I work in a municipal government where Windows is used throughout our institution. People use computers in our institution mainly for browsing the internet and using MS Excel or Word. I was thinking of suggesting that we change to a GNU/Linux operating system, such as Ubuntu. (Primarily because I think the annual license costs for using Microsoft systems takes a big cut out of the municipal government's budget.)

Can a large organization switch from Windows to Ubuntu? (Is it feasible, will it save money and how can I demonstrate this?)

P.S. One concern people in my institution have raised is that we need secure operating systems and that Ubuntu is not safe enough. They think that because Ubuntu is open-source then it is not safe enough?

  • 8
    This is not a discussion site. Try ubuntuforums for questions like this. In my opinion: users need to want to. – Rinzwind Oct 14 '17 at 14:50
  • 22
    Because our government, health care systems, military, and schools should not train our people and children to be dependent of any corporation to function and alternates to Microsoft and Apple must exist period. – Panther Oct 14 '17 at 15:08
  • 7
    It is opinion based but it is an important issue see Ubuntu bug No1 - bugs.launchpad.net/ubuntu/+bug/1 – Panther Oct 14 '17 at 15:11
  • 8
    "One concern people in my institution have raised is that we need secure operating systems and that Ubuntu is not safe enough. They think that because Ubuntu is open-source then it is not safe enough?" Common misconception, security by obscurity. You could mention Linus's Law and the fact that most websites use Linux. Facebook, Google, etc. – Léo Lam Oct 14 '17 at 17:18
  • 21
    @PålGD: But, please read down to the comments, where it is explained that the article is no more than sensationalist click bait. In reality, the mayor commissioned a report about the future of Munich's IT infrastructure. Nothing more. There was no "considering switching back". The report was simply open and unbiased, and thus also among many other options explored switching back to Windows. I also don't think that Munich is a good example. Linux on the Desktop was much less mature in 2003 than it is in 2017. Much more stuff is web-based and thus OS-independent now than in 2003. – Jörg W Mittag Oct 15 '17 at 17:41

13 Answers 13


Sensible organizations don't choose their computer systems because of the cost of the operating system. They choose them because the system supports the business operation at an acceptably low level of risk.

If you compare the potential cost of disrupting the business to the cost of continuing to run Windows, quite likely the cost of Windows licenses doesn't even show up on the radar. I would expect an average-sized municipal government organization could incur costs of millions of dollars/pounds/[insert your currency here] per day from an organization-wide IT disruption that took more than a few minutes to fix.

If you want to make a case for change, forget about computer based technical arguments, and grand philosophical propositions about the evils of closed systems and the delights of open ones. Leave that sort of thing to people like Richard Stallman. Make a business case instead!

And don't forget to include the costs (financial, disruption, temporary loss of productivity, additional support required, etc, etc) of retraining every computer user in the organization to work with new systems and software that most of them may never have even heard of, let alone used, and will see no good reason for "change just for the sake of change".

The above is not meant to be "anti change" or "pro Microsoft." It's just being realistic about how an organization works - and that is something which many IT staff seem completely ignorant about.

  • 25
    You bring up an important point. However, democratic institutions are driven by other incentives than sheer business cases. – David Foerster Oct 14 '17 at 20:29
  • 5
    I don't believe Stallman , FSF, or the pro opensource community is suggesting we make change in such a drastic means that we bring such institutions to a standstill as you suggest in this post. Most of your post applies to any OS as there is quite a loss of productivity with windows 10 and the new look and feel of microsoft office, certainly takes the average person more then a few minutes to find their way around the new layout of windows 10 and the new office interface. By your logic we would all be running DOS and wordperfect. I would agree, however, opensouce laggs the most in busness apps – Panther Oct 15 '17 at 11:59
  • 4
    @bodhi.zazen The time it would take to train someone from windows 8 to windows 10 would be negligible compared to moving from windows (any) to ubuntu IMO. When I moved up I watched their little walkthrough (about a minute and a half) and felt right at home. I do understand your point though, and I've worked with computers all my life, so I'm not exactly the average user. I could just imagine trying to teach ubuntu to my mom, and I'm cringing. – user746008 Oct 15 '17 at 18:04
  • 3
    For most end users the transition to Ubuntu is as easy as the transition to windows 10. Logging in and opening a program such as open office is just as easy. I have not really had any difficulty with end users. End users do not do sys admin , they call the help desk every time they have a problem, so the training for end users is negligible. Now the IT department is different. Just fire them and hire people how are RHEL certified done. I would cringe at the thought of training you mom too =) – Panther Oct 16 '17 at 1:21
  • 3
    @bodhi.zazen I think we live in different worlds. The thought of deploying any linux distro in my workplace frankly gives me nightmares. – J... Oct 17 '17 at 11:57

