They switch back to Windows in a time of total vendor lock-in in the office software space. In the time when they started their Linux project, one really could switch from Office 2003 to Staroffice/Openoffice. But now, in the days of cloud and absolute technical domination of Office 365, this is a bit sad to see.
Office 365 is affordable, well-supported, easy to use, and updated frequently. It does what businesses and governments want. Their formats are well-supported. The software is also very stable and very mature. Other than ideology, there are few reasons not to use Office 365.
I wasn't aware any cogent argument was being put forward. "Does what businesses and governments want" is so broad as to read like advertising copy. I mean this is about as blatant as it gets. Clearly the only reason to use open source software is because one is an idealogue... please.
> Clearly the only reason to use open source software is because one is an idealogue... please.
If you taken a few more seconds to read, that is not what your parent comment said when they wrote "Other than ideology, there are few reasons not to use Office 365."
It breaks the site guidelines to insinuate astroturfing or shillage unless you have specific evidence, so would you please not do that?
Such things no doubt exist, but internet users are overwhelmingly more likely to just invoke them ex nihilo when someone happens to post a view they don't like. This is toxic to discussion, so we don't allow it.
One reason not to use it is because it makes German government dependent on a foreign company. Another reason not to use it is because it's a cloud product and shouldn't be used to store private data about German citizens. Another reason not to use it is because the FOSS alternatives are free.
Well, they certainly have a number of products, but I fail to see how their giving away licences has anything to do with reducing those ongoing operational costs? Not only that then, "first one's free" certainly applies: I would hate to see the renewal costs they charge!
> I fail to see how their giving away licences has anything to do with reducing those ongoing operational costs? Not only that then, "first one's free" certainly applies: I would hate to see the renewal costs they charge!
Your parent comment is referring to the following model:
1. Microsoft provides free or nearly-free licenses to schools. They don't charge the schools for software now, and they won't charge the schools for software later, because that would defeat the purpose of giving the software away. Renewal fees are not an issue anyone sane would worry about.
2. Incoming government workers already know how to use Microsoft software, because that's what they used at school. They complain if they're supposed to use something else.
> Renewal fees are not an issue anyone sane would worry about
Right. The presence of that threat is enough to discourage looking for alternatives. "we hear you've been looking into Linux and teaching Python instead of C#/.Net next year... sure would be a shame..."
they apply to proprietary software. they most certainly do not apply equally. it's a lot easier (and therefore cheaper) to find staff to support windows than it is to find staff to support linux.
For market dominating software, those costs can be reduced by sites like stackoverflow, youtube, and blog posts/samples.
Try searching for how to perform an arbitrary function in Word or Excel. Hundreds of hits, and probably 25% will solve your problem. Try the same search for Google Sheets or OpenOffice. Crickets was always my experience.
Of course, but do notice that with open source software, if one has the motivation, it's all right there in black and white code. Doesn't work for you? change it. Don't know what's going on? It's because you haven't looked - you're not prevented from doing anything.
That's assuming you can't get in touch with the person who wrote it. Which you probably could quite easily do.
Yes; I've sent patches to open source projects before.
When I'm in "accomplish a task using a spreadsheet" or "write a document/prepare a presentation" mode, I'm probably not going to be inclined (nor have time) to go rathole on changing the source to the office application, recompiling, testing, doing my original task, and submitting a patch.
There was a time, many moons ago, when secretaries with zero IT knowledge learnt Wordstar key sequences and got very productive using the software; now keyboard usage is considered a highly technical skill among IT workers and rocket science among casual users.
It's just a market driven untraining; everyone can learn to understand and use FOSS, but it needs strong motivation.
You've got to take the context into account. At that time, the alternatives were things like typewriters that were even more labor-intensive to use. They were learning a tool that made them more productive, and therefore more valuable.
Nowadays, Microsoft's Office suite was, is, and, all indicators suggest, will continue to be a bigger productivity multiplier than any of its competitors. For people who primarily work in office software, choosing something like it is the sensible default for any company whose primary motive is profit. Sure, it costs money, but the cost of the software is tiny compared to a skilled office worker's salary - small enough that I'd estimate it wouldn't have to make someone even 1% more productive to justify the retail price, let alone what it costs if you qualify for any sort of volume discount.
There are other motivating factors that could come into play, sure. But if we're talking places that have to pay for their labor (i.e., not home use) and make heavy use of office software, going with the one with the lowest TCO is a sensible default.
True, completely different context, but I was referring to the common belief that using something non mainstream, or train someone to, is that hard. If we continue walking this path we soon will end up with programmers using only the easiest development system around, something we are already experiencing with hardware that gets faster every year but software doesn't. To me letting the market decide what tools one has to use is very bad in the long term.
Same as with MS Office. I remember me and my parents doing trainings in MS Office, which wasn't exactly free either.
> Securing the environment.
Same as with MS Office.
> compatibility with proprietary formats isn't 100%.
MS Office' compatibility with open formats isn't 100%.
> Maintaining the environment by addressing bugs.
That is a win for MS Office. Now offset that against the license cost and vendor lock-in risk.
I know we conclude the same, namely choosing FOSS for many kinds of environments, but your post sounds a lot like there are many costs unique to FOSS while in reality there are lots of costs involved with both options.
Those details are dealt with the contract between Msft and the Munich government. Microsoft, along with Amazon and Google, provide on-premise storage; so specified data never ends up leaving the system.
The home country of Microsoft still has big and active military bases in Germany and spies even on Germany's highest politicians. If they wanted they could pulverize Germany withhin minutes. You can't honestly say that the German government is really independent under these circumstances.
You are right. It sounds like that. I actually wanted to say more:
"Look, fellow Germans. We are called inudstry leader and powerhouse of Europe. We export lots of goods for money they owe us. So much that we can bankrupt other nations. Our politicians say that yes, we are a souvereign nation. At the same time we cannot protect our own borders. We give control to the EU and other nations. We let other countries fly attacks on countries we are not at war with from our own soil. We host weapons that could instantly pulverize our whole country. We are spied to up to the highest levels of government. Even down to the lowest levels we fail. We use software that we don't have the source to. Here is this little project of providing some sanity and some freedom. They use software that is actually open to look at and we are free to develop it to our needs. We can take these matters into our own hands and share our work with the world to create a part of a free and open society. Yet, we have kill this project. Not because it doesn't work. It works, just not as well as the foreign made closed software. It doesn't look as good. It is killed by some politicians who want to, yet again, give away some of our freedom to foreign corporations. They want to run the government like a company and they succeed as such. Being a company, not a government. Let's not be that stupid. Many countries have the same needs for office software. Let's work together on solutions that mutually benefit us. Let's try again. Maybe not in Munich but in the next city. Maybe not replacing everything at once but step by step creating open formats being able to run more and more open software. One day we will look back at this time as the dark ages where the latin bible (proprietary formats) could only be read by the clergy (closed software)."
Can't protect your own borders? What do you mean? You should be letting in more refugees. What do you mean you give control to the EU and other nations? Stop being so isolationist and xenophobic!
I jest - but it's fun to see this written from a German perspective, especially when the EU is so critical of the United States and the United Kingdom over these exact things.
I don’t know if Office 365 has - isn’t the on-premises deployment of that simply Exchange Server?
