South Korean government to switch to Linux: ministry
koreaherald.com
koreaherald.com
I work in a Danish municipality with 100.000 citizens. We operate more than 300 IT systems, some of these are one of a kind systems that have no alternatives and some of those only run on Windows.
We also have 8000 employees, a lot of them couldn’t tell you if the smartphone they use each day is an android or an iPhone. They know windows and they know office, switching them to pages and macOS would be almost impossible, Linux and libre office is out of the question.
We have an IT staff of skilled people who were willing to skip pay for ideology they are all certified Microsoft and have little other education. They could all easily leave for other positions in the Microsoft infrastructure. It would cost around a million to replace just one of them.
We could easily switch our tech to Linux, we would need to build a lot of replacement software but that could be done in a reasonably manner, but moving the business side to Linux would be the most expensive thing we ever did.
Which is why I’m a skeptic of something I believe should happen. We pay a tech tax to Microsoft, by we can’t stop. It’s just not doable.
Specialist software is a big problem of course, likely leading to a parallel Windows infrastructure of some kind, either dedicated Windows workstations for the workers that need it or a terminal-server setup of some kind - I know companies that used the latter to keep Windows XP applications available in an isolated environment too. But it's all a bespoke process.
That doesn’t really make sense though, because our IT staff can’t support it, so we would need 1-2 guys extra which is more expensive than the windows licenses for 1200 employees.
There's no doubt there'll be increased support costs in the short term, but this should be a long-term investment to get off of proprietary software and embrace open formats, (in my humble opinion, governments should by law be required not to produce documents that require proprietary software to open properly).
Governments should be the ideal vehicles for long-term investments. It's usually public corporations that are mostly concerned with the short term, it shouldn't be governments. If the SK government would seriously commit, in 10 years Linux would be the new normal and it would be Windows that requires extra staff, support etc.
We simply have too much software that only runs on; and too much staff schooled for Windows. You’re right that we could change if we wanted to, but it’s just hard to justify such an investment in an area where it would mean that we had to fire teachers and nurses to pay for it.
We don’t even have a good business case. I mean, our total Microsoft licenses are around the same expense as 4 employees. It would be literal decades before we brought the investment home if we made a full switch.
That doesn’t really work in political organisations, and it’s likely why all the previous attempts to switch have failed. To do this, we’d need massive public support for it, so the political leadership doesn’t get cold feet during the “short” term investment, and the truth is, the public couldn’t care less.
That doesn’t mean we can’t get there eventually. We try to buy more and more open source, we set up co-operations across municipalities to define architectural standards so suppliers can’t define our infrastructure anymore. But it’ll take a long time, and maybe it’ll never make sense to leave Microsoft. I mean, what’s the open source alternative to office365?
That's nice to hear. I think your best bet is indeed to slowly make sure you're not dependent on Microsoft. Even if Windows never wholly goes, at least Linux could then be a viable option for new deployments, or for smaller departments.
As someone who supports both Windows and Linux laptops, if you switched to Linux your support staff would eventually halve. The control you have over Linux is so tight, its relative immunity viruses, to staff playing with it and adverse Microsoft "upgrades" so good that support literally drops to whatever it takes to keep the hardware going. Or to put it another way: if you need to re-image a Linux laptop, it's because the disk drive broke.
Mind you, it takes a while to get to that point. And you need someone with real Linux expertise. Real Linux expertise unfortunately means when a problem arises gets to the point you would give up and re-image Windows, instead with Linux you haul out the source, and find the root cause. But if you can do that you will be able to drive the software defect rate to near zero.
Or to put it another way again: Windows licences are the price to pay if you don't / can't pay for that sort of expertise, so you are trapped into paying to do it for you. Unfortunately Microsoft is a monopoly, so you end up paying a fair bit more. I'll grant you small enterprises don't have much choice. But I'd expect someone with 1200 employees does have the choice.
I would love to see some countries adopting Linux for government systems and funding research and development e.g. maybe fund Libre Office or the KDE one more and then build out other tools to be cross platform.
For example, IIRC, they use a smart card to authenticate on their workstations, and also on their web applications. For that you need to customize a bit (better out of the box support for pkcs11 devices, integrate the Certificate Authorities needed, drivers).
There are a few contributions to open source projects:
A few years back, they were contributing to a fork of Thunderbird called Trustedbird for example.
There are also a few individual contributors inside the Gendarmerie, for example Xavier Guimard creator and one of the current developers of LemonLdap (SSO portal).
Also just as a side note, this distribution is used by the Gendarmerie, which is not exactly the national Police. In France you have both a Police Nationale and the Gendarmerie National. The Gendarmerie is military personel (the Police is civilian). I'm guessing national sovereignty plays a small role in that (VS using MS Windows, which is completely controlled by a US company, bounded to the US government).
When you run Linux you do it by creating custom packages and changing the package manager to point to a different server, at which point you have created a "custom distro".