This is a critical issue to Linux and Ubuntu as a whole, see bug No1


The 2 issues are:

  1. Because our government, health care systems, military, and schools should not train our people and children to be dependent of any corporation to function and alternates to Microsoft and Apple must exist period.


  1. I side with GNU / FSF on this issue, in the digital age people have the right to examine, modify, and distribute the software that runs on their hardware.


Our mission is to preserve, protect and promote the freedom to use, study, copy, modify, and redistribute computer software, and to defend the rights of Free Software users.

This does not mean free as in free beer, people are free to charge for modification, distribution, and support of such code it means free as in the freedom to examine, modify, and fix (source) code without being dependent on such support systems.

Nor does it mean that I agree with everything Stallman/GNU/FSF proclaims, but I agree with the fundamental philosophy of GNU/FSF .

The "problem" is that our social systems are dependent on Microsoft and you can not ask schools / health care / business / government to stop functioning for the FSF philosophy, but you can make them aware they do have alternates and ask them to support such alternates.

Chromium, Libre/Openoffice, Firefox, and other open source options do exist so why not try them and consider paying to support them same as they use and support their current corporations.

Open source communities need to develop alternates to the systems our society uses and the most glaring deficiencies, IMO, are business and professional related apps, accounting, photography, digital processing, medical records, and voice recognition technology. It is sad my Android phone has better voice recognition out of the box than Linux desktops.

My current employer is in fact using Chrome and Open Office on windows rather than IE and Microsoft Office so it is a work in progress but a step in the right direction.


Migrating to Linux is no different than migrating to any other OS, including windows 10. I do not know any business that simply rolls out Windows 10, or any other OS without prep.

  • IT Does the IT department have someone who knows the new OS ? If not they will need to train or hire someone who does. For Linux I suggest someone who is RHEL certified or similar. I have never seen a business roll out an OS the IT department is not intimately familiar with. Windows 7/8/10 was "under consideration" for many many years and many businesses remained and still remain on windows 2000 rather than migrating. Migrating to windows 10 is not as simple as claimed by the pro windows posts here and certainly was not in any way overnight.

  • Hardware. All OS, windows included, have hardware requirements. Many of the problems here, on this site, and other support channels are hardware related. Business does not purchase hardware with no OS installed and then struggle to install windows, track down the drivers, etc, they purchase hardware with windows pre-installed or they purchase windows compatible hardware nand generally contract out hardware support. They need to do the same with Linux, purchase computers with RHEL / Ubuntu pre-installed and contract out support with RHEL or Canonical, same as they do with Microsoft.

  • What software is needed ?