But the Azure Cloud itself has a supported on-premises deployment, disconnected if you want - https://youtube.com/watch?v=qLIzMwgY_Cw so maybe there is an O365 in there too?
This is silly. You live in globalization. You're dependent on foreign companies one way or another, whether it's cheap Chinese made products, Russian gas, or Saudi oil. But you're worried about being 'dependent' on Microsoft Office, from the United States? Get a grip.
> Another reason not to use it is because the FOSS alternatives are free.
Free as in "free beer" isn't free. There are many costs beyond the cost of acquisition that are often overlooked.
Usability isn't nearly as good with the FOSS alternatives. Microsoft has a paid team of usability experts designing and testing the usability of features for many different user audiences, including disabled users who rely on special usability features.
People likely have some familiarity with Office applications, so training costs are lower. For the training employees need, there is a large wealth of training resources out there, from free to pricey.
There is also no question about being able to get support in the future, as Office as been around almost 30 years, and is going strong. Office has no real competitors in the marketplace, so as long as there is money to be made in selling word processors, spreadsheets and other office-centric desktop software, those document formats and applications will be supported.
Office has a very flexible plug-in architecture for customization by thousands of software devs who can develop for that.
Microsoft isn't naive to the needs of European governments, they have government editions of their software and cloud services to align their software with government regulations.
The foreign company argument isn't really meaningful these days. No country is an island (except island nations!). Everything in IT is dependent on foreign resources, PC parts from Asia, tech support from India, Software from US companies, cloud services, etc. If you don't want to have to rely on foreign companies, go back to typewriters.
Can it be used completely offline and off-the-grid while still supporting work-flows between peers? Can the documents created easily be shared with peers not using it? Is one tied to other products due to this product, like OS, certain hardware and other things? Who has authorative power over the software and it's updates?
Here's my main problem with MS productivity software: Quality degradation when including standard formats like png, pdf (not sure about svg these days).
With Apple desktop software you can pretty much copy & paste from A to B and never worry about the result looking bad. With MS I have to follow very specific workflows to get an acceptable result. I don't understand why this is still such a hassle after 25+ years of development. I'd love to have some of Office's advanced features (collaboration, referencing, most excel features, powerpoint's line and endpoint snapping), but with Pages/Keynote/Number's GUI and layouting engine.
> Other than ideology, there are few reasons not to use Office 365.
Isn't that a sufficient reason? Other than ideology, there are few reasons not to kill someone and take his money, if there's no chance of being caught.
Proprietary software is immoral; that is IMHO a sufficient reason to forego it.
There is one very good non-ideological reason not to use it: one should never ever use a sole-source vendor for a mission critical product on a subscription basis because if that vendor decides to raise their price, you have no option but to pay.
This same logic can be applied to FOSS software. It could stop being supported at any moment. There’s also zero financial incentive to fix any critical bugs that arise.
As for non-subscription based software, you are simply delaying a pricing increase as the newer versions are needed eventually as new versions contain key features.
No. With FOSS you always have the option of hiring someone to fix bugs and make enhancements, or form a coalition of customers to hire people and spread the cost, or doing the work yourself. Moreover, the market for providers willing to do that work will be competitive, which will tend to drive costs down.
WRT key features: I'm still using Microsoft Office 2008. In ten years I have never once needed a feature it didn't have.
Imagine a world where all German city governments use FOSS software. Every tax Euro spent on adding critical features automatically benefits hundreds of municipalities. Some of those features might even benefit the general public that can also use the same free office software.
Compare this to money spent on licensing fee going straight to Microsoft.
Now, the data is fully managed by a US company and may be peeked at. Or not.
This is a reasonable concern when you put you sensitive information over there.
>> days of cloud and absolute technical domination of Office 365,
Very weird - for my clients of dozens of companies, Google Docs dominates. Never saw even one client using Office 365 just heard someone negotiating with Microsoft.
I'm all for FLOSS solutions and love Linux and many of the things that are freely available to us. I wouldn't trade Linux for Windows for any sort of development work.
However, I'll be the first to admit that MS Office simply blows Libre Office out of the water. I'm very good with Excel and Calc, but I'm many times slower with Calc. The hotkeys, variety of equations, and simple features simply aren't available in Calc. I know that if I'm say 3x slower, the cost associated with that across a year and across a team is going to outweigh the MS subscription.
There is nothing wrong with ceding the crown to the better product.
Nope. From https://wiki.documentfoundation.org/Development/Java - "Note though that very little of this Java code is actually part of the shipped product." A lot of the Java stuff is ancillary, and chunks of Java code have been rewritten to C++ over time.
Part of the reason that LibreOffice is slow is that it was essentially debugged into existence. Michael Meeks from Collabra would often post to Planet Gnome about his optimization work. There is certainly hope looking forward. It is a lot faster than it used to be, and they will hopefully be enabling Multi-threading in Calc for LibreOffice 6.0.
I've used MS Office, Libreoffice and Google Apps and find them all equally appalling. Terrible, clumsy, complicated, inaccurate UI, unusable version control, ugly output. (Compared to git + emacs + TeX/docbook/perldoc/HTML/plain text).
"Terrible, clumsy, complicated and ugly" are exactly the kind of adjectives that spring to my mind when thinking of git, emacs and (to a lesser degree) TeX.
The switch now announced does not include switching back from LibreOffice to MS Office. They have a lot of existing LO templates which would cost too much (work/money) to migrate.
And really, this is just to say we need to encourage better open source solutions. Open source will never win when closed source is better: Everywhere open source solutions have won is when the quality and user experience has beaten the closed source offering. Consumers, and largely, most businesses, will never choose open for open's sake.
They are going to spend tens of millions of Euro on the migration. The cost of contracting someone to add hotkeys, equations and simple features to libreoffice is nothing compared to that.
The cost of figuring out exactly what to contract somebody to build, though — especially for a non-technical bureaucratic agency — is likely to be fairly astronomical.
(I doubt they are sitting around the conference table saying "Fuck! If only the Linux office suite had these seventeen keyboard shortcuts and the equation editor had these thirty-nine additional features, there would be no need to switch!")
Windows probably really does work better out of the box for their use cases and user base, but I doubt they could tell you why exactly.
The $50 million is just for the switch. They'll probably pay a few million every year for the licenses. Yeah, I don't think they can justify it from a cost point of view. This move was political and Microsoft convinced them to do it by building an HQ there.
> I doubt they are sitting around the conference table saying "Fuck! If only the Linux office suite had these seventeen keyboard shortcuts and the equation editor had these thirty-nine additional features, there would be no need to switch!"
What are the arguments supporting the switch, then? One would think they determined there are some features missing from the Linux based setup and found out that the Windows solution has those, then decided that they have to switch. If this wasn't the case, then what are they basing their decision on? Flipped a coin, or simply corruption, or what?
Not necessarily, one investment buys you the software that meets your needs now, the other is a donation that buys you at most a promise for software that maybe meets your needs sometime in the future.
This assuming they had a need that OSS Office didn't cover, of course.
Don’t disagree, but playing devils advocate, they won’t be on Windows / Office till 2023 per linked article. Five years is a long time in the land of software to resolve promises.
> MS Office simply blows Libre Office out of the water
Graphically, sure. But after not using MS Office for a few years, I really prefer LibreOffice. It does everything I need (and a lot more, actually) and I know where all the buttons are. Other than that it looks prettier, there is zero reason for me to use MS Office. (I can only think of reasons why I wouldn't want to use MS Office.)
