Are there any studies done on how much linux can save governments money if they do a whole migration?
Are there any studies done on how much linux can save governments money if they do a whole migration?
Contact Mathias Waack. He can probably give you plenty of resources.
mathias.waack@schwaebischhall.de
He manages the only successful Linux based infrastructure in governmental institutions in Germany.
That's interesting. Genuinely interested in how you can have desktop Linux pcs managed as effectively. I would love it if institutions in the U.S use Linux for their desktops.
Any thoughts about doing the same for Canada, Trudeau?
Cnada
Microsoft are deep into the government with exchange and Active Directory with most being migrated to Microsoft365 and Azure.
Add in MS Teams, SharePoint, MS SQL, 30 years of business rules living in old excel macros that ends up running the entire company.
Windows enterprise licences would be a tiny part of their spend and far too costly to mitigate away from. Most large corporations are virtualising old windows version just to keep their existing legacy apps runnings.
Totally agree on Microsoft having a vested interest with being the US Gov main software provider and the spend a lot to keep it that way
I am a pretty big fan of Open Source and have used Linux myself since the early 90’s. Most governments are not going to save money switching to Open Source. At least not within say the term of a politician or an election cycle. Probably the opposite.
This kind of significant shift costs money. Training costs. Consultants. Perhaps hardware. It would not be at all surprising if there are custom software solutions in place that need to be replaced. The dependencies and complexities may be significant.
There are quite likely to be savings over the longer term. The payback may take longer than you think though.
I DO believe governments should adopt Open Source. Not just for cost through. One reason is control and a reduction of influence ( corruption ). Another is so that public investment results in a public good. Custom solutions could more often be community contributions.
The greatest savings over time may actually be a reduction in forced upgrades on vendor driven timelines. Open Source solutions that are working do not always need investment. The investment could be in keeping it compatible longer. At the same time, it is also more economic to keep Open Source up to date. Again, it is more about control.
Where there may be significant cost savings is a reduction in the high costs of “everything as a service” product models.
Much more important than Open Source ( for government ) are open formats. First, if the government uses proprietary software, they expect the public to use it as well and that should not be a requirement. Closed formats can lead to restrictions on what can be built on top of these formats and these restrictions need to be eliminated as well. Finally, open formats are much, much more likely to be usable in the future. There is no guarantee that anything held in any closed format can be retrieved in the future, even if the companies that produced them still exist. Can even Microsoft read MultiPlan documents these days? How far back can PageMaker files be read? Some government somewhere is sitting on multimedia CD projects that can no longer be decoded.
What about in-house systems that were written in proprietary languages or on top of proprietary databases? What about audio or video in a proprietary format? Even if the original software is available, it may not run on a modern OS. Perhaps the OS needed is no longer available. Maybe you have the OS too but licenses cannot be purchased.
Content and information in the public record has to remain available to the public.
The most important step is demanding LibreOffice ( or other open ) formats, AV1, Opus, and AVIF. For any custom software, it needs to be possible to build it with open compilers and tools. Web pages need to follow open standards. Archival and compression formats need to be open.
After all that, Open Source software ( including the OS ) would be nice. It bothers me less though. At that lobby, it is more about ROI and Total Cost of Ownership. Sometimes, proprietary software will still make sense.
Most proprietary suppliers actually do stuff for the fees they charge. Are governments going to be able to support their Open Source solutions? Do they have the expertise? Can they manage the risks? Consultants and integrators may be more available, better skilled, amd less expensive on proprietary systems. Even the hiring process can be more difficult as local colleges and other employers are producing employees with expertise in proprietary solutions but maybe not the Open Source alternatives. There is a cost for governments to take a different path from private enterprise. How do you quantify those costs?
Anyway, the path to Open Source may not be as obvious, easy, or inexpensive as you think. It is a good longer term goal though and we should be making progress towards it.
Good comment. I'm pretty sure "public money, public code" used to be a slogan a while back. It didn't get a lot of traction but it resonated with me.
A comment on the open format part:
I got insanely pissed off by recent news (even though I usually laugh at bad news) that the Czech Government cannot have an interconnected eGovernment system between different organizations because each part is made by a different supplier. Jeez just make a fucking github repo, it's not that hard. I just don't get how they can be this stupid, especially considering part of the Government is the Pirate party, The supposed IT guys. These "suppliers" aren't supplying airplane parts, it's merely software.
rant over. I tried to keep the f-bomb count to one, but I'm telling you, it was f- insanely hard. Writing this comment and remembering that atrocity reignited my rage.
edit: explicitly added "eGovernment" system
French gendarmerie moved to Linux like a decade ago. They certainly have something about it.