I don't know about that, I've seen what kinds of systems the existing distros build and I'm not impressed. Which isn't to say that Microsoft would definitely better, seeing as they are doing their best to become a Linux distro themselves while embracing all the bad behaviors that entails, but your statement certainly isn't a given.
What do you mean by this?
Run Linux on a Desktop and it's a whole different story. Shit breaks all the time, the package dependency spaghetti is insane, no one seems to test anything (or they don't listen to the people who do), etc.
Technically interesting, though.
There's a reason Microsoft was successful for years: they built turn-key solutions, and added new turn-key features their customers requested.
Switching from that sort of IT to a linux-based, 'we need a system architect' model doesn't happen overnight. There's organization shift, cultural shift, firing unneeded people who won't retrain, retraining people who are worth it, and hiring talent to fill gaps.
Not to mention that talent's more expensive and less easily certified (e.g. all the official MS rubber stamp programs).
Not impossible, but it's a LOT of work. With dubious chances of success.
[edit: TOC -> TCO]
If I have a 50% chance to cut my IT budget by 15%, if I remake my org, is it worth it?
Honestly, because most customers sponsoring projects were morons. If you told them there was a 25% chance of failure, they'd go with someone else who claimed a 0% chance of failure.
I finally decided most of this was just sponsors covering their ass. If they initiated a project with risk and it failed, that reflected on them. If they initiated a project and the consultants claimed (winkwinknudgenudge) there's no risk, then if it failed those lying consultants were the only ones to blame.
A big difference compared to one or two decades ago is that a lot of domain-specific services have moved to web applications. So, there are probably a lot of machines out there that are just glorified web browsing machines.
Also, the popularity of iOS and Android have made more applications cross-platform.
Of course, once Office documents or legacy Win32 applications are part of the equation, you can't really beat Windows. It is a common misconception that LibreOffice could replace Office. I use a Linux desktop and a MacBook and most of the Word/Excel paperwork that I have to do (luckily irregularly) does not open correctly in LibreOffice. Of course, it is an uphill battle for the LO folks.
By that criteria (which are real criteria from real people), the only viable replacement for Microsoft Office, is the next version of Microsoft Office.
The only way to break out of this, I think, is for governments to forbid themselves to use proprietary software that was not developed on their own soil. Because let's face it, it's not really about money (especially not about short term money). It's more about national security, sovereignty, and freedom.
Dealing with this at work now.
And it's not like they're disinterested or incapable of learning, literally no one thought "Gee, what those folks are doing looks a lot like code."
The biggest boost to interop would be requiring software vendors to release spec sheets for necessary software components and file formats... if they want government work. No more "It's too hard" or "We can't." You want the job, those are the rules.
So, business teams invented the all-too-common version of agile development even before “real” software developers, without the disadvantage of ignorance of the body of knowledge around the SDLC, got around to it.
For that use case, there's always SoftMaker Office[1].
By that criteria, the only viable replacement for Microsoft Office is the _same_ version of Microsoft Office.
The delta between two Microsoft Office versions has often been larger than between a Microsoft Office version and a LibreOffice version...
Walk a mile in our shoes. You understand truths about your job we may not.
Perceived change generates just as much stress and associated opposition as actual change. You've just enabled me to make that explicit for myself. I feel stupid that took me so many years. Thank you for that.
Writing this out somewhat softens my frustration with people refusing to use anything but the blue e or similar intellectual laziness.
Not everything is an adversarial thing about two opposing groups of people.
Citation needed.
The company I work for switched wholesale away from Microsoft. Management moved first and the rest of us shortly afterwards. Haven't heard any complaints.
The company you work for is one of the enlightened few. Or maybe it's a tech company with an enlightened management?
It is. We are a tech/consulting company, but management is management and we also have a good number of employees older than me (and I think that is great :-)
Income inequality is a concern, and some people are certainly overpaid. But at the C* or technical architect level, the cost of pinching pennies is dominated by the cost of choosing a wrong direction (or not recognizing a changing situation and correcting quickly enough).
It's true they depend super heavily on ActiveX and other weird applets/software that runs only on Windows and/or IE.
A lot of the communication with the government (yearly tax submission) as well as banking software would not run on our Linux or OSX devices, and so we had these 'loaner' windows 'tax machines' that got reimaged every few days that people could use to do their business.
They also have their own Office-like suite by Hancon[1], that is a pain to support and only runs on Windows. There were attempts to have gDrive to be able to read the files but afaik that never worked out.
Hancom did release a Linux version of their office suite a few years ago, and regularly updates the OSX version. Moreover, Hancom has been dabbling in Linux for over 20 years. They even maintained their own Red Hat-based distro at some point. They also support OpenDocument and OOXML fairly well. In short, Hancom has been hedging their bets very thoroughly. So if the government adopts Linux by any chance, it won't be difficult for them to go fully cross-platform and keep their lucrative government contracts.