    • Linux can run some software and most if not all servers needed.
    • Desktop software that is no problem would include "desktop" functions such as web browsing, email (many email servers have web interfaces similar to gmail), word processing (Liibreoffice is similar enouhg to word and it is rare an end user would need something that can not be done in Libreoffice. Biggest problem with libreoffice is converting documents), PDFs are fine, etc.
    • Server side Linux works just fine as well. Linux can easily handle most if not all servers including web servers, ftp, mail servers, file servers, active directory, virtualization, etc.
    • Specialized software - Linux struggles with certain specialized software for example voice recognition, digital image processing (Photoshop), accounting (Linux just does not have a program to manage accounts receivable and payroll), video editing, music studio. Linux has programs that do some of these features but for the most part not up to par with Windows.
    • Propriety software - Many business use propriety software that is either specifically written for them or is not available on Linux. generally these are database driven apps such as medical records or accounting apps. Although there may be some attempts on the linux side, such as http://www.open-emr.org/ , generally they are not on par because business has not spent the millions on them they have the closed source options. https://www.americanactionforum.org/research/are-electronic-medical-records-worth-the-costs-of-implementation/ . 5 physician practice spend more than 200,000 in the first 60 days for a EMR. The reason I started this post with philosophy is we need to divert the funds spent on propriety software to open source. As it is now, the fact of the matter is most if not all of these propriety specialized applications want to lock business into long term contracts and more often than not there is no way to convert from one system to another.
    • I just used an EMR as an example as I am guessing people are going to understand that example better than others. Imagine the type of database amazon.com uses to track users on the web site, inventory, shipping, suppliers, reviews of products, etc of all they offer.
  • Training - They need to train the help desk and support staff same as with any other OS. No business would roll out windows 10 without training the support staff.
  • Implementation. No business rolls out windows 10 overnight as suggested by some of the windows supporters here. Doing so would cripple the business. You would roll out linux similarly , a certain quota of machines starting with the simple tasks and progressing to the more complex tasks in a measured step wise fashion with the old system to fall back to. Many businesses have a limited number of "power users" more or less beta test and then help train. Again, using an EMR as an example, an office of 5 physicians may have one of the 5 start using the new EMR on one or two encounters a day, falling back to the old if there are problems. Over days or months they would migrate to the new EMR and then bring the other 4 up to speed on the new much faster. No business would migrate to a new OS, including windows 10, or a new large software application, like an EMR, overnight and the suggestions of many posts here from windows supporters are frankly unrealistic and not the way business does things.


Government is going to go through a similar process but much slower and more drawn out. Look at how they handle any large scale project, there will be meetings, analysis, political whims, pilot programs, migration plans, funding crisis, etc.


Schools and Universities need to teach out children and young adults open source solutions when available. This will obviously take longer, but as a parent I can tell you the schools my children attend have zero tolerance for open source OS and open source programs. Try submitting a paper in anything other than a .doc or .docx, lol . At least universities are starting to teach open source. Hopefully with bash on windows, firefox, libreoffice, and other successful open source options the migration will be easier moving forward.

Bottom line

  • Like a drug, we as a society did not become dependent on Microsoft overnight, it took some time, decades in fact and we are not going to migrate to open source solutions overnight.
  • The posts made here by the supporters of Microsoft are unrealistic and oversimplified.
  • No one in the open source community is advocating an overnight change or bringing business to a stand still.
  • The process of change starts with fundamental philosophy, do we as a society want to be dependent on Microsoft ? If you answer that question with a "yes" or a "who cares?" change will not happen. Change starts with a change in ideas, perceptions, and goals, and an implementation plan follows. This is evidenced by the pro Microsoft posts here, most of the posts are at step one, not even considering change.
  • 5
    Maybe you can't see the irony in "Bug #1". Reported in 2004, still not fixed in 2017. Why would any organization want to rely on a system with that abysmal level of support? ;) – alephzero Oct 14 '17 at 18:28
  • 4
    @alephzero: The only reason why Microsoft software has good support infrastructure is because of money (from its customers). The same would happen if organisations showed interest in and thus created a market demand for Linux support. What's missing is a critical mass for the latter to become feasible and competitive. As with almost all technology markets, it's a hen-egg acceptance problem. As bodhi explained democratic organisations have additional incentives over other organisations to use free, open-source software and are thus in a better position to spear-head this critical mass. – David Foerster Oct 14 '17 at 20:23
  • 4
    This critical mass was achieved by Linux and support providers a long time ago in other market segments (i. e. servers) btw. I see no reason why it couldn't be achieved on the desktop and for groupware and ERM-like applications. – David Foerster Oct 14 '17 at 20:25
  • 4
    @alephzero Bug 1 is closed bugs.launchpad.net/ubuntu/+bug/1/comments/1834 (but not for the reason we all hoped) – lofidevops Oct 14 '17 at 20:44
  • 4
    ... Microsoft (unlike Apple, who make incompatible changes every OS release) have learned the basic business lesson that long-lived organizations want tools that have long-term stability - and "long term" here means "50 years is better than 25," not "you can download a new latest build of the app every night." But then, IMO Apple's strategy has been to reposition itself from "computer system supplier" to a company that markets expensive consumer toys to bling-lovers. There's nothing wrong with that - so long as the customers understand what they are really buying! – alephzero Oct 16 '17 at 4:51

Your argument is sound. In the long run, free software lowers costs because it keeps vendors competitive. (As an organization you will always depend on some vendor for service, even if it's in-house IT staff.) In the medium term, you need to consider that there need to be enough vendors around to make prices low. But in the short term, there are financial and managerial costs to making the transition that are difficult to swallow.