Well, France exported a lot of socially important technology to underdeveloped nations. For example their "La Révolution" we imported in 1917 was core component of improving quality of life including reducing workday by 4 hours down to 8 hr/d, 5-day work week, universal literacy and later full universal education including higher education, universal healthcare and universal housing. Another technology we imported in same period was "Etat laïc", but we lost it in 2010s.
France certanly has something about it.
There's a bunch of things french are doing themselves instead of importing exactly because they want to maintain technological independence and promote local industry.
I wasn't clear I think but I meant that they certainly have data about the migration to Linux.
It’s shameful that I feel the need to preface this by saying that I’ve used Linux for 26 years now:
The consensus is that it’s a massive cost increase rather than savings.
Well, I would argue that depends very much on the basis of your calculations. Closed source software means public services are held hostage after a company winning a contract. In Norway some Finnish company won a contract for some digital system in the health services and later wanted them to ship all their computers to Finland so that they could update their software. In a paradigm were governments commited to Linux and open source software, there would most likely be a lot less overhead in adapting and developing solutions for Linux.
I actually agree with you, under communism we could run public services on open source software no problem.
When the externalities of training people to use that software, integrating with outside systems, using state power to influence standards&norms and contributing back to the development only exist on the balance sheet of the switch though, it’s not possible.
You would have to calculate it assuming that msft wouldn't deliberately make the process more difficult and impractical, which they have demonstrated they are willing to do.
(Refer to the section labeled "The Microsoft Playbook": https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html)
Fixed link. https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html
Edit: I didn't have time to read all of it but what I did read was very interesting.
I believe the question is missing somehow the main points... even if the switch cost the double or triple there are several strategic advantages that should take into account:
- use Linux allow to growth the number of high specialized professional workers, investing on local resources;
- invest in a local network of specialized companies, instead of financing the silicon valley with ours public money;
- be less dependent by abroad technologies, get a major control of the system used;
These are few that come to my mind...
What would be really interesting to know is the percentage of the investment that stay in the region/country following a linux-based/opensource IT infrastructure for public bodies vs the current closed M$|OSX paradigm.
I agree. A lot of profits wouldn't be cash saved, for one taxes that you aren't losing to multinational corporations headquartered in Ireland or Cyprus.
Cybersecurity costs would also likely go down due to most malware being exploited isn't targeting desktop Linux.
Cybersecurity costs would also likely go down due to most malware being exploited isn't targeting desktop Linux.
Which is going to change once any sort of widespread adoption happens.
But at least in my circles, malware really isn't that big of a deal in security. Phishing is where the danger is these days, where the costs occur.
You know how you make the expertise in a large organisation?
You call for volunteers to run a pilot, move one team or a product to open source alternatives, learn what skills are needed in your tech people, what transition training is needed for staff
Have the pilot group select the desktop environment, change it if the choice generates too many tickets
Take that and roll it out more broadly, possibly aligned with new desktop hardware rollouts
Add to the good - you get to know that the US government couldn't lean on a single American company for access to your organisation's secrets
I hear it didn't go well in the German government, something about the cost of training and skyrocketing tech support calls for basic tasks.
That sounds like a bad transition plan. For sure there's some lessons to learn from that experience.
They seem to be pushing for it still. Did you hear about that grant the German government gave to gnome?
One of them main reasons for that, I think, is how the average non-tech computer user perceives UI/UX, when they have been exposed to only a single type of interface for most of their lives (most probably Windows).
And even though they tend to pick up different UIs in mobile phones fairly quickly, that seems to not be the case for computers.
Back that up with earlier versions of middle-school computers studies in being mostly like:
- How to print a file in Microsoft Word?
- How to copy a file to USB drive? (with the implicit - using Explorer on Windows XP)
And you have most of the population thinking that's the only way to do it. That was the case with me until I learned programming.
The Germans also fell prey to Microsoft telling them that they would give them all the free copies of Windows they might need and build a new facility providing a ton of jobs in their area if they would abandon the Linux thing.
The city in question also built their own distro based on an older version of an existing distro rather than going from off the shelf.
Less than you think. Existing staff needs to be dragged kicking and screaming to learn the new systems.
Increasing the size of the helpdesk due to the increased call volume, more experienced non helldesk IT staff to babysit data migration and legacy systems. Now you have the administrative burden of all those new staff members.
Thats just the bad transitional phase, I think op means longterm
Still not as much as you think.
Let's assume they have M365 E5 at $57/m/user. A small government is several hundred people let's use 300.