And if MS Office is what you're going for, then it still runs in Wine. No need to convert the entire infrastructure.
I have very little experience with OpenOffice. If I remember correctly, OO was bought by Oracle after which the community forked LibreOffice, and that is now installed by default in many GNU/Linux distributions. I think I haven't used OO in at least five years, so I can't really say anything about it.
Read the article (and, if you are able, the ones in German) – this has nothing at all to do with Excel having more hotkeys. LiMux has been around for a long time, and those wrinkles had been pretty much ironed out.
Mayor Reiter openly calls himself a "Microsoft fan". He got Microsoft to move their German HQ to Munich, which was a significant deal for the city. They then hired Accenture, an official MS partner company, to write an report on the city's IT problems. It concluded that LiMux administration is poorly coordinated between departments, and that was all the justification they needed to push for a migration back to Windows.
I am surprised this Techcrunch piece is so "balanced", including a parroting of the understated official cost estimates. This deal is a stupendous waste of taxpayers' money, and German-language news (at least those I've read) are much less subtle about it.
> I wouldn't trade Linux for Windows for any sort of development work.
Same. Windows has all the Linux dev tools plus a huge host of tools that Linux devs can only dream of. The only thing I need/want Linux for is cloud deployment. If I'm deploying an Intranet server, Windows servers is what people want there.
It amazes me that Linux devs speak so highly of their own tools and yet they can't even build a decent desktop system that just works, that masses of people want to actually use. And devs are people too :)
I wouldn't trade Windows for Linux for development and everyday computing, even gaming.
Microsoft and Apple both have slow and clunky UIs. I am actually more efficient on Gnome with extensions than either of the two.
Only use Mac OS and Windows for their compilers and application packaging with cross-platform code bases.
"Masses of, 1st world, people" are also the ones that don't understand or care about concepts such as security, privacy and true right to own. Never found the masses argument that supportive to a valid argument since group-think doesn't actually prevent problems but increases them.
Most developers aren't even using a Linux desktop OS though. Linux desktop distros are so bad that Apple was able to come along and eat up whatever market share it has among developers just by being stable.
> Never found the masses argument that supportive...
It is supportive when discussing things such as necessities. So, for instance, we can all agree that everybody needs to breathe air, right? Likewise, the masses (developers or otherwise) need stability in their OS (among many other things).
GNU/linux usage has been growing along with MacOS. If you include just kernel variants or anything other than desktop computing GNU/linux has been the market leader for a bit now comprising most supercomputers, servers, ChromeOS(variant not GNU), and Android (variant not GNU).
I use Ubuntu for everyday work and have no more problems (sometimes less) than my Mac and Windows co-workers.
Let me know when any real modern OS becomes "unstable". Aside from literal testing and beta releases modern platforms for Windows, MacOS, and linux based platforms are all relatively stable and show significant improvements over 10 years ago.
The only reason I can see for using a Microsoft based platform over any other platform is if you're heavily invested in VBA and macros. If you're not then there is no reason a company cannot use a GNU/linux based platform such as Ubuntu. LibreOffice, Google Apps, Zoho, Collabora/Nextcloud all provide platform agnostic tools that allow for easy use no matter what you're using.
You're working from a very old mindset of operating systems.
> GNU/linux usage has been growing along with MacOS.
That's a euphoric way of saying that Linux desktop systems (the actual thing we're discussing) has nowhere near the same market share, but it's growing I swear!. And of course Linux fanatics can't help but mention how widely the kernel is used even if it's completely off-topic. Should we mention how widely Intel ME/Minix is installed too just for kicks?
> I use Ubuntu for everyday work and have no more problems...
That's cool. Some people also bike to work instead of driving a car and they don't have any problems with that either.
> Let me know when any real modern OS becomes "unstable".
Right now. Linux doesn't even have an ABI for drivers and just this year I think Linus promised a whole 6 years of stable interfaces for drivers. Furthermore, historically and currently, next to nobody makes serious commercial software for Linux desktop systems because of the wildly unstable environment. Look at any discussion thread comparing Linux to Windows or Mac and you'll find plenty of people complaining about Linux drivers, lack of software, unstable/unfit desktop UIs and general fragmentation.
> The only reason I can see for using a Microsoft based platform over any other platform is if you're heavily invested in VBA and macros...
Oh, I thought your opinion would have been completely different after reading the first couple of paragraphs. Let me guess, you also think that command line apps are better than GUIs right? Oh and you think that configuring your system by opening up text files and typing in free-form text is better than using a GUI too.
Obviously the city of Munich and just about every other city and business on earth disagrees with you because none of them want Linux on their desks.
> You're working from a very old mindset of operating systems.
Nope. Your ideas about the state of the Linux desktop are altogether fanciful.
I wonder about the telemetry... or have they decided that the available methods of disabling it are sufficient? I remember hearing that a lot of Germans were wary of Win10 specifically because of that.
If you have Windows 10 Enterprise edition, all telemetry can be disabled. Most people don't talk about that much because at $500 a seat... well, most people don't buy Enterprise edition. (Of course, organizations of certain sizes, government institutions, etc. may have better pricing options available to them as well.)
Also, bear in mind, if you're an organization with a proper modern firewall solution, you can detect and block unwanted network traffic at the edges of your network as well.
It's so confusing to disable that a (admittedly self-proclaimed, but still) "security expert" got it wrong several times, and even after that it still continued sending data to Microsoft. It reminds me of the same scummy tricks used by spyware/adware in the past e.g. "do not disable for the best user experience", double-negative wording, etc.
I believe it's still (deliberately) really easy to pirate/crack Windows, so I could get the Enterprise edition if I wanted --- but I wouldn't want it even if it was legally free, just like all sorts of other malware and unwanted software.
No doubt that it's easy to mess up Group Policies. He's not wrong in that you have to very clearly read each one's description because different policies expect you to do different things sometimes to do the same things. That's definitely not new to Windows 10, and arguably just bad design on Microsoft's part; there appears to be no standardized design pattern to how GPOs should function across the whole Windows team.
But it is still entirely possible to disable all telemetry with it.
There are, however, a number of hacks that can trim that down even further. But being hacks they aren't guaranteed to always work--things frequently change.
Since writing the article mentioned a couple comments up I have tried completely ripping out all telemetry components from the install CD, which seems to be the most effective option. Tools such as Blackbird (http://www.getblackbird.net/) also are quite effective. Nonetheless, monitoring traffic still shows connections to Microsoft-controlled servers. There are so many ways Microsoft (and therefore possible governments) can gather--or at least infer--information about your system.
To completely protect yourself would require manually monitoring, selecting, and installing updates, manually updating certificate revocation lists, constantly watching for new settings (including firewall and hosts entries) that must be addressed, etc.
So no, it is not possible to completely disable all forms of Windows 10 telemetry without also committing to an unreasonable amount of work.
Just a small information that's missing from the discussion: I'm sure this decision doesn't have anything to do with Microsoft (a major taxpayer) moving its German offices from the suburbs (Unterschleißheim) to Munich proper in 2016 (https://mspoweruser.com/microsoft-germany-moves-into-a-new-h...)...
Key users hated Linux and their know-it-all IT administration prescribing them software that nobody else uses. During the time the IT department failed to provide a solution that was better than the MS offering. The user interface is different and the compatibility lacked badly.