(Along the way you will also encounter nonsensical arguments like you described. Make a note of them and address them, possibly with the help of a vendor, but they are a smaller issue than costs and change management.)

Start talking to several vendors about those costs. An obvious consideration would be Canonical themselves. They offer commercial support called "Ubuntu Advantage". They specifically have experience with migrations, so they should be able to advise you on feasibility. You should also investigate if there are any local vendors in your city / region / country (they may support different distributions). What are their maintenance costs? Costs to train your IT staff?

You'll probably need to build up some case studies. Consider similar projects and why they succeeded or failed. Someone in the comments mentioned Munich (where LiMux is installed), and results there have been mixed not least because of vendor delays and political opposition. You can find other examples at the LiMux link. An interesting one is the ISS switching to Debian as their laptop operating system. Skippy points out the GendBuntu project in their answer. It is an excellent case study not just because it continues to be successful, but because it is well-documented and security is an important requirement.

You may want to contact the FSF for advice. They have an ethics-based approach rather than a financial one (so they wouldn't recommend Ubuntu, for example) but they may be able to point you to case studies as they are particularly interested in getting free software into government.

You could propose a single-user pilot project, where you are the test case. The cheapest way would be to dual-boot Windows/Ubuntu on your work machine -- but that's why you tagged your question right?

Trying to make this small step happen will give you a good feel for:

  • how much influence you might have on such a big decision
  • any entrenched opposition (e.g. an IT department who aren't comfortable shifting / worried about job security) and potential allies
  • immediate infrastructure issues (can you authenticate against your existing LDAP, can you function without sudo privileges)
  • short-term usability issues (can you open documents sent from other offices, do you inadvertently rely on PDF Forms or Flash, can you get your email and can others read yours)
  • your ability to demonstrate that you aren't losing time to these issues

As you can see even technical issues are just the root of organizational ones. Once you have your bearings you might want to post more specific questions about motivating for change at The Workplace Stack Exchange. You will get very realistic assessments and advice. But come back here for technical questions :)

Finally, please blog about your efforts, even small ones, so that others may learn. (Try get them syndicated, would Planet Ubuntu be appropriate? And obviously use a free license like CC-BY-4.0 so that others may build on your work!)


Can a large organization switch from Windows to Ubuntu?


Proof: the French National Gendarmerie has made such a transition on more than 80,000 computers. They needed to develop their own distribution, GendBuntu, but of course as a police force they have very specific needs. See the GendBuntu Wikipedia article for details about the deployment timeline.

More anecdotically (577 computers), the French National Assembly has also switched to Ubuntu since 2007.

Will it save money?

It seems so. Quoting this outdated article:

A report published by the European Commission's Open Source Observatory provides some details from a recent presentation given by Gendarmerie Lieutenant-Colonel Xavier Guimard, who says that the Gendarmerie has been able to reduced its annual IT budget by 70 percent without having to reduce its capabilities.

Is it safe enough?

Well, the French National Gendarmerie use it.

  • this is an excellent and well-documented case study! thank you! – lofidevops Oct 18 '17 at 10:00

Your question is going about things the wrong way. It's like saying "Here are the wheels I want". Now what kind of car will fit them?

First you have to look at the size of your municipality. How many employees what application software do you need?

For example you might need water & sewage billing software. What operating systems and hardware does it require to run?

You might need payroll and benefits software. What operating systems and hardware does it require to run?

You might need fixed assets, inventory control, equipment preventative maintenance and work order software. What operating systems and hardware do they require to run?

About the only thing I see as Linux likely only I see for Municipality is the website software. For all the rest of the application software I imagine your options are 95%+ Windows only.