Hancom's Hangul software was popular in Korea in the 90s because of its splendid Hangul support (Word wasn't that good at it back in the day). So it became a de-facto "standard" for Korean documents. But the problem was, piracy was really rampant at the time, because it was before online DRM protection technologies came into light. (They've tried hardware DRM, like the serial port dongles, which eventually were cracked.) They almost went bankrupt, up to the point where Microsoft was thinking about buying them out (there are a lot of conflicting info about this in the internet, so I don't know exactly what deal might have gone through)
But then, when Koreans learned about this imminent danger, they launched a national campaign to save the company, and raised over $10 million for them. As for why this was possible: nationalism was pretty strong back in the late 90s, partly because of Korea's IMF crisis. (If you want to read more about the Korean economic recession and how people responded to it with activism, read more about the Gold-collecting campaign in 1998 [0]) Because of this, Hancom was able to recover and still be alive to this day.
Nowadays more and more Koreans are using Word these days, but Hancom's software still has lots of uses, especially in the government. The Hangul word processor had some unique characteristics other than just language support: most notably it was one of the first word processors to have nested table support. Because of this, many Hangul documents have layouts made of nested tables (I know some people at HN would be utterly disgusted with this, but this has many benefits: you can have total control of your document layout, and it's good for making documents with dense information such as government forms.) This is why Hangul documents cannot be simply converted to Word; the two have very different conventions about layout in general.
To conclude, reflecting this as a Korean (and an avid Linux user), I have both good and bad feelings about this. On one hand, Koreans were forced to use a proprietary, non-free program to open their documents, including important government stuff. This also sucks for Mac and Linux users (The Mac version was horrible the last time I've used it, and I couldn't get the Linux version to install on Arch Linux). On the other hand, if Hancom wasn't alive, then Koreans would be forced to use Microsoft Word with subpar Korean support (because Microsoft wouldn't really care about Korean language support unless there was competition). While I have mixed feelings about Hancom, I really hope that they would provide good support for Linux users in Korea.
You can tell that Word was never designed with CJK typography in mind when you write a sentence in English, try to add a word in Korean, and watch in horror as the entire line (and sometimes the entire page!) jiggles up and down while Word tries to figure out the line spacing (leading). By default, Word doesn't know how to keep leading consistent when you mix CJK and Latin characters in the same line, resulting in a very unprofessional-looking document. Hangul gets it right the first time.
On the other hand, Word blows Hangul out of the water when it comes to Latin typography.
https://en.wikipedia.org/wiki/Export_of_cryptography_from_th...
So depending on which version of Windows you are talking about you might both be right.
I think the language packs might not have been available in all SKUs; not very sure about that part. The CJK support, I think, was.
The default system locale thing (MBCS) was a separate issue.
It still is common. They just replaced ActiveX plugins with native Windows apps because ActiveX has a bad rep. It's still Windows through and through, and most of those "security" apps are borderline spyware.
I have to use Korean banks and government websites daily for business, and I keep them safely segregated inside a VM.
They stoped using ActiveX in the 00s but then started requiring the user to install some rootkit-like security software that was nearly impossible to uninstall and was probably able to spy on users.
Now they some of them have an app that is basically a bundled browser accessing their regular website.
The crypto nerds (including me) might think this is a splendid idea--perfect security!
The downside: they did this to shift the blame to the consumer if something goes wrong. If there is a credit card fraud, the consumer is automatically at fault by default, since obviously the consumer mishandled the public key crypto file. This makes sense, but this also means no chargebacks and other nasty side effects.
Also, this also meant that you had to install a bunch of ActiveX verging-on-malware-ware to get your banking shit settled. Now it supports Chrome and Firefox but holy crap, they install a ton of weird malware protection software and I'm pretty sure some of them take a cut from your computer's performance and stability. I'm just glad they don't display fucking ads.
I would believe it much more if they said that they will try to use more Linux, try to mix them up, or create APIs for everything etc. Nowadays it's so easy to create services and use LDAP or whatever thing you want with whatever programming language or service. You can host it on Linux and start changing some computers and systems to Linux.
But "switch from X to Y" seems a terrible idea. Even as a personal computer user, for me, it's impossible to "switch". Imagine for a government. It's a hilarious statement, one of those that politicians do say, but they have no fucking clue of what they are saying.
uhh... no, here in France a lot of public infrastructure runs on linux.
Apart from the obvious benefit of lower cost, security and personal individualization are further benefits. No matter what the politicians or business will tell you: Any government running Windows on their systems is basically broadcasting their data straight into US secret services. There's nothing stopping them from doing that. Who would prosecute them? Would US law enforcement have motivation to do that?
And when it comes to individualization: Any open standard that allows any other third party to develop products with will (if left to its own means) always win out against proprietary solutions, since you can pitch all the third parties against each other to compete for the lowest bid.
But since it's not about getting cost-effective IT, the governments don't appear too concerned of the US knowing all their secrets and Microsoft is definitely more capable of pampering politicians than 3rd party Linux developers could, absurd requirements dot every government contract to bar said Linux developers and leave Microsoft as the only "viable" choice.
