The real risk to MS here is when people take the plunge and see how much nicer a (well supported, well put together) Linux distro is, compared to Windows etc! Will they want to buy their next laptop at home with MSFT installed?
Plus, if countries moving off Microsoft put a fraction of the savings into libre office product development, it might get an even more lopsided comparison!
Tbh the only reason I can't put an ubuntu for my parents/grandparents is Microsoft office.
Libre is generally ok, but they deal with lots of documents from different ppl/orgs that just need to have similar rendering. Windows is so bloated it's insane
Yes, because when they go to Media Market and FNAC they will only have macOS and Windows laptops to chose from.
Having Tuxedo, System76 and co is great, but what most people care about is what there are in regular consumer shops.
Linux adoption among consumer hardware has only been a success where it is invisible to the user, like the Linux kernel on Android or WebOS, firmware on DVD/BlueRay players,...
> how much nicer a (well supported, well put together) Linux distro is, compared to Windows
As someone who regularly uses both, and even in spite of the ad avalanche that (a bog-standard home install of) Windows has become: [X] to doubt. I regularly have issues on my corporate Linux laptop that I suspect wouldn't ever happen on Windows. Fleet management is easiest on Windows, with Active Directory, Group Policy, images, etc. Maybe package managers make software management easier on Linux, but fleet sysadmin is basically built-in on Windows.
Linux is superb for servers—I use Proxmox VE as the hypervisor on my homelab—but it still has a rather long way to go as a sharp-edge-free desktop OS, regardless of whether it is for home or corporate use.
Not sure about goverment requirements, but many apps I use nowadays are on the web, Figma, Canva, google docs, outlook and so on. There is not much need for a dedicated OS for most users.
the browser is definitely becoming more OS-like. that’s the Chromebook bet. but for proper office work you need good file management. if they’re not keen on trusting cloudy solutions by the big corps then FOSS on bare metal is the way to go.
I’m always surprised to see how naive people are when comparing the enterprise/government usage of software vs. Personal usage.
Linux and open source software are a good alternative for personal usage, but for government usage, wouldn’t they have to port every custom software they have built over decades to Linux? Not to mention driver support for legacy/custom hardware. Also, from where would they get the support and guarantee that critical bugs and issues are patched and distributed without disruption to every single computer in time?
Replacing MS office would be the easiest and what’s the point of migrating away from MS Office while still using Windows eco system? Is it just political symbolic migration because they don’t have anything better to do with tax payers money?
Their custom software probably runs fine under Wine? Also, if it's actual custom software, they might have the source and could even recompile it. And also, the software of the Swiss government is usually surprisingly good quality, so they probably have competent engineers who they could call upon to help with such porting.
Porting software nowadays might prove easier than expected. Exactly how much custom software should one expect? Centralized web applications running on Java are probably the standard anyway, and they are not OS-dependent.
The point of migration seems to be data-storage compliance anyway. They can later migrate off Windows easily once openDesk gets them going.
Custom software they have build over decades is pile of crap. Mostly could be replaced with web interface, gpg encryption, and batch processing of CSV files with bash or python.
It will fail because employees want "mouse point and click", "mouse point and double click" interfaces and are unable to use anything else.
Office is going to be a blocker - not just because of the two XML file formats, but because of the different UI. I personally find the Microsoft version more productive despite a mental list of complaints I have with it.
The correct choice, in my opinion, would be to invest in a new EU_Office project that uses the Microsoft XML formats. If you're trying to run your government on a new platform, that would also be a hedge against TDF pulling in a direction you don't want: an "independent self-governing meritocratic entity" has possible failure modes that need to go on the risk register.
Office is very much not a problem. LibreOffice has UI plugins that arrange everything in a standard Office pre-365 layout. I've been using it for years and I can't recall a single time I've ran into any issues. I think Excel macros are where the start of dealbreakers begin, which is a niche issue. 90+% of Office users won't hit that wall.
Second, the EU has been pouring money into an OpenOffice fork for official use and replacement of Office 365 for a while. https://office.eu/
I have been having some fun lately with some extra tokens from my work, experimenting with setting it loose on replicating software.
The results have been pretty surprisingly good. There’s a lot of functionality missing on first use, but I think if I let it churn through a few million more tokens it could get close.
I would be very concerned if I were a MS, SaaS or Adobe about what’s to come in the near future.
The biggest challenge with these kind of migrations is training users, especially government employees. Switzerland is relatively highly educated so this might be less of a challenge than in other countries but it won't be easy.
The technical part is probably relatively easy in comparison. Especially because the stakes are low (parallel run with the existing system, good will from everyone involved).
