CPUs to count, … They just have to run the software any way that makes sense.
The application-lockin is illusory. Whatever those expensive applications do, a large business or a bunch of small businesses can get together and replace with FLOSS. Linux, LibreOffice, FireFox, Apache, etc. are all proof of that. Google, SUN, FaceBook, Munich, French National Police, government of Brazil, schools all over the world are proof of that. Whatever software is lacking gets written. The world can and does make its own software. The world has more and better programmers than M$ and all its “partners”. There is no monopoly on intelligence.
So, the trolls can continue to beat their drums as the tide of GNU/Linux adoption continues to rise. Maybe when their drumheads get wet they will stop.
Does anyone believe that if GNU/Linux cost more than that other OS that India would adopt it?
Does anyone believe that if GNU/Linux cost more than that other OS that Germans couldn’t do the maths?
Buck was wrong. A proper cost/benefit analysis works for FLOSS and GNU/Linux. It’s the right way to do IT. When Buck was typing, the trolls were claiming that only GNU/Linux geeks used GNU/Linux and that it would never exceed ~1%. They were wrong and still are. I could show data from Ethiopia or Uruguay where GNU/Linux has beaten most versions of that other OS and students are doing it.