[0]: https://www.libreoffice.org/discover/who-uses-libreoffice/
What Munich did was essentially created a homegrown solution which suffers from all the problems of homegrown solutions. Yes, you get to tailor it to your business, but then you have to maintain it for decades. Inevitably industry passes you by, and you end up paying for a sub-standard proprietary (ironically) solution that has no widespread support and you're own your own for everything - from new features, to security updates, to training, to compatibility and interoperability. It's always great at the beginning, and it's always miserable 10 years later when the excitement goes away and the original champions have moved on.
In other words, the same situation that exists everywhere.
Not exactly. When the decision to go back to Windows was announced, they were even still running Windows 2000 on some machines. [1]
The initial reason for overhauling their infrastructure was certainly due to the proliferation of outdated software, but the consultancy they hired to evaluate the situation made the more conservative recommendation to update everything to the latest version and let each department choose which operating system to use. (Since they were de facto already doing that, just without a reasonable process.)
That this turned into "100 % Windows, no compromises" was probably due to Microsoft politics, i.e. someone convinced the mayor that Linux was the cause of their infrastructure issues rather than bad infrastructure management.
I bet they still have the same problems, except now they also have a few machines running outdated versions of their homegrown LiMux distribution.
There are options such as PolarisOffice and Naver Office, but given that Hancom Office has a stake in the game, their cloud service HancomSpace along with Hancom Online may gain momentum. In the worst case scenario, I could see Korean government see Office 365 as a fallback.
https://www.zdnet.co.kr/view/?no=20190516093822 https://namu.wiki/w/%EA%B5%AC%EB%A6%84%20OS https://space.malangmalang.com/ https://www.hancom.com/product/productNetfficeMain.do
Add to that, (at least here in Germany) BK is a franchise that sells licenses to franchise operators, so I doubt they have strict requirements on the actual IT setup, just the pictures they need to display.
So you have very normal, run-off-the-mill managers and IT services, the former unable to grasp that Windows 10 with constant updates and fuck-ups will constantly require the services of the IT guys, and the latter naturally profiting of that. The increased margin of not using Windows wouldn't be enough to compensate that, and an increased price for a Linux package is something very difficult to convince those managers of.
And then also IT-people well versed in Linux usually have clients who pay substantially more than simply outfitting non-IT businesses with IT-equipment.
* Windows has been approved by management to be the only OS allowed on the computer network * The software for showing the static images / videos and whatever manages it will have been written on Windows or even DOS * The manufacturer of said machines whoever wrote their software was Windows Developer and it is cheaper just to ship a copy of Windows Embedded then to rewrite the software , QA and integrate it.
I have a piece of software that I get small support contracts for. I tell clients it is Windows only. There is nothing stopping me from deploying to Linux or even one of the BSDs (I think). I just can't be bothered to deal with differences in distros, going through a QA process etc when it won't really get me many more sales.
The companies that make that stuff started out with Windows, they have built a workflow to manage these devices on top of Windows, why switch now? It's not cost effective.
I'm pretty sure I've never seen anything like that on LKML.
Such dependence on Microsoft technology is not only found in government but also in corporations. Recently I needed to access my bank account there and the bank required me to install some anti-virus software that only runs on Windows.
I don't know how things are now in general, but I welcome this change.
Here's something more likely: Munich was ready to move back because the project was expensive to maintain, cumbersome in practice, killed productivity and was generally unpopular with the end-users, and Microsoft's new branch was just a cherry on top.
You might not like this line of argument, but it's not a "conspiracy theory" since it's well-supported. It's disingenuous to completely dismiss the role of political influence on a government decision.
https://www.techrepublic.com/article/linux-in-munich-no-comp...
But it isn't. You just asserted it. I don't dispute the facts at hand, but you can't just link those together and assert causality when there is no evidence for it .
>It's disingenuous to completely dismiss the role of political influence on a government decision.
I don't know what to do with that. There is no evidence and everyone disputes the corrupt link you assert. Not only that, there is an alternative explanation: the Linux deployment wasn't popular, killed productivity, and wasn't cheap to maintain. What you're doing is the definition of FUD.
I did not assert that Microsoft's move to Munich was the only reason for the city's change from LiMux to Windows. But, Microsoft did gain political influence in Munich by moving their headquarters there. Politics played a role in the city's decision, and there is sufficient press coverage to support this. See the TechRepublic article in my parent comment.
You're arguing that the deployment of LiMux itself was the one and only reason for Munich's return to Windows, and dismissing the role of politics in Munich's decision as a "conspiracy theory".
>Microsoft did gain political influence in Munich by moving their headquarters there.
And did it actually influence the decision of Munich to change their infrastructure? And how much? 10% or 90%? Let's say it was a minor consideration, on the order of 1%, and the rest was the result of the failure of the LiMux deployment itself - see what I mean about needing evidence. You can't just say things because of your FEELS.