Educated or not, people that have been using the same system for many years will have a tough time IMHO. Looking at my mom and dad interactions with any UI, it seems to me they don't really read what that menu is about, they just use "muscle memory" to navigate it. So I expect a drop in efficiency until that memory is rebuilt.
> Switzerland is relatively highly educated so this might be less of a challenge than in other countries but it won't be easy.
True, but that's not accounting for a bunch of typically Swiss cultural factors:
- Swiss people tend to love to death the "one standard solution for one particular problem" approach
- Swiss people tend to be quite change-averse relative to other nearby cultures.
Don't get me wrong I love the Swiss to death, and the above cultural biases actually have their advantages, but re-training government employees is not going to be easy, however well educated they may be.
Fortunately, there is another cultural Swiss bias that's going to help: most large Swiss organizations have been designed using the army as a model, which mean Swiss employees tend to obey orders quite well.
So if the mandate comes from the top, it'll likely succeed. Eventually. Cause in Switzerland, they tend to take their time (yeah, that's another base cultural cliché, but I couldn't resist. And we're already deep into those already, why stop :-) ).
I'd guess the hurdles are always more structural and organizational - as even with Microsoft stack itself. See for instance the Limux/Munic story: https://lwn.net/Articles/737818/
Not really - this is akin to doing a complete rewrite of your 30 year old codebase to Rust.
Every organization has their own hacks, bespoke tools, processes and hardware that has to be painstakingly replaced, reprogrammed and users retrained. In many cases, Linux alternatives don't exist so they either have to be written or processes/hardware replaced.
These are OLD organizations and none of them is really the same as any other. Legendarily, companies like SAP made their whole fortunes managing this kind of complexity.
It's just a lot of work with a lot of edge cases you've never thought of (or seen in your life) that takes years.
The real risk to MS here is when people take the plunge and see how much nicer a (well supported, well put together) Linux distro is, compared to Windows etc! Will they want to buy their next laptop at home with MSFT installed?
Plus, if countries moving off Microsoft put a fraction of the savings into libre office product development, it might get an even more lopsided comparison!
Tbh the only reason I can't put an ubuntu for my parents/grandparents is Microsoft office.
Libre is generally ok, but they deal with lots of documents from different ppl/orgs that just need to have similar rendering. Windows is so bloated it's insane
The rendering won't be 1:1 but installing the most common MS fonts already does a lot
Yes, because when they go to Media Market and FNAC they will only have macOS and Windows laptops to chose from.
Having Tuxedo, System76 and co is great, but what most people care about is what there are in regular consumer shops.
Linux adoption among consumer hardware has only been a success where it is invisible to the user, like the Linux kernel on Android or WebOS, firmware on DVD/BlueRay players,...
Stores are dead to anyone who is under 35 years of age.
> how much nicer a (well supported, well put together) Linux distro is, compared to Windows
As someone who regularly uses both, and even in spite of the ad avalanche that (a bog-standard home install of) Windows has become: [X] to doubt. I regularly have issues on my corporate Linux laptop that I suspect wouldn't ever happen on Windows. Fleet management is easiest on Windows, with Active Directory, Group Policy, images, etc. Maybe package managers make software management easier on Linux, but fleet sysadmin is basically built-in on Windows.
Linux is superb for servers—I use Proxmox VE as the hypervisor on my homelab—but it still has a rather long way to go as a sharp-edge-free desktop OS, regardless of whether it is for home or corporate use.
I think the issues you're having on your "corporate linux laptop" are due to the "corporate" part.
I'm also having a lot of issues on my corporate linux laptop (such as dbus crashing) that just do not happen normally.
They are due to all the proprietary ~~spyware~~ security software they install, that hook pretty deep into the system and isn't as well tested.
Not sure about goverment requirements, but many apps I use nowadays are on the web, Figma, Canva, google docs, outlook and so on. There is not much need for a dedicated OS for most users.
the browser is definitely becoming more OS-like. that’s the Chromebook bet. but for proper office work you need good file management. if they’re not keen on trusting cloudy solutions by the big corps then FOSS on bare metal is the way to go.
SAP, Oracle, weird Office macros from decades ago …
They might know their own use case better than you and I?
I’m always surprised to see how naive people are when comparing the enterprise/government usage of software vs. Personal usage.
Linux and open source software are a good alternative for personal usage, but for government usage, wouldn’t they have to port every custom software they have built over decades to Linux? Not to mention driver support for legacy/custom hardware. Also, from where would they get the support and guarantee that critical bugs and issues are patched and distributed without disruption to every single computer in time?
Replacing MS office would be the easiest and what’s the point of migrating away from MS Office while still using Windows eco system? Is it just political symbolic migration because they don’t have anything better to do with tax payers money?