When you are paying $100 for Windows on a client PC you are often spending $1000 per seat to run the application software which costs $100,000 plus for the core on the server(s). But writing the software from the ground up on a different OS (like Linux or Windows) can cost over $1 million.

Let's not forget any city > 30 years old may have IBM S/390 mainframes and AS/400 minicomputers running their mission critical applications. They may have Windows PC clients (using terminal emulators) that can be converted to Linux PC clients on a department by department basis.

Don't restrict yourself to the "When all you have is a hammer everything looks like a nail" approach.


Speaking from experience, I can say this transition is doable, rewarding in every way, and frankly I don't know why this is even still a question today.

Anyone saying Linux or open-source software "isn't safe" has no idea how networking works -- they're certainly already using several Linux-based devices and servers in city buildings right now. Linux and UNIX machines vastly outnumber Windows installations today; you can't use the Internet without interacting with hundreds of Linux machines per hour.

Next time someone tells you they don't want to use Linux, UNIX, or open-source software, remind them that their smartphone runs one of those two operating systems. And yeah, no more Internet.

I've had some success getting around prejudices by explaining to people what the RFCs are, and how those are the Internet, and the relationship between open standards and open software, but that's a longer discussion.

Having said all this, and institutional inertia being what it is, I think you'll have an easier path advocating a switch from Windows to Chromebooks or ChromeOS workstations for the average user, with Ubuntu for more complex workstations only. If you really want Ubuntu everywhere, then look at techniques that are ChromeOS-like in management: containerize as much as you can, use snap, look at management tools such as Chef. Other than virus prevention, the real win in institutional Linux is the lower TCO and higher reliability from automated administration.

Either way, step one is to set up a Google Apps domain for the city, and get people started using google docs, deprecating Word and Excel. You'll see immediate productivity gains just from the collaborative editing capabilities. If you're in a situation where this can't be done for legal or firewall reasons, then it might not be time yet for you to push the migration. You might be able to accomplish similar "cloud-like" results with internal servers and something like ownCloud.

In California, Chromebooks in each classroom have just about supplanted the Windows-based computer lab in most schools. This provides a more secure, more manageable platform with a much lower total cost of ownership; so low that schools are approaching a 1:1 ratio of systems to students, where before they could only afford around 1:20.

Another data point: The New York financial industry figured out a couple decades ago that Linux TCO was a fraction of the cost of Windows. Even UNIX, with its higher license fees, was cheaper, since most of TCO is labor. I used to build trading floors; by the time I left NYC, we had 5 guys managing thousands of UNIX machines with primitive tools we wrote ourselves. The ratios just don't favor Windows, and never have.

Google around for Gartner Group and other consulting agencies' reports on Linux TCO -- you should find more stories like this there.

A final data point: among other things these days I'm the "IT guy" for our family business. We have around a couple dozen linux machines and one Windows machine. That Windows machine consumes about half of my sysadmin time, and the only thing it runs is quickbooks. Not cost effective at all.

Windows started life as a personal computer UI, optimized for home use by a single user, and even after gaining its own kernel it's always struggled to grow out of that mindset; the Windows development community is still stuck there. In terms of security, it's a toy. In terms of institutional use, it's completely inappropriate.


Let me try. These would help to ease their pain.

  1. Cost of OS and applications:

    • Windows is more expensive to run.

      Yes it would be expensive to make a switch but in the long run, you will spend less on Linux compared to Windows.

  2. Safety:
    • Let them see the current malware statistics with respect to Windows and Linux (a.k.a Ubuntu) incidents. See here and here among others.
  3. Software:
    • Show them free alternatives to costly Windows applications. Linux holds its own with lots of good alternatives to Windows applications.
  4. Hardware Cost:
    • Linux runs better on old hardware as compared to Windows. Linux out performs in this regard as it can easily be run on an old hardware when compared to Windows, hence making buying a new system less likely.
  5. OS Support
    • We have both Canonical and RHEL providing one of the best paid support for Linux systems out there. With the standard of today both REHL and Canonical provide one of the most outstanding paid support available and comparable to that of Windows.