Again, one thing we know, it wasn't a great deployment.
https://www.techrepublic.com/article/linux-in-munich-no-comp...
There are additional sources describing politics as a significant factor in the city's decision:
https://www.networkworld.com/article/3170104/munichs-great-l...
https://translate.google.com/translate?hl=en&sl=de&u=https:/...
https://translate.google.com/translate?sl=auto&tl=en&u=https...
Reiter based the decision to switch back from LiMux to Windows on a report commissioned from Accenture, "the leading Microsoft services provider for the ninth consecutive year" with an obvious conflict of interest:
https://newsroom.accenture.com/news/accenture-and-avanade-na...
https://www.zdnet.com/article/microsoft-partner-claims-munic...
I invite you to provide evidence for your assertion that the role of politics in Munich's decision is a "conspiracy theory" and "FUD".
EDIT: > role of politics
Role of politics is a somewhat different claim than just "pressure/influence by Microsoft" (which for sure happened, but it's hard to gauge how large the influence was), and includes lots of other factors that are likely to have also played a role. The entire thing is quite murky and I think "it happened because X" is too easy whatever the X is.
AGAIN, the BIG picture - this deployment didn't turn out to be the success Munich had hoped it was going to be. They could limp along, sure, and I'm sure there are plenty of people who like it, especially the guy who runs the entire project, and the people who are ideological inclined against Microsoft, and the Green party...
I mean Jesus, the only way they could get this LiMux thing to work was by maintaining Windows licenses for legacy software and interoperability, virtualizing Office because OpenOffice/LibreOffice sucks (and it does), and moving to Exchange for Calendar and Email. This is a chimera. It sucks.
There were good reason to ditch Windows in the first place, and now there are good reasons to ditch this Linux experiment as well. That's the big picture.
>I invite you to provide evidence for your assertion that the role of politics in Munich's decision is a "conspiracy theory" and "FUD".
I invite you to provide evidence that your bias has not coloured your interpretation of events. Have fun proving a negative.
(I read all the things you cited - it's the same kind of stuff - high on speculation, light on evidence).
Here is where you are misinterpreting all of the above. I did not say there were any "secret nefarious reasons". All involved parties made the decision that would be most beneficial to them ("business reasons" in the case of Microsoft), which influenced Munich's reversal.
Munich's switch to LiMux was an embarrassment to Microsoft, and one of the reasons for their move to Munich was to gain political clout by becoming a more important stakeholder to Munich. Generally speaking, companies have more influence on a city's decisions when they are actually located in the city, because they pay taxes to the city and have more employees living there. This isn't a "conspiracy theory", it's common sense.
In summary:
You: Munich switched back to Windows solely because there were issues with the LiMux deployment, and Windows software is superior to Linux software.
Me: Munich switched back to Windows because there were issues with the LiMux deployment, and because Microsoft made itself more important to Munich by moving its headquarters to the city.
We can agree to disagree.
And yes, Googlers & co., that includes you, even if the small lib you released is Open Source, the magic sauce powering Page Rank, Analytics, AdWords and such isn't.
https://www.theregister.co.uk/2017/11/13/munich_committee_sa...
> Hübner said the city has struggled with LiMux adoption. "Users were unhappy and software essential for the public sector is mostly only available for Windows," she said. > She estimated about half of the 800 or so total programs needed don't run on Linux and "many others need a lot of effort and workarounds". > Hübner added, "in the past 15 years, much of our efforts were put into becoming independent from Microsoft," including spending "a lot of money looking for workarounds" but "those efforts eventually failed."
https://www.neowin.net/news/munich-germany-realizes-that-dep...
> https://www.neowin.net/news/munich-germany-realizes-that-dep...
Looks like the open source equivalents for the software didn't deal well enough with different file formats.
"In 2018, journalistic group Investigate Europe released a video documentary via German public television network ARD, wherein it is claimed that the majority of city workers were satisfied with the operating system, with council members insinuating that the reversal was a personally motivated decision by lord mayor Dieter Reiter. Reiter denied that he had initiated the reversal in gratitude for Microsoft moving its German headquarters from Unterschleißheim back to Munich."
This is also worth a read:
https://blog.documentfoundation.org/blog/2017/02/14/statemen...
In short, it _was_ MSFT lobby grease.
I don't doubt that there were issues with 'different file formats', .docx and .xslx right up front (open standards that they are one wonders how there could have been an issue).
https://mspoweruser.com/microsoft-germany-moves-into-a-new-h...
Probably has a lot more to do with it than file formats ever did.
A lot of gov / HR departments etc have embedded forms in things like word documents and PDFs. The same problem with alternative PDF viewers, they don't do PDF forms well if at all.
You don't know the requirements of the different departments and saying "this free alternative works fine when writing a doc" doesn't cover their use cases.