Let's be honest. How do you beat MS Office using Linux solutions? Will your Excel documents with macros still work flawlessly? How would you replace Exchange? Maybe Thunderbird with what calendar solution? There are tons of material and training for everyday MS Office use and most people are already familiar with the details they need. For OpenOffice... not as much... OpenOffice isn't even really developed anymore as people switch to LibreOffice.
From people in Munich I talked to, there were a number of interwoven problems. In particular there was a major IT reorganisation at the same time as the switch.
Users also tend to blame everything on Linux, whether related or not.
An yeah at the end of the day, interop with MS documents is not seamless.
That makes sense. When you try something new, every hurdle or setback gets blamed on the new system. When you stick with something old, they're accepted as part of the daily grind.
Munich should have switched to ChromeOS/ChromiumOS or Android or RedHat/CentOS or Ubuntu on thin clients, as the software is anyway usually web based in the cloud/intranet. But neither Google nor RedHat or Ubuntu are known to lobby.
@sz4kerto:
> ChromeOS proprietary
Why you pick this one of a list of many? ChromiumOS is open source, as is Android, as is RedHat or Ubuntu. But Windows is proprietary.
> Google spends a lot on lobbying, by the way, afaik more than what Microsoft is spending.
Really? Source please. The question is how much do they all spend lobbying for an OS? How much do they spend lobbying in Munich? (hint German HQ got relocated to Munich) You read like a employee.
Here is why - even if I risk more downvotes or getting flagged down:
If someone flags your comment on HN, you can only post about 4 comments per day and news submit feature is "practically" disabled - HN displays a "slow down, don't submit so fast" Error message which of course makes little sense. Also your comments will start with a lower/negative value behind the scene, you will find your posts on the bottom even if it says 1 votes, and other comments with negative value may be above yours.
There is a lot of monitoring by companies contracted obviously by $$ going on, to shape PR and public opinion. Be it employees, fanboys, or paid PR drones, it doesn't matter, they are quite toxic to the community at large I would say. I would wish more moderation in regard, and the removal of the downvote- and flag-functionality for everyone (but moderators).
Okay, I understand. I agree that the Hacker News comment system isn't optimal.
Downvotes shouldn't hide a comment. Unpopular opinions shouldn't be suppressed.
I'm not sure how the flag-functionality works, but if it can hide a comment without moderator input, then I'm all for replacing it with a "report"-function.
Too many times one sees throwaway accounts being used for expressing possibly unpopular thoughts. I can only imagine how many comments weren't written, because somebody was afraid of downvotes and losing the good standing of his account. Personally I too have started writing a couple comments and not posted them in the end, because I was afraid of downvotes and the associated consequences. Having to think like that doesn't exactly help to make HN a diverse and interesting place.
Even the giants can die. And as they do, their willingness to help you export your data from their proprietary formats and products will decrease at a rate faster than the neccessity of doing so. Properitary software is a short sighted choice, and probably one made by people with greased pockets.
It's difficult to get funding to work on open source software. Governments are a natural source, but time and time again bribery fucks everything up.
Would it be wrong to say that Microsoft is "too big to fail", similar to some banks? I'd expect governments to actually try and save Microsoft precisely because of this situation.
> Would it be wrong to say that Microsoft is "too big to fail", similar to some banks? I'd expect governments to actually try and save Microsoft precisely because of this situation.
It doesn't really work like that. If Microsoft fails, it will be slow. You don't go from 90% installed base to 0% installed base overnight. There is no point at which they still have enough customers to be too big to fail but don't have enough to continue operating.
But as businesses transition to something else, they have to pay the transition cost. Which is a lot higher when changing market conditions force you to transition unexpectedly than when you have the foresight to plan and get ahead of the curve.
Office's formats are now open, and have been for a long time, now - they're basically just zip archives full of XML files.
Whether they're easier or harder to work with is maybe a topic for debate. My experience from working with them is that OXML tends to be easier to work with than ODF in simple cases, in part because of better open source library support on more languages, and more StackOverflow-type advice. (Yes, open source support for the Microsoft formats really is quite a bit better than for the OpenOffice ones these days.) But it also tends to be the one that produces really gnarly compound documents that are just impossible to make sense of.
I'm not sure that's because the format allows for more complexity, though, so much as because MS Office makes it easier for the kinds of people who use templates that look like spiral bound notebooks with Comic Sans type to create trouble.
Background : I am IT for 25 years, run Linux on all servers (pro and home), develop open source.
I tried to use Linux desktop at least 5 times over the last years (again a few days ago on the home shared laptop).
It is ok-ish when I use the browser and IDEs (vcs, pycharm) but anything else is horrendous. I looked for an equivalent of ShareX, I felt like I was 25 again.
Linux desktop is on for the 86% if the train which only uses a browser ("Internet") but I doubt anything more specialized is reasonable.
Not that Linux desktop is bad (though it is not good either), it's because there are no ports of that specialized software.
If someone is fully (apps, I mean) html5 then why not.
Gaming is also covered, with 2500+ commercial games available on Steam only.
The KDE applications Kdevelop, Kdenlive and Krita has improved a lot the latest 1-2 years, providing a pretty great IDE, video editing and also painting application.
3D animation, modeling and texturing isn't too shabby either, with Maya, Houdini, Blender and 3D-Coat.
For ShareX, there are many alternatives for grabbing screenshots. A direct clone is ShareNix (video: http://www.hnng.moe/f/3CI)
If ShareX is what's stopping you, I think that you're mainly stuck with a Windows desktop habit.
Regarding gaming: it doesn't matter how many games are there, what matters is WHICH GAMES.
Playing games is a social activity after all, especially for teenagers. If all of your Windows-using friends are playing Assassin's Creed: Origins, and you can't because there isn't a Linux version of it then those 2500+ games on Steam do not really matter to you.
No - I am stuck with a solution which works out of the box. I mentioned in an earlier comment that downloading Minecraft and opening it by double-clicking opens an archive. This is simply unacceptable for a desktop version.
At the same time I run and ran hundreds of servers on Linux and I would never run anything serious on Windows Server. I am probably stuck with a Linux habit here.
Download the file from Github - check
Start several commands out of the blue - no idea what to do.
I call a friend, he tells me to start a terminal and type it over there - check.
(98,3% of the population which expected to swap Windows with Linux Desktop is now gone)
Honestly - this is zillions of light years behind Windows.
And again: I run hundreds of servers on Linux (and my home server/router/fw). I hacked in 1994 some of the networking kernel (in pre-modules time) on a 386. So I am hardly a Linux hater.
> Linux desktop is on for the 86% if the train which only uses a browser ("Internet")
Ugh. Stop it. Just stop it. It isn't the late 90s or early 2000s anymore. Hardly anyone uses a desktop computer just for the internet because that market is well served by phones and tablets. That segment of the market existed only for a brief time between the rising popularity of the internet and the invention of purpose built consumer devices for browsing the internet.
Linux Desktop community continues to prop up this strawman user because it is an easy target. Tell them something is wrong with how the OS works that's keeping people from wanting to use it and they'll point to their strawman and say "most people just browse the internet, and we can do that just fine, it must be something else!". Users hate your software? I'ts because they're drooling morons who don't understand it because they just use a computer to browse Facebook. It's a convenient ego-saving lie, like pretty much every excuse that comes out of the Linux desktop camp for why they aren't better. Excuses is what the Linux Desktop community is best at.