300*€57*12 is a yearly cost of ~34KE5 license includes
Office 365. That can be replaced with Open/Libre Office at minimal cost.
Teams unified communications suite. You would have to go Slack/Zoom combo to get the same capabilities at a monthly cost per user for each.
SharePoint/OneDrive. Not sure of Linux versions.
Email with anti spam filtering. Postfix with MTA that filters maybe.That is just off the top of head.
Windows to KDE is a smaller change than major windows version changes. Pre-ribbon office to newer office
Reminds me of the US swapping to the metric system.
Short to mid term would be miserable and confusing for people. Long term would probably work out better. Will it happen: never.
Short term: no breathing
Long term: improved breathingYou can at least see why that’s a bad plan right?
Just hold your breath silly
The issue with Linux is getting middle management to support it. I'm my experience is based on them laying you off and hiring somebody else. Linux is great but management needs support contracts.
And thus RedHat was born
What's the cost to rewrite all of the existing software to a Linux version?
Zero if it can run on wine
Govts and large companies moving to Linux isn't about costs, security or studies. It is about plain simple corruption.
- Govts/companies like blame someone when things go wrong, if they chose open-source there's isn't someone to sue then;
- Buying proprietary stuff means you're outsourcing the risks of such product;
- Corruption pushes for proprietary: they might be buying software that is made by someone that is close to the CTO, CEO or other decision marker in the company, an old friend, family or straight under the table corruption;
- Most non-tech companies use services from consulting companies in order to get their software developed / running. Consulting companies often fall under the last point that besides that they have have large incentives from companies like Microsoft to push their proprietary services. For eg. Microsoft will easily provide all of a consulting companies employees with free Azure services, Office and other discounts if they enter in an exclusivity agreement to sell their tech stack. To make things worse consulting companies live of cheap developers (like interns) and Microsoft and their platform makes things easier for anyone to code and deploy;
- Microsoft provider a cohesive ecosystem of products that integrate really well with each other and usually don't require much effort to get things going - open-source however, usually requires custom development and a ton of work to work out the "sharp angles" between multiple solutions that aren't related and might not be easily compatible with each other;
- Open-source requires a level of expertise that more than half of the developers and IT professionals simply don't have. This aspect reinforces the last point even more. Senior open-source experts are more expensive than simply buying proprietary solutions;
- If we consider the price of a senior open-source expert + software costs (usually free) the cost of open-source is considerable lower than the cost of cheap developers + proprietary solutions, however consider we are talking about companies. Companies will always prefer to hire more less expensive and less proficient people because that means they're easier to replace and you'll pay less taxes;
- Companies will prefer to hire services from other companies instead of employees thus making proprietary vendors more compelling. This happens because from an accounting / investors perspective employees are bad and subscriptions are cool (less taxes, no responsibilities etc);
- The companies who build proprietary solutions work really hard to get vendors to sell their software, they provide commissions, support and the promises that if anything goes wrong they'll be there. This increases the number of proprietary-only vendors which reinforces everything above. If you're starting to sell software or networking services there's little incentive for you to go pure "open-source". With less companies, less visibility, less professionals (and more expensive), less margins and less positive market image, less customers and lesser profits.
Unfortunately things are really poised and rigged against open-source solutions and anyone who tries to push for them. The "experts" who work in consulting companies are part of this as they usually don't even know how to do things without the property solutions. Let me give you an example, once I had to work with E&Y, one of those big consulting companies, and I realized some awkward things while having conversations with both low level employees and partners / middle management, they weren't aware that there are alternatives most of the time. A manager of a digital transformation and cloud solutions team that started his career E&Y, wasn't aware that there was open-source alternatives to Google Workplace and Microsoft 365 for e-mail. I probed a TON around that and the guy, a software engineer with an university degree, didn't even know that was Postfix was and the history of email.
Holy hell, a lot of what you just described hit right home with me.
I started off as one of the cheap developers ("technical consultant") for one of those Microsoft business products. Almost every single one of our customers are already ingrained into Microsoft ecosystems and setting up the system we customize and sell is mostly a matter of integrating into their existing AD, Exchange Mail Server and sometimes their private cloud. I was pretty ignorant of open source tools that would tremendously help even if you're mostly using Microsoft. Ignorant might not be the right word. It would be more correct to say "afraid to peek out of the comfortable Microsoft bubble". It wasn't just me, a lot of propriety consultants don't really bother with anything else. If something's beyond our capabilities we can always get the support of Microsoft, supposedly. This chain of responsibility give end customers assurance somehow. Like you said, assurance on who to blame and sue at least.