The Linux community like to think that Microsoft is the "big evil" that is trying to thwart them at every turn. I know quite a few people in Microsoft and they tell me that the left hand doesn't know what the right hand is doing.
So sorry I think it is a conspiracy theory and as someone that earns his living processing documents from Office I know that incompatibilities exist between Libre Office (which is a fork of Open Office) and MS Office.
> The Linux community like to think that Microsoft is the "big evil"
I really don't think I speak for the Linux community but I do think that MS is still the big evil, just in slightly nicer packaging. Fantastic PR, crappy company.
No. Please don't put words into my mouth.
The problem is that there is no specs / documentation for anything generally. Most of my time in dev has been spent dealing and finding work around for all sorts shitty bugs in applications when dealing with either web browsers or generating documents.
This is partially because of deliberate vendor lock-in and partially due to nobody being given time to actually produce specs as these are often an after thought. Customers don't care about me making proper documentation, even when they do other developers that do integrations DON'T RTFM and I have to guide them through it anyway. Customers don't pay for specs / docs etc, they pay for the product that lets them get shit done.
Fanatics of Open Source (and they are that) always argue about cost of the software, vendor lock in etc. The costs of the software are almost nothing compared to hiring people. I was on a team of 6 people doing a major rewrite (which shouldn't have been happening IMHO) and it cost the business about £500,000 just in staff. Licenses for SQL Server and Windows server was a small percentage of that and were about 1 man months worth of cost. These costs are nothing for large councils / businesses that will either recoup that cost from the software itself or will just raise taxes to cover the shortfall (which is what Gov always does).
The really good reason to use Open Source software is security, privacy and being able to run what you want on what you have purchased, but that never gets mentioned, it always the bullshit cost angle because £400 sounds a lot to the kids in University which are broke.
Also vendor lock-in is the norm for a vast number of reasons other than "M$ is the evilz, stop using Windoz". Just look at the 8bit/16bit era, almost everything was lock-in and much more expensive than today (by factors of ten). The free PC ecosystem is NOT the norm in computing.
> I really don't think I speak for the Linux community but I do think that MS is still the big evil, just in slightly nicer packaging. Fantastic PR, crappy company.
LOL. Google, Facebook, Paypal, Twitter are far more powerful than Microsoft and are much much more evil, you know why? Because they think they are the good guys. Microsoft are basically like IBM these days, they sell loads of proprietary stuff that office use because they have won in that sector, they are dominate almost nowhere else.
A great deal of the internet is now hosted on a about 5 or 6 providers, the largest being Amazon's cloud. That is far more frightening to me than Microsoft. Every single competitor (except Microsoft) uses Amazon's Object Storage API because they are the largest.
This isn't the early 2000s where Microsoft were a monopoly they simply aren't anymore because the market has changed so drastically in 20 years.
Just throw a custom graphical shell over a hardened Unix os, then use Valve's proton project for application comparability. You could be so rich using other people's work Microsoft.
Can't do it, Google owns that business model. I mean, they managed to do that twice? Must have taken them some serious effort!
Apply Hanlon's Razor, and dont assume everyone has a grand plan.
OOTH, my W10 laptop (xps 13 9343) is forced by microsoft to update. This means that every week my lappy spends around 90 minutes downloading, installing, failing, and finally rolling back an update. If I can remember I just keep delaying it.
As have many colleges
What I was talking is the central govt and all govt undertakings. It'll be huge. Windows and IE6 is standard and they are bad, they hang a lot
Although the penetration is quite low. The project looked stalled few years back, seems like they are back again. Last release is dated 28.8.2018
Building a one time theme or anything of that sort is okay, I think. But it'll be costly to maintain a distro
Surely at the level of app/drivers or something, but probably not core functionality.
The whole point of the process was to save money, and one of the ways they can do that is by standardization.
Any deviation for de-facto standards can be costly in many ways, especially in terms of risk.
By 'pooling' I mean paying some smart people to do that. Like Rad Hat. Or --> Micrsoft! Or whoever.
Software is like anything else: it has to be made and maintained, it takes work, and someone has to pay for it.
The fundamentally different issue is 'platform monopolies' which we have to watch out for. But other than that it's just a product.
S. Korean gov. probably spends a zillion dollars making and maintaining roads. Do they design and build their own construction equipment? No way. Because it makes more sense for all of us if they pay some specialized people to do that, i.e. 'just buy it'.
More specifically to your point, it might make sense for a 'state version' of some linux distro, that really wasn't a functional variation, but more or less a 'bundle'. Maybe RedHat could do this ...
I'd bet 5,000 won that they won't even contribute their translation fixes (i.e. their only positive contribution to the world) upstream. They will also rather go out of business than to follow the LTS cycle and keep their distro up to date. It has happened before. It will happen again.
I'd say the main difference is that just an additional repository on existing system might break if that base distro updates, while you control that if you go entirely through your own repo.
Thanks for letting us know, couln't have guessed it.