I could literally rant for hours about the problems with Linux as a desktop, but the short version is that it sucks because the community behind it sucks, and probably likes that it sucks because it allows them to feel smugly superior to all the "average" (read: drooling browser-only strawman) people.
Even when Linus himself bitches them out they don't pay attention:
"making binaries for Linux desktop applications is a major fucking pain in the ass. You don’t make binaries for Linux, you make binaries for Fedora 19, Fedora 20, maybe even RHEL5 from 10 years ago. You make binaries for Debian Stable…well actually no, you don't make binaries for Debian Stable because Debian Stable has libraries that are so old that anything built in the last century doesn’t work." --Linus Torvalds (DebConf 2014)
I don't think the "browser-only strawman" is condescending elitists dismissing users --- I'm a self-described GNU zealot on most days the only software I use on are apps like firefox, ssh, and vim and a few other bits like gnome/xmonad, liquid haskell, latex, and rustc.
I have no clue how the "average person" uses a computer, but for me it really is all web. Hell, if I had consistent LTE on my laptop, I could very just spend the whole day in a full-screen firefox with a web-ssh connection to my work server for development.
Average home user non-Internet use case # 0: Printing.
If for no other reason, I will never suggest Linux as a desktop option to any of my non-IT-professional friends & family for this one reason.
Use case # -1 would be "owning a laptop." Linux continues to be finicky on laptops - even when all the hardware works, you still end up with rando issues like poor battery life and fans that never turn off due to imperfect laptop power management.
(That said, there is one version of Linux that has cracked both of these nuts for the case of people who really do just want to use the Web: Chrome OS. And I do recommend that to anyone I think it'll work for.)
If you really were all-web, you would just use Chrome OS. As you said, though, you're actually a user of liquid haskell, latex, and rustc. In other words, either a computing hobbyist and/or developer.
There are actually pretty decent webIDEs for all of those things (shareLaTeX's vim mode is pretty nice, too), and I did consider switching to a chromebook. I was actually about to buy one before I decided that the max 8GB of RAM was way too little for my usual browsing habits. That, and I'm dependent on some firefox addons that just don't have good chrome equivalents, so chromeOS is sort of a no-go.
Edit: apparently there's a sizable segment of the market that agrees with both of us!
> In the third quarter of 2016, Chromebooks made up 54% of computer shipments to K-12 classrooms in the United States, says IDC analyst Linn Huang. That market share figure even factors in iPads, which themselves have been successful in education.
I honestly do not understand what you are talking about
I am super lucky that my browser, Visual Code Studio and the Jetbrain suite works on Linux Desktop.
My children downloaded Minecraft. Then double-clicked on it to start it. It opened an archive... WTF? How can a desktop environment can do that?
Then they asked me how to start the game their friends play (Star Wars something) . Sorry kids - Windows only.
So yes - people like me who have uber-restricted needs will be relatively fine - the others will be stuck. Try to get a complicated Excel sheet and open it on Linux, good luck.
The community comment makes no sense: there is no Windows community either so what to compare?
Also: what is the Linux equivalent of Active Directory / Windows domain? How do you centrally manage workstations? (push software updates, configure firewall rules and other access policies, etc.)
> what is the Linux equivalent of Active Directory / Windows domain?
LDAP
> How do you centrally manage workstations?
SSH
This is how my school does it: there is a system user with an ssh key which they use for pushing updates and such (e.g. recently fail2ban was installed on all computers), and we can all log into any system using LDAP.
Active directory isn’t equivalent to a basic LDAP directory not even close.
To have similar functionality you need a commercial directory solution.
Same goes for remote management SSH is how you connect to machines remotely not how you manage them.
Automation of Linux desktop again requires either a commercial product or a huge development undertaking.
The problem with Linux on the desktop is that if you get out of the Microsoft ecosystem you need products to centrally manage users, workstations, configuration and patch management and all of these usually require you to go for other enterprise solutions even if they are OpenSource in their core but still commercial in practice.
Hello. I am a linux lover. I have choosen ubuntu gnome because IMHO, it is the closer to desktop linux and I do not want to spend my life tuning my system. Recently there was some automatic updates and now the chrome scrollbar jumps crazy. I am pretty sure this will be fixed in a couple of days (weeks), but all these little glitches are becoming tiring. Since release 16.04, I have stopped advising people to migrate to linux. I see discutions about removing the features that count for me (network transparency, remote display, GPL license, ...). As a tax payer, I really want my administration to stop giving money to compagnies that work against my freedom. I feel very sad about it but Munich is probably reverting to Microsoft for sensible reasons and not for politic.
At work, I have just migrated to RHEL 7. With RHEL 7.3, there was a problem with gedit (impossible to resize). This ought to be solved in 7.4. In 7.4 most of the applications crash because GTK does not check correctly the values returned by the RandR extension (which is broken in my version of exceed). The current workarounds are either to use xming or deactivate extension RandR (unrealistic in our case because many people use laptop). The package bash-completion is installed by default. It causes so many issues that we decided to mandate its removal in all installations.
138 comments
[ 6.2 ms ] story [ 179 ms ] threadIf you taken a few more seconds to read, that is not what your parent comment said when they wrote "Other than ideology, there are few reasons not to use Office 365."
Such things no doubt exist, but internet users are overwhelmingly more likely to just invoke them ex nihilo when someone happens to post a view they don't like. This is toxic to discussion, so we don't allow it.
https://news.ycombinator.com/newsguidelines.html
Training employees.
Maintaining the environment by addressing bugs.
Securing the environment.
Plus compatibility with proprietary formats isn’t 100%.
I’d still choose FOSS for many kinds of environments but it’s not free.
Your parent comment is referring to the following model:
1. Microsoft provides free or nearly-free licenses to schools. They don't charge the schools for software now, and they won't charge the schools for software later, because that would defeat the purpose of giving the software away. Renewal fees are not an issue anyone sane would worry about.
2. Incoming government workers already know how to use Microsoft software, because that's what they used at school. They complain if they're supposed to use something else.
Right. The presence of that threat is enough to discourage looking for alternatives. "we hear you've been looking into Linux and teaching Python instead of C#/.Net next year... sure would be a shame..."
Try searching for how to perform an arbitrary function in Word or Excel. Hundreds of hits, and probably 25% will solve your problem. Try the same search for Google Sheets or OpenOffice. Crickets was always my experience.
That's assuming you can't get in touch with the person who wrote it. Which you probably could quite easily do.
When I'm in "accomplish a task using a spreadsheet" or "write a document/prepare a presentation" mode, I'm probably not going to be inclined (nor have time) to go rathole on changing the source to the office application, recompiling, testing, doing my original task, and submitting a patch.
There was a time, many moons ago, when secretaries with zero IT knowledge learnt Wordstar key sequences and got very productive using the software; now keyboard usage is considered a highly technical skill among IT workers and rocket science among casual users.
It's just a market driven untraining; everyone can learn to understand and use FOSS, but it needs strong motivation.