Their custom software probably runs fine under Wine? Also, if it's actual custom software, they might have the source and could even recompile it. And also, the software of the Swiss government is usually surprisingly good quality, so they probably have competent engineers who they could call upon to help with such porting.
The greater the dependence, the greater the risk.
Porting software nowadays might prove easier than expected. Exactly how much custom software should one expect? Centralized web applications running on Java are probably the standard anyway, and they are not OS-dependent.
The point of migration seems to be data-storage compliance anyway. They can later migrate off Windows easily once openDesk gets them going.
> but for government usage, wouldn’t they have to port every custom software they have built over decades to Linux?
Maybe not the case for everyone, but this has been the reason for making as much in-house software as possible behind a web frontend.
Custom software they have build over decades is pile of crap. Mostly could be replaced with web interface, gpg encryption, and batch processing of CSV files with bash or python.
It will fail because employees want "mouse point and click", "mouse point and double click" interfaces and are unable to use anything else.
Office is going to be a blocker - not just because of the two XML file formats, but because of the different UI. I personally find the Microsoft version more productive despite a mental list of complaints I have with it.
The correct choice, in my opinion, would be to invest in a new EU_Office project that uses the Microsoft XML formats. If you're trying to run your government on a new platform, that would also be a hedge against TDF pulling in a direction you don't want: an "independent self-governing meritocratic entity" has possible failure modes that need to go on the risk register.
Office is very much not a problem. LibreOffice has UI plugins that arrange everything in a standard Office pre-365 layout. I've been using it for years and I can't recall a single time I've ran into any issues. I think Excel macros are where the start of dealbreakers begin, which is a niche issue. 90+% of Office users won't hit that wall.
Second, the EU has been pouring money into an OpenOffice fork for official use and replacement of Office 365 for a while. https://office.eu/
[delayed]
I have been having some fun lately with some extra tokens from my work, experimenting with setting it loose on replicating software.
The results have been pretty surprisingly good. There’s a lot of functionality missing on first use, but I think if I let it churn through a few million more tokens it could get close.
I would be very concerned if I were a MS, SaaS or Adobe about what’s to come in the near future.
Do migrations like this get easier each time? With prior art being available, and an ecosystem and pipeline starts to form around the new norm.
The biggest challenge with these kind of migrations is training users, especially government employees. Switzerland is relatively highly educated so this might be less of a challenge than in other countries but it won't be easy.
The technical part is probably relatively easy in comparison. Especially because the stakes are low (parallel run with the existing system, good will from everyone involved).
Educated or not, people that have been using the same system for many years will have a tough time IMHO. Looking at my mom and dad interactions with any UI, it seems to me they don't really read what that menu is about, they just use "muscle memory" to navigate it. So I expect a drop in efficiency until that memory is rebuilt.
> Switzerland is relatively highly educated so this might be less of a challenge than in other countries but it won't be easy.
True, but that's not accounting for a bunch of typically Swiss cultural factors:
- Swiss people tend to love to death the "one standard solution for one particular problem" approach
- Swiss people tend to be quite change-averse relative to other nearby cultures.
Don't get me wrong I love the Swiss to death, and the above cultural biases actually have their advantages, but re-training government employees is not going to be easy, however well educated they may be.
Fortunately, there is another cultural Swiss bias that's going to help: most large Swiss organizations have been designed using the army as a model, which mean Swiss employees tend to obey orders quite well.
So if the mandate comes from the top, it'll likely succeed. Eventually. Cause in Switzerland, they tend to take their time (yeah, that's another base cultural cliché, but I couldn't resist. And we're already deep into those already, why stop :-) ).
With a couple of exceptions, my guess is that these migrations are part of the commercial negotiation pack of government procurement agencies.
At a certain point I’d expect that tech sovereignty concerns will tip while governments out of the US-vendor mindset.
As exemplified by a major project recently started by the French government:
https://www.rfi.fr/en/france/20260417-france-to-remove-windo...
https://www.techradar.com/computing/windows/france-has-ditch...
I'd guess the hurdles are always more structural and organizational - as even with Microsoft stack itself. See for instance the Limux/Munic story: https://lwn.net/Articles/737818/
Not really - this is akin to doing a complete rewrite of your 30 year old codebase to Rust.
Every organization has their own hacks, bespoke tools, processes and hardware that has to be painstakingly replaced, reprogrammed and users retrained. In many cases, Linux alternatives don't exist so they either have to be written or processes/hardware replaced.
These are OLD organizations and none of them is really the same as any other. Legendarily, companies like SAP made their whole fortunes managing this kind of complexity.
It's just a lot of work with a lot of edge cases you've never thought of (or seen in your life) that takes years.