The way Linux systems are constructed is by gluing together a giant pile of disparately developed software, usually with fragile scripts, and trying to pretend that it is a robust and integrated system. Consequently, they require huge amounts of effort to alter, which will all be undone or in conflict with the next update. So what do you do when you get a Linux system working the way you want? You make your own distro, because it's the only way to keep it like that.
So the working scenario is that a Korean company creates its own distribution, adapts it to the needs of government, sells it and then maintains it for the fee. This could also have a benefit of supporting a local vendor instead of sending public funds to US.
I wonder how different it would be were they using an OSS stack. Any stories?
[1] https://opensource.com/education/14/9/interview-charlie-reis...
[2] https://lancasteronline.com/opinion/editorials/penn-manor-s-...
I've heard something like this in the south-Indian state of Kerala as well.
Piracy just isn’t mostly worth it anymore.
(It's super common for information displays everywhere to run Windows)
You're doing it wrong. The government should have a robust hierarchy of support staff in place. While the transition will consume a significant amount of their time and therefore cost money, it's money that you were spending anyway, and was already budgeted. New PCs are not required for Linux, except in cases where their existing hardware is damaged or insufficient, which again should already be in the budget.
[1] https://www.theregister.co.uk/2017/11/13/munich_committee_sa...
Why does it suck? Is it an issue of compatibility or performance? At least WSL2 will fix the former.
It sucks because the Linux distros available for it are now outdated, it utilizes a specific outdated kernel, it is extremely slow.. the file system is about 20x slower doing most file-system related tasks. Copy/paste doesn't work well without XLaunch (and then stills sucks). There was never a good terminal for it, and the new one doesn't address if it will provide mouse support for things like Vim, it didn't support Docker.
There are many more nuances, but that is just a few. Let's not forget though that Windows package management is the worst out of all the operating systems, so even if you have WSL, you are still going to be suck maintaining a crappy OS using terrible bloated installers that leave files everywhere and a crappy package manager like Chocolatey that is falling further behind repos like AUR and Arch.
That's what apt-get is for.
> it utilizes a specific outdated kernel
Lots of distros do that.
> it is extremely slow.. the file system is about 20x slower doing most file-system related tasks
That's a windows problem, not a WSL problem.
> There was never a good terminal for it
Don't you have the choice of almost all existing windows and linux terminals? How can none of those be good?
And normal windows installers have nothing to do with it.
The lack of docker support is important, but let's not ignore that they're in the middle of fixing it.
Various Linux desktop distros?
If so, have mercy on anyone who will be using Ubuntu Desktop for the first time. I hope they aren't afraid of using Terminal, because the internet will shout them down for wanting a GUI solution.
Love Ubuntu server, but could never get into linux desktop.
Also you're not really supposed to read through the entire man page for an app that you only touch once or twice, you're supposed to just search for what you need and leave.
My personal opinion is that GUIs are often less user friendly, they just look friendlier. GUIs make it easier to sell software, not to use it. Free software isn't being sold and that's why it tends to not have a GUI.
I think it’s a mixture of two factors:
1. From the perspective of the writer, it’s much easier to describe instructions via text when publishing those instructions via a predominately text-based medium such as web sites. The alternative is taking screenshots and editing those images to highlight relevant widgets. This is more time consuming and not ever site offers an easy way to upload those images (meaning you’re then sometimes having to manually upload them) nor even embed them into the discussion (eg HN). And on top of that, you still need some text to accompany the images anyway. So you might as well go with a text driven solution. And if you’re having to describe a UI in text, it’s obviously going to be easier to do so with a text based UI than a graphical UI.
This is a problem I’ve personally experienced as someone who used to write a lot of documentation for graphical UIs. It can be annoying and remember that a lot of time, people providing guides are those volunteering their time. So naturally people will opt for a lazier approach.
2. The other major factor for why I think CLI instructions are so common even on desktop Linux is because there isn’t a single agreed standard desktop environment. Not even a standard GUI toolkit, let alone DE. Arguments about whether this choice / fragmentation is a net good or bad thing aside; it makes much more sense to offer a guide that is going to be more universal (as the CLI on Linux tends to be - notwithstanding some district-specific file system hierarchies). Where as a guide with images will not only be very specific to using that unique management tool but would also be subject to any changes to the UI of that tool even if the underlying configuration files haven’t changed. So with CLI examples you end up with a guide that is not only more universal but also has greater longevity. The longevity point is something that can frustrate less technically capable Windows users who are left confused each time Microsoft decide to redesign their Control Panel applets (et al) and Linux isn’t any less subject to UI changes in its graphical frontends either. So it does make some sense to provide instructions to configuring things directly in the backend - even if it seems counterintuitive to a desktop user.
Ultimately though, I don’t see anyone who uses Linux being rude about those who use GUIs. At worst they might recommend using the CLI if someone has tried configuring something from a GUI and found themselves getting stuck because options aren’t getting applied in a way that was being expected. However there are examples when the reverse is also true. For example on HN a few weeks ago some CLI users moaning about the complexity of kemu was repeatedly (and rightfully) recommended to use a frontend instead.