It's a fairly complex calculation, because you also need to take a lot of other cost factors into account:

  1. Windows has a lot more built-in support for central management. You need to replace this with your own system, so you need an experienced Linux administrator team. During the transition period, expect to have both a Windows and a Linux team.

  2. Software might need to be ported to run.

  3. Users need training.

This can certainly be done, but it is a multi-year process where you convert one department after the other.


You have some great answers already, so let me briefly share my personal experience on a related matter of theoretical as well as practical interest.

A lifetime ago I switched from Qwerty to Dvorak. Both are "open source" and yet the switch is costly. The costs are non-monetary and yet they are fundamental to the decision to switch or not. The benefits are: comfortable typing, speed, being part of the cool gang. The costs are: about one month of lost productivity, having to install a Dvorak keyboard every time you use someone else's computer, being part of the ostracized gang.

Was the switch easy? No!

Was it worth it? Absolutely!

Would I recommend it? That depends: how serious is your RSI? My Repeated Strain Injury was so bad that I could hardly type, so in effect I experienced no net loss in productivity by switching from Qwerty to Dvorak. My RSI has never reappeared and I have never looked back. (To understand why Dvorak solves RSI, it has to do with the position of the letters being more finger-friendly for the English language, and in my experience exact same benefits in several other European languages I use)

Note that studies purporting to show that Dvorak is no better than Qwerty are always conducted by Qwerty users (and vice versa). For similar reasons, you will have a hard time getting objective opinion on Windows versus Ubuntu.

To de-analogize: if your institution is currently suffering from using Windows, either because Windows is repeatedly down from Virus attacks, slowed to a crawl because of competing anti-malware, or busy getting the latest Windows update, ask yourself whether Ubuntu will solve these problems. In my experience, these Windows issues will be solved and gone for good. Great... except they will be replaced by Ubuntu-specific issues (usually some missing drivers).

Is Ubuntu really superior? In theory yes, because... open source is superior, right? but in practice, it's not like Windows costs a fortune, does it? To me they are close substitutes. And therefore, in my view, the crucial question is: will Ubuntu solve problems you are having with Windows?

If you have no specific problems, it is hard to see how to justify a switch: the time it will take employees to adjust to the new look of Ubuntu will itself be a distraction costing several thousands/billion dollars (depending on the size of your local government).

Examples of why I prefer Ubuntu: all the free software that comes with the package managers and is so easy to set up and such a mess with Windows. But your mileage may differ.

One way to entice employees to desire a change would be to distribute Ubuntu computers to volunteers. Make sure they have all the drivers!


I don't know which country are you from, but many legislations like Russia and Italy are committed to Open Source Software, Open Document Formats and Open Data.

So, when a government entity need software, it is mandatory by law to consider OSS first.

That means if there are no objective needs to run a specific OS (like if we have some piece of software/hardware which runs only with a specific OS like Windows or Mac OS) then an open source OS shall be used.

The same thing about Microsoft Office; if you really need it you can use it, but citizens should receive, or at least be able to receive under request, documents in an open format. If I'm not, I can always fill a complaint.


I have considered this question myself in the past and whilst I see your point I think the risks / costs far outweigh the advantages here:

  • There is likely to be a significant amount of disruption changing an entire PC (and server) estate from one OS to another - what is the real cost of that?
  • There are both financial and time costs for end users to be trained. How many users have used any other desktop OS to Windows? How many are even aware other OSes exist? How many are even competent with Windows let alone something else they have never used? If there are 1000 users in the business both the financial and time costs are massive.
  • What is the skill set of the IT support team? If there are no Linux skills in the support team, that's an entire department who need training / replacing (at a time when they will be needed most)
  • Compatibility of current systems - you say users only ever use Word / Internet etc but I'll bet there's Active Directory in the background. What is involved in implementing an alternative? Other compatability issues may include smaller things such as Word docs - yes you can open open a .doc file in OpenOffice / LibreOffice but it does not always look 100% like it does in MS Word, another reason for your already fed up with this users to give further criticism.
  • Finally, as I have eluded to above - staff goodwill. How may staff is this going to annoy to the point that it affects morale and goodwill? Some staff are IT haters at the best of times and a minefield such as this one could send them over the edge

So yes, you may save in the long run not having licence fees etc but the cost of implementation to the business will be massive and other non tangible.