Took me a while to break out of Microsoft bubble and now I do open source ERP. I do get by okay, but I think it's mostly because my country cannot afford Microsoft license fees.
Microsoft will easily provide all of a consulting companies employees with free Azure services, Office and other discounts if they enter in an exclusivity agreement to sell their tech stack.
I never thought about that. my job we use some software made by oracle. I have wondered how much it would cost to make a linux version or wine compatible
I watched the process where a set of Unix machines we had were up for replacement. The first version of the request had as the preferred option IBM hardware and Linux, the second version - after it had been to the executives - had the preferred option as IBM hardware and AIX
Like the exec knew what either were, but they would have had a consultant check the proposal. Then they pretend to have respect for the professionals they employ
The city administration of Munich switched to Linux, migrated all data and users, trained them etc. for millions of Euros, and then eventually switched back some years later since staff productivity was way down, and users didn't feel comfortable in the OS environment.
You can't enforce a change. Linux is great, especially so for tech enthusiasts, but the average (or probably below average) user might have a hard time to adjust.
And when performance is measured in workforce efficiency, then you have to accept that it's simply not suited for every environment.
They actually flip flop a lot.
2006: Migration to LiMux begins
2008: 1200 out of 14,000 have migrated to the LiMux environment
2013: Over 15,000 LiMux PC-workstations (of about 18,000 workstations)
2016: Microsoft moves german HQ to Münich
2017: Dumping Linux https://www.linuxinsider.com/story/munich-city-government-to-dump-linux-desktop-84307.html
2020: Going back to Linux https://www.zdnet.com/article/linux-not-windows-why-munich-is-shifting-back-from-microsoft-to-open-source-again/
2023: Microsoft opens new Experience Center in Münich https://www.munich-business.eu/meldungen/neues-microsoft-experience-center-emea.html
2023: Analysing what needs to be done to switch to Win10 before new vote https://www.tweaksforgeeks.com/ditching-linux-for-windows-after-wannacry-is-too-risky-for-munich-green-party-warns/The switch back to Windows was not because of vad productivity. They switched back when the new Major got a visit from the microsoft CEO.
https://www.sueddeutsche.de/muenchen/muenchen-microsoft-linux-verwaltung-1.5562006?reduced=true
The problem with that is a few years is a bit short to get real benefits out of it. And the Wikipedia article contradicts the statement that productivity went down. Actually issues and errors went down, half the workforce was alright with it and they saved tens of millions of Euros. And then they cancelled it. That decision wasn't backed by technical or factual reasons at all. Many people said they were fine with Linux. Issues were for example that they had old and outdated computers. As a reason to switch back they claimed: sync to mobile phones had issues (...as if government workforce syncs their calendar to their mobile phones...) and these were issues with the groupware suite. Nothing had anything to do with Linux, productivity or the people who sat in front of the computers and actually had to use it. There were quite some benefits and from the technical side things were going well despite admins not being backed by their superiors and the city. They did a final study which contains quite some / mostly dubious statements, and Microsoft also was involved in the switching back.
You CAN enforce a change. Sure, change is always hard in the beginning. But we do it all the time. The story of LiMux is more: You can destroy anything if you really want to. And politics likes to twist things so it suites their narrative. (And lobbyism is a thing and Microsoft is better at it than the Linux community.)
Sounds like very poor management since everytime a business company switches system infrastructure, the end-users will receive courses. I was working in a factory which changed the automation system and every end-user spent 4+4 hrs in the lecture room and after 1 month of use they had again 4 hrs advanced use cases lecture.
After just 6 months every worker said the new system is easier and better, which first seemed to be impossible transfer.
I agree. And for most end users they are just clicking buttons or accessing web based applications where the OS doesn't play into "needing to switch"
Interesting!
From my experience as an executive I recon they haven‘t factored in the side effects like vendor lock-in, customizability and application speed of changes. Those are pretty hefty sums over years/decades.
How did they measure productivity of a city administration? !?
Hehe, you think the words 'administration' and 'productivity' can be used in the same sentence?
When you don't know something, it's completely ok to not say anything.
Doesn't Germany use a lot of Linux in their government?
Some places. Munich tried to migrate but went back.
The price of the software is a tiiiiiny component of TCO. Support, compatibility, training, availability of companies etc forms a much bigger thing.
And of course the new Microsoft hq in munich that they build in 2017 after Munich announced to transition back to Windows.
Watch the movie "zeitgeist moving forward" and learn why they don't and why life sucks.
I'm sure I'm going to get made fun of for this post.
For anyone who didn't see “zeitgeist moving forward” yet, just don't.