Nowadays, Microsoft's Office suite was, is, and, all indicators suggest, will continue to be a bigger productivity multiplier than any of its competitors. For people who primarily work in office software, choosing something like it is the sensible default for any company whose primary motive is profit. Sure, it costs money, but the cost of the software is tiny compared to a skilled office worker's salary - small enough that I'd estimate it wouldn't have to make someone even 1% more productive to justify the retail price, let alone what it costs if you qualify for any sort of volume discount.
There are other motivating factors that could come into play, sure. But if we're talking places that have to pay for their labor (i.e., not home use) and make heavy use of office software, going with the one with the lowest TCO is a sensible default.
Same as with MS Office. I remember me and my parents doing trainings in MS Office, which wasn't exactly free either.
> Securing the environment.
Same as with MS Office.
> compatibility with proprietary formats isn't 100%.
MS Office' compatibility with open formats isn't 100%.
> Maintaining the environment by addressing bugs.
That is a win for MS Office. Now offset that against the license cost and vendor lock-in risk.
I know we conclude the same, namely choosing FOSS for many kinds of environments, but your post sounds a lot like there are many costs unique to FOSS while in reality there are lots of costs involved with both options.
Sure, you can swap Intel with AMD, but isn't that also foreign?
It's a little different for operating systems and all applications that run on top of it...
The home country of Microsoft still has big and active military bases in Germany and spies even on Germany's highest politicians. If they wanted they could pulverize Germany withhin minutes. You can't honestly say that the German government is really independent under these circumstances.
"Look, fellow Germans. We are called inudstry leader and powerhouse of Europe. We export lots of goods for money they owe us. So much that we can bankrupt other nations. Our politicians say that yes, we are a souvereign nation. At the same time we cannot protect our own borders. We give control to the EU and other nations. We let other countries fly attacks on countries we are not at war with from our own soil. We host weapons that could instantly pulverize our whole country. We are spied to up to the highest levels of government. Even down to the lowest levels we fail. We use software that we don't have the source to. Here is this little project of providing some sanity and some freedom. They use software that is actually open to look at and we are free to develop it to our needs. We can take these matters into our own hands and share our work with the world to create a part of a free and open society. Yet, we have kill this project. Not because it doesn't work. It works, just not as well as the foreign made closed software. It doesn't look as good. It is killed by some politicians who want to, yet again, give away some of our freedom to foreign corporations. They want to run the government like a company and they succeed as such. Being a company, not a government. Let's not be that stupid. Many countries have the same needs for office software. Let's work together on solutions that mutually benefit us. Let's try again. Maybe not in Munich but in the next city. Maybe not replacing everything at once but step by step creating open formats being able to run more and more open software. One day we will look back at this time as the dark ages where the latin bible (proprietary formats) could only be read by the clergy (closed software)."
I jest - but it's fun to see this written from a German perspective, especially when the EU is so critical of the United States and the United Kingdom over these exact things.
But the Azure Cloud itself has a supported on-premises deployment, disconnected if you want - https://youtube.com/watch?v=qLIzMwgY_Cw so maybe there is an O365 in there too?
Free as in "free beer" isn't free. There are many costs beyond the cost of acquisition that are often overlooked.
Usability isn't nearly as good with the FOSS alternatives. Microsoft has a paid team of usability experts designing and testing the usability of features for many different user audiences, including disabled users who rely on special usability features.
People likely have some familiarity with Office applications, so training costs are lower. For the training employees need, there is a large wealth of training resources out there, from free to pricey.
There is also no question about being able to get support in the future, as Office as been around almost 30 years, and is going strong. Office has no real competitors in the marketplace, so as long as there is money to be made in selling word processors, spreadsheets and other office-centric desktop software, those document formats and applications will be supported.
Office has a very flexible plug-in architecture for customization by thousands of software devs who can develop for that.
Microsoft isn't naive to the needs of European governments, they have government editions of their software and cloud services to align their software with government regulations.
The foreign company argument isn't really meaningful these days. No country is an island (except island nations!). Everything in IT is dependent on foreign resources, PC parts from Asia, tech support from India, Software from US companies, cloud services, etc. If you don't want to have to rely on foreign companies, go back to typewriters.
With Apple desktop software you can pretty much copy & paste from A to B and never worry about the result looking bad. With MS I have to follow very specific workflows to get an acceptable result. I don't understand why this is still such a hassle after 25+ years of development. I'd love to have some of Office's advanced features (collaboration, referencing, most excel features, powerpoint's line and endpoint snapping), but with Pages/Keynote/Number's GUI and layouting engine.
Isn't that a sufficient reason? Other than ideology, there are few reasons not to kill someone and take his money, if there's no chance of being caught.
Proprietary software is immoral; that is IMHO a sufficient reason to forego it.
Nonsense.
As for non-subscription based software, you are simply delaying a pricing increase as the newer versions are needed eventually as new versions contain key features.
WRT key features: I'm still using Microsoft Office 2008. In ten years I have never once needed a feature it didn't have.
Compare this to money spent on licensing fee going straight to Microsoft.
It would go to some supa-duppa FOSS consulting company, or to yet another Berlin airport.
Now, the data is fully managed by a US company and may be peeked at. Or not. This is a reasonable concern when you put you sensitive information over there.
Similar to what China does.
Very weird - for my clients of dozens of companies, Google Docs dominates. Never saw even one client using Office 365 just heard someone negotiating with Microsoft.
https://lwn.net/Articles/737818/
However, I'll be the first to admit that MS Office simply blows Libre Office out of the water. I'm very good with Excel and Calc, but I'm many times slower with Calc. The hotkeys, variety of equations, and simple features simply aren't available in Calc. I know that if I'm say 3x slower, the cost associated with that across a year and across a team is going to outweigh the MS subscription.
There is nothing wrong with ceding the crown to the better product.
Part of the reason that LibreOffice is slow is that it was essentially debugged into existence. Michael Meeks from Collabra would often post to Planet Gnome about his optimization work. There is certainly hope looking forward. It is a lot faster than it used to be, and they will hopefully be enabling Multi-threading in Calc for LibreOffice 6.0.
https://www.phoronix.com/scan.php?page=news_item&px=LibreOff...
IT team at a previous employer had to setup Windows VM for some sales folks (who otherwise loved Mac) because it was so bad.
Do you think that someone who has never touched computer except for Facebook, MS Office and emailing should know Lisp and UNIX environment?
(I doubt they are sitting around the conference table saying "Fuck! If only the Linux office suite had these seventeen keyboard shortcuts and the equation editor had these thirty-nine additional features, there would be no need to switch!")
Windows probably really does work better out of the box for their use cases and user base, but I doubt they could tell you why exactly.
What are the arguments supporting the switch, then? One would think they determined there are some features missing from the Linux based setup and found out that the Windows solution has those, then decided that they have to switch. If this wasn't the case, then what are they basing their decision on? Flipped a coin, or simply corruption, or what?
Besides, if they using Excel for finance stuff, they're doing it wrong anyway:
https://www.wsj.com/articles/stop-using-excel-finance-chiefs...
This assuming they had a need that OSS Office didn't cover, of course.
Graphically, sure. But after not using MS Office for a few years, I really prefer LibreOffice. It does everything I need (and a lot more, actually) and I know where all the buttons are. Other than that it looks prettier, there is zero reason for me to use MS Office. (I can only think of reasons why I wouldn't want to use MS Office.)
And if MS Office is what you're going for, then it still runs in Wine. No need to convert the entire infrastructure.