I don't mean to paint a glum picture but my experience is that change is rarely seen positively unless there is some major benefit. The only benefit I see here is cost savings (if any) which even if realised only please management.

Caveat - I have no experience in any project such as this and the above is just my opinion

  • 1
    Your observations are the end result of training out population from childhood to use windows and holds little to no water. All these issues apply when upgrading to windows 10 for example which is nothing like previous versions and is quite confusing as the OS interface and microsoft office interface has changes quite drastically. On the other hand, for the most part, firefox, open/libre office, and many open source projects can be implemented and most people in an office setting do not need all the features of microsoft office. – Panther Oct 15 '17 at 12:09
  • 1
    "All these issues apply when upgrading to windows 10 for example" - really?? When I migrated from Win 7 to Win 10, the learning curve took me about 15 seconds to start doing productive work - which was considerably shorter than the time to install the upgrade! Oh, and I tried open office once on Win 7 - and deleted it after 24 hours as some sort of silly joke, not an office productivity suite. Sheesh, even MS Office 98 was more functional and way more reliable than that pile of poo! – alephzero Oct 19 '17 at 4:54
  • The simple fact that everyone is just talking about "Linux" says a lot. Nobody is bothering to specify which of the many competing versions they are actually referring to! I've used enough different versions of **ix to know very well that the devil is in the detail here. – alephzero Oct 19 '17 at 4:57
  • I agree about OpenOffice. Even as a power user and persisting for over a year with it, I just had to go back to MS Office and all has been glorious since! – SEarle1986 Oct 19 '17 at 17:04

My birth country's government once decided to do just that, on a federal level. All agencies and institutions in the executive branch were to migrate to open software by X date. It did not work, but its shortcomings may help with your case.

1. People are avert to change

You likely have a bunch of people who have been doing things one way since ever. They will not like the disruption (remember when Microsoft tried to abolish the Start Button in Windows 8?). One thing my birth country did, which was a good thing imo, was baby-stepping it by first switching Microsoft Office for LibreOffice. It's something that deeply messes with people's habits, but is easy to revert in case of a total productivity meltdown. You'll get a good view of their propensity to change with this.

2. Training is a must

This was perhaps the biggest reason the FOSS initiative failed in this case, the government just switched the softwares and told people to get on with their lives. No matter how similar they may seem, there are always differences, and some folks won't pull one step forward without having their hands held. Especially old timers who already struggle to do everything digital at all.

In my country's case, after the initial shock they hurried with courses, but it were too many people, and by the time many finally got training, weeks of frustration had gone by and they developed an internal blockade that "this software is no good" or "I'll never manage to learn this", and learn they didn't. Prepare all your training beforehand, and switch clusters at a time instead of everybody all at once.

3. What about support?

The main difference between a home office or small business and a big business is that the big business cannot fail for even a second. So support is a must. If your municipality is way too small you may be able to go by with just a few IT guys, but if you deal with sensitive information, or have more than 100 people working there (random threshold btw), you will need a support network working around the clock.

There are many such services for Linux, Red Hat likely the most famous. If you can't get it for whatever reason, a robust IT team is necessary, and that is not easy to gather. One agency changed a particular type of software for a FOSS alternative a few years ago, and it's a pretty successful story, but they have a large dev team to depend on.


It is possible, and admirable, but this has to be thought through well. Adaptation problems aside, an OS is not something you just go and uninstall with ease, you must have a solid bailout plan in case everything goes wrong. I'd advice you to start small and with a select group of more tech-oriented workers, and preferably with a double install (or a VM if you have the memory). Observe and gather data for many months, then do a new round of trials and so on. You're probably looking at a 1-2 year long switch, depending on how big your local government is, and how much the upper management will support you.

On a side note: once my country's government decided on the switch, Microsoft came onboard with an irresistible offer, slashing prices of OS, Office and support alike. Let MS know you plan on abandoning the ship, you may just get yourself a sweeter deal!


I think this list of Linux adopters might also help. It would be useful to show evidence of this sort of migration being done before with success and certain gains have out of this process

Not the answer you're looking for? Browse other questions tagged or ask your own question.