Every workplace is domain-specific, and domain specific software is simply a tool to get the job done. Windows/Linux is completely irrelevant and of absolutely zero consequence IMHO. You pick a tool that you know will get the job done, and simply use whatever OS it works on. We use a lot of scientific software at work and a 100% of it requires Windows. Yes, it could easily be re-written to work on Linux or OSX but that is not the case at this moment. If we were an accounting firm, we'd just pick Quicken or Quickbooks or whatever tool people were familiar with and use it. Or if we were an magazine or an architecture firm, or an electrical company, we'd again use whatever tools we need to get the job done. As far as the OS goes, the only thing people care about is whether their tools still work on version N+1.
For various reasons, Linux and OSX require much more involvement from the vendor to keep the software running. OSX is very user-friendly, but very hostile to developers who just want to push code out, and not be shackled with supporting the new flavor of the month API that Apple comes out with.
Windows requires very little time investment from the vendor to make it work on newer versions. Heck we're using Windows XP software (like.. binaries compiled decades ago) on Windows 10 without issue. Yes, obviously, giant caveat here - I'm not saying windows never has API breaking changes .. I'm only saying the probability is much much lower on Windows than any other platform. (discounting esoteric options that nobody is using)
People who are cheer-leading Linux or Windows or w/e, never seem to "get" this. Nobody really cares about the OS.
Sounds like your graphics card doesn't work well with Linux. Is it Nvidia ?
Which one? There's a crapton.
PS - "evil" is not a religious term, it's actually the substitute for a religious term, which is "wickedness" (see Widipedia.)
Windows, on the other hand, not only has special characters that cannot be used but also reserved words that cannot be used as file names (eg COM). Then there’s the issue with magic file extensions and even Windows Explorer hiding extensions entirely by default.
On balance I prefer Linux’s approach. Sure it creates some problems but they’re easy to be worked around and usually only rear their head when you start using lesser polished UIs or shared file systems with Windows. Plus Linux empowers you to use those names if needed, or change them if it’s a problem. Whereas Windows has these weird hidden show stoppers - many of which were only relevant in the days of DOS but have been carried forward in every iteration of Windows since (including NT) for compatibility reasons. That causes way more problems to developers - which is my personal use case.
https://docs.microsoft.com/en-us/windows/desktop/fileio/nami...
Worst scenario is you’re dealing with a non-technical team but in those instances you’d keep your storage centralised and use an NT based solution (or something with first party support for AD) so you catch any Windows related bugs at save time rather than further down the road when you need to access them again. Given file names in Windows is limited by both the file system and the OS, whereas Linux is limited only by the file system, all you really need to do is catch those Windows-related errors. So running an NT-based storage array would solve that problem.
It seems fine to me. I'm running it on my laptop right now, and I don't find it significantly more annoying than Windows or MacOS, both of which I've recently used on work machines.
I'm sure there are plenty of people for whom that's not true, especially people who need niche commercial software. But a great deal of what people use is on the web these days anyhow, so I think it matters a lot less what OS Chrome is running on top of.
they are still on Windows 7? ouch!!!
Installing is not straightforward though, you have to load drivers into the installation environment if you're using M.2 or USB 3.x.
IMHO the effort I need to learn that causes me far less frustration than having to use Windows 10. My total experience (helping colleagues/friends with their Win 10 systems) is probably less than 60 minutes in total, but one look at that start menu with actual internet-loaded ads (!) was enough to dissuade me from ever installing that OS on any of my systems. If they're greedy enough to blatantly do that, who knows what kinds of other dark patterns lurk within that OS...
Windows 10? Man I feel like I'm on an ad-infested website and I need to be careful of where I click. But I'm definitely not objective here.
Lolz, I heard that already... with Windows 2000. So many (me included) spent serious time trying to strip XP of all the crap, because "Win2000 was just perfect".
In reality, XP (like 2000) was just another pile of shit. They patched the hell out of it with SP2, and still it would get owned and owned.
Vista and Windows 7 have made using Windows much safer, and Win 7 refined the interface enough to be a substantial improvement over XP. Win 7 was probably the pinnacle of pre-touch/pre-cloud Microsoft.
The main difference so far is the shortcut used to switch keyboard layout, and enabling Wi-Fi sharing in 10 is very straightforward.
The software I use in both feels just the same.
If you think that's painful, wait until you find out about Windows XP[0]. :)
[0] - https://www.windowslatest.com/2018/04/04/windows-xp-is-still...
In a more modern OS, KoToR doesn't render grass even running in the same hardware.
I think a lot of people were fine with 7.
Doesn't really jibe with the moves they have been making lately that seem to show that if anything, Microsoft is de-prioritising the Windows business.
They are happy for you to move off Windows provided you still use some of their cloud offerings. There's a reason Office 365 is doing as well as it is.