When it comes to formatting in Word though, it should be worth noting that MS Word supports ODT files by now.
Mayor Reiter openly calls himself a "Microsoft fan". He got Microsoft to move their German HQ to Munich, which was a significant deal for the city. They then hired Accenture, an official MS partner company, to write an report on the city's IT problems. It concluded that LiMux administration is poorly coordinated between departments, and that was all the justification they needed to push for a migration back to Windows.
I am surprised this Techcrunch piece is so "balanced", including a parroting of the understated official cost estimates. This deal is a stupendous waste of taxpayers' money, and German-language news (at least those I've read) are much less subtle about it.
Same. Windows has all the Linux dev tools plus a huge host of tools that Linux devs can only dream of. The only thing I need/want Linux for is cloud deployment. If I'm deploying an Intranet server, Windows servers is what people want there.
It amazes me that Linux devs speak so highly of their own tools and yet they can't even build a decent desktop system that just works, that masses of people want to actually use. And devs are people too :)
The masses want to use Windows. To make the masses want Linux, it would have to be like Windows.
Microsoft and Apple both have slow and clunky UIs. I am actually more efficient on Gnome with extensions than either of the two.
Only use Mac OS and Windows for their compilers and application packaging with cross-platform code bases.
"Masses of, 1st world, people" are also the ones that don't understand or care about concepts such as security, privacy and true right to own. Never found the masses argument that supportive to a valid argument since group-think doesn't actually prevent problems but increases them.
> Never found the masses argument that supportive...
It is supportive when discussing things such as necessities. So, for instance, we can all agree that everybody needs to breathe air, right? Likewise, the masses (developers or otherwise) need stability in their OS (among many other things).
I use Ubuntu for everyday work and have no more problems (sometimes less) than my Mac and Windows co-workers.
Let me know when any real modern OS becomes "unstable". Aside from literal testing and beta releases modern platforms for Windows, MacOS, and linux based platforms are all relatively stable and show significant improvements over 10 years ago.
The only reason I can see for using a Microsoft based platform over any other platform is if you're heavily invested in VBA and macros. If you're not then there is no reason a company cannot use a GNU/linux based platform such as Ubuntu. LibreOffice, Google Apps, Zoho, Collabora/Nextcloud all provide platform agnostic tools that allow for easy use no matter what you're using.
You're working from a very old mindset of operating systems.
That's a euphoric way of saying that Linux desktop systems (the actual thing we're discussing) has nowhere near the same market share, but it's growing I swear!. And of course Linux fanatics can't help but mention how widely the kernel is used even if it's completely off-topic. Should we mention how widely Intel ME/Minix is installed too just for kicks?
> I use Ubuntu for everyday work and have no more problems...
That's cool. Some people also bike to work instead of driving a car and they don't have any problems with that either.
> Let me know when any real modern OS becomes "unstable".
Right now. Linux doesn't even have an ABI for drivers and just this year I think Linus promised a whole 6 years of stable interfaces for drivers. Furthermore, historically and currently, next to nobody makes serious commercial software for Linux desktop systems because of the wildly unstable environment. Look at any discussion thread comparing Linux to Windows or Mac and you'll find plenty of people complaining about Linux drivers, lack of software, unstable/unfit desktop UIs and general fragmentation.
> The only reason I can see for using a Microsoft based platform over any other platform is if you're heavily invested in VBA and macros...
Oh, I thought your opinion would have been completely different after reading the first couple of paragraphs. Let me guess, you also think that command line apps are better than GUIs right? Oh and you think that configuring your system by opening up text files and typing in free-form text is better than using a GUI too.
Obviously the city of Munich and just about every other city and business on earth disagrees with you because none of them want Linux on their desks.
> You're working from a very old mindset of operating systems.
Nope. Your ideas about the state of the Linux desktop are altogether fanciful.
Also, bear in mind, if you're an organization with a proper modern firewall solution, you can detect and block unwanted network traffic at the edges of your network as well.
I remember reading that this common belief is wrong, but unfortunately I don't remember where/who claimed that. Did it recently change, perhaps?
It's so confusing to disable that a (admittedly self-proclaimed, but still) "security expert" got it wrong several times, and even after that it still continued sending data to Microsoft. It reminds me of the same scummy tricks used by spyware/adware in the past e.g. "do not disable for the best user experience", double-negative wording, etc.
I believe it's still (deliberately) really easy to pirate/crack Windows, so I could get the Enterprise edition if I wanted --- but I wouldn't want it even if it was legally free, just like all sorts of other malware and unwanted software.
But it is still entirely possible to disable all telemetry with it.
There are, however, a number of hacks that can trim that down even further. But being hacks they aren't guaranteed to always work--things frequently change.
Since writing the article mentioned a couple comments up I have tried completely ripping out all telemetry components from the install CD, which seems to be the most effective option. Tools such as Blackbird (http://www.getblackbird.net/) also are quite effective. Nonetheless, monitoring traffic still shows connections to Microsoft-controlled servers. There are so many ways Microsoft (and therefore possible governments) can gather--or at least infer--information about your system.
To completely protect yourself would require manually monitoring, selecting, and installing updates, manually updating certificate revocation lists, constantly watching for new settings (including firewall and hosts entries) that must be addressed, etc.
So no, it is not possible to completely disable all forms of Windows 10 telemetry without also committing to an unreasonable amount of work.
Let's be honest. How do you beat MS Office using Linux solutions? Will your Excel documents with macros still work flawlessly? How would you replace Exchange? Maybe Thunderbird with what calendar solution? There are tons of material and training for everyday MS Office use and most people are already familiar with the details they need. For OpenOffice... not as much... OpenOffice isn't even really developed anymore as people switch to LibreOffice.
Users also tend to blame everything on Linux, whether related or not.
An yeah at the end of the day, interop with MS documents is not seamless.
@sz4kerto:
> ChromeOS proprietary
Why you pick this one of a list of many? ChromiumOS is open source, as is Android, as is RedHat or Ubuntu. But Windows is proprietary.
> Google spends a lot on lobbying, by the way, afaik more than what Microsoft is spending.
Really? Source please. The question is how much do they all spend lobbying for an OS? How much do they spend lobbying in Munich? (hint German HQ got relocated to Munich) You read like a employee.
Google spends a lot on lobbying, by the way, afaik more than what Microsoft is spending.
If someone flags your comment on HN, you can only post about 4 comments per day and news submit feature is "practically" disabled - HN displays a "slow down, don't submit so fast" Error message which of course makes little sense. Also your comments will start with a lower/negative value behind the scene, you will find your posts on the bottom even if it says 1 votes, and other comments with negative value may be above yours.
There is a lot of monitoring by companies contracted obviously by $$ going on, to shape PR and public opinion. Be it employees, fanboys, or paid PR drones, it doesn't matter, they are quite toxic to the community at large I would say. I would wish more moderation in regard, and the removal of the downvote- and flag-functionality for everyone (but moderators).
It's difficult to get funding to work on open source software. Governments are a natural source, but time and time again bribery fucks everything up.
It doesn't really work like that. If Microsoft fails, it will be slow. You don't go from 90% installed base to 0% installed base overnight. There is no point at which they still have enough customers to be too big to fail but don't have enough to continue operating.
But as businesses transition to something else, they have to pay the transition cost. Which is a lot higher when changing market conditions force you to transition unexpectedly than when you have the foresight to plan and get ahead of the curve.
Whether they're easier or harder to work with is maybe a topic for debate. My experience from working with them is that OXML tends to be easier to work with than ODF in simple cases, in part because of better open source library support on more languages, and more StackOverflow-type advice. (Yes, open source support for the Microsoft formats really is quite a bit better than for the OpenOffice ones these days.) But it also tends to be the one that produces really gnarly compound documents that are just impossible to make sense of.
I'm not sure that's because the format allows for more complexity, though, so much as because MS Office makes it easier for the kinds of people who use templates that look like spiral bound notebooks with Comic Sans type to create trouble.
I tried to use Linux desktop at least 5 times over the last years (again a few days ago on the home shared laptop).
It is ok-ish when I use the browser and IDEs (vcs, pycharm) but anything else is horrendous. I looked for an equivalent of ShareX, I felt like I was 25 again.
Linux desktop is on for the 86% if the train which only uses a browser ("Internet") but I doubt anything more specialized is reasonable. Not that Linux desktop is bad (though it is not good either), it's because there are no ports of that specialized software.
If someone is fully (apps, I mean) html5 then why not.
The KDE applications Kdevelop, Kdenlive and Krita has improved a lot the latest 1-2 years, providing a pretty great IDE, video editing and also painting application.
3D animation, modeling and texturing isn't too shabby either, with Maya, Houdini, Blender and 3D-Coat.
For ShareX, there are many alternatives for grabbing screenshots. A direct clone is ShareNix (video: http://www.hnng.moe/f/3CI)
If ShareX is what's stopping you, I think that you're mainly stuck with a Windows desktop habit.
Playing games is a social activity after all, especially for teenagers. If all of your Windows-using friends are playing Assassin's Creed: Origins, and you can't because there isn't a Linux version of it then those 2500+ games on Steam do not really matter to you.
While CS:GO might not be your thing it should be proof that the sitiation is improving.
90% of the game library or 95% of game performance Will not do it.
Until then, having a windows license even if only for gaming is a no brainer.
At the same time I run and ran hundreds of servers on Linux and I would never run anything serious on Windows Server. I am probably stuck with a Linux habit here.
Download the file from Github - check Start several commands out of the blue - no idea what to do. I call a friend, he tells me to start a terminal and type it over there - check.
(98,3% of the population which expected to swap Windows with Linux Desktop is now gone)
I am IT, the terminal is my friend.
I start ` sharenix` and get
(sharenix:2817): GModule-CRITICAL : g_module_close: assertion 'module != NULL' failed Gtk-Message: Failed to load module "canberra-gtk-module"
So much for the clone and the experience.
Honestly - this is zillions of light years behind Windows.
And again: I run hundreds of servers on Linux (and my home server/router/fw). I hacked in 1994 some of the networking kernel (in pre-modules time) on a 386. So I am hardly a Linux hater.
Ugh. Stop it. Just stop it. It isn't the late 90s or early 2000s anymore. Hardly anyone uses a desktop computer just for the internet because that market is well served by phones and tablets. That segment of the market existed only for a brief time between the rising popularity of the internet and the invention of purpose built consumer devices for browsing the internet.
Linux Desktop community continues to prop up this strawman user because it is an easy target. Tell them something is wrong with how the OS works that's keeping people from wanting to use it and they'll point to their strawman and say "most people just browse the internet, and we can do that just fine, it must be something else!". Users hate your software? I'ts because they're drooling morons who don't understand it because they just use a computer to browse Facebook. It's a convenient ego-saving lie, like pretty much every excuse that comes out of the Linux desktop camp for why they aren't better. Excuses is what the Linux Desktop community is best at.
I could literally rant for hours about the problems with Linux as a desktop, but the short version is that it sucks because the community behind it sucks, and probably likes that it sucks because it allows them to feel smugly superior to all the "average" (read: drooling browser-only strawman) people.
Even when Linus himself bitches them out they don't pay attention:
"making binaries for Linux desktop applications is a major fucking pain in the ass. You don’t make binaries for Linux, you make binaries for Fedora 19, Fedora 20, maybe even RHEL5 from 10 years ago. You make binaries for Debian Stable…well actually no, you don't make binaries for Debian Stable because Debian Stable has libraries that are so old that anything built in the last century doesn’t work." --Linus Torvalds (DebConf 2014)
I have no clue how the "average person" uses a computer, but for me it really is all web. Hell, if I had consistent LTE on my laptop, I could very just spend the whole day in a full-screen firefox with a web-ssh connection to my work server for development.
If for no other reason, I will never suggest Linux as a desktop option to any of my non-IT-professional friends & family for this one reason.
Use case # -1 would be "owning a laptop." Linux continues to be finicky on laptops - even when all the hardware works, you still end up with rando issues like poor battery life and fans that never turn off due to imperfect laptop power management.
(That said, there is one version of Linux that has cracked both of these nuts for the case of people who really do just want to use the Web: Chrome OS. And I do recommend that to anyone I think it'll work for.)
Edit: apparently there's a sizable segment of the market that agrees with both of us!
> In the third quarter of 2016, Chromebooks made up 54% of computer shipments to K-12 classrooms in the United States, says IDC analyst Linn Huang. That market share figure even factors in iPads, which themselves have been successful in education.
https://www.fastcompany.com/3067267/how-chromebooks-aim-to-f...
I am super lucky that my browser, Visual Code Studio and the Jetbrain suite works on Linux Desktop.
My children downloaded Minecraft. Then double-clicked on it to start it. It opened an archive... WTF? How can a desktop environment can do that?
Then they asked me how to start the game their friends play (Star Wars something) . Sorry kids - Windows only.
So yes - people like me who have uber-restricted needs will be relatively fine - the others will be stuck. Try to get a complicated Excel sheet and open it on Linux, good luck.
The community comment makes no sense: there is no Windows community either so what to compare?
The browser platform makes the kernel completely irrelevant, ChromeOS is a very good example of it.
Politicians never fail to disappoint, feels like.
That said, I often can't replace Windows due to legacy applications.
Regarding MS Office, some people buy iPhone. These people also seem to need their Microsoft Office.
LDAP
> How do you centrally manage workstations?
SSH
This is how my school does it: there is a system user with an ssh key which they use for pushing updates and such (e.g. recently fail2ban was installed on all computers), and we can all log into any system using LDAP.
To have similar functionality you need a commercial directory solution.
Same goes for remote management SSH is how you connect to machines remotely not how you manage them.
Automation of Linux desktop again requires either a commercial product or a huge development undertaking.
The problem with Linux on the desktop is that if you get out of the Microsoft ecosystem you need products to centrally manage users, workstations, configuration and patch management and all of these usually require you to go for other enterprise solutions even if they are OpenSource in their core but still commercial in practice.
There are different solutions, such as Red Hat's Satellite:
https://www.redhat.com/en/technologies/management/satellite
Or Canonical's Landscape:
https://landscape.canonical.com/
But a tech-savvy team could do the same with Ansible, Puppet, Chef, etc.
How would that even be possible? Are they going to rewrite all the GPL code?
It would be interesting to see if they're getting an unusually good deal. Though I suppose there might be ways to hide that if they wanted to.
https://news.ycombinator.com/item?id=15765990
https://news.ycombinator.com/item?id=15780556
https://news.ycombinator.com/item?id=15661372