The costs seem to mention training and support, but not comparing productivity makes it useless. It's like arguing that you can use the Gimp rather than photoshop, which is only true until you try.
The whole point of software is that it increases productivity (word processor rather than typewriter). While it may be the case that Linux and open source alternatives saves a bunch on licenses, I don't buy the argument that it is an amount "saved" until thurough research is made in the productivity of the users.
The study MicroSoft commissioned from HP put the total cost at about €60 million or between two and three times the costs Munich claims and substantially more than their estimate for using Windows.
That said and even if it's fabricated from whole cloth, the numbers are all chump change in the budget for a city with 1.4 million inhabitants, particularly when the spending is spread over more than a decade. Ten million euros in savings is less than a Euro per inhabitant per year since the project kicked off in 2002. Using Munich's numbers, they could do the entire project every year with a head tax of €25 per year.
€10 million is a lot of money if it's yours. At the scale of a city it's repaving a few streets.
One might as well expect productivity to increase switching away from proprietary software. I would, at least.
Taking your example of Gimp, I expect institutional adoption of free software would result in some people having it who didn't have Photoshop at all before. They can now become more productive and take less of other people's time.
Well, it could work either way I suppose, but in my experience, usability/productivity of desktop software is severely lacking in OSS software (unlike e.g server software). Perhaps due to problems attracting non-developers to work on them. Polish is expensive, but polish is productivity.
The potential upside like you mention would be more licenses at lower cost. Whether that was a bottleneck depends on the organization in question. Either way I don't think it's possible to make any conclusions without research.
I'd expect a loss of productivity from using OpenOffice that far exceeds the cost of Microsoft Office, plus ongoing losses in productivity as staff try to cope with its incompatibilities with complex Office documents that probably arrive on a daily basis.
However, staff productivity is not generally a major concern in local government offices, as far as I can tell. Maybe Munich is different, but I doubt it...
>One might as well expect productivity to increase switching away from proprietary software. I would, at least.
That doesn't make any sense. What is inherent about closed-source that makes it hard to use and what is inherent about open-source that makes it easy to use?
I think the important phrase here is switching away. That implies there's been a change, which brings increased training (either formal or learn-as-you-go) to learn the new system. If they need Word for their job and now they have OpenOffice, that's not a productivity gain, that's a loss at least until they get up and running (and then it's an ongoing loss every time they need to convert a docx with custom formatting that OpenOffice doesn't understand.)
There is such a thing as inherent productivity in the software though. Switching from one piece of software to another has a fixed cost, but switching to software that makes you more productive will help offset that cost.
I agree on OO. The risk in my eyes here is that OSS (desktop) software is almost invariably less productive than proprietary software for some reason. Switching that way can make sense if you save a lot on licenses, administration etc., but if you have to pay once for a switch AND then pay continuously by using less productive software you are taking a much higher risk. It may give a warm fluffy feeling to be using open standards, but the problem is that the only standard that exists is a de-facto proprietary standard set by Microsoft. It may be noble to try to change that, but I certainly wouldn't want to spend tax money being the first major city to do so.
> Well while we don't need to state reasons, there is also the risk that all proprietary software is less productive for some reason.
Only from simple empirical studies of a few software packages (OO vs MS Office, Gimp vs PS, Eclipse vs IDEA, Blender vs. many) I think it's fairly obvious that the "normal" open source model doesn't match the megacorp development model for complex desktop software. My best guess at the reason is that 1) Ivory Tower is better than committee design for user interfaces 2) you need a LOT of expensive non dev resources such as designers and focus groups which just don't seem to materialize for the open source projects like they do when Adobe or Microsoft waves a pay check.
I think there may be lots to save in administration, servers etc when moving to Linux, but 2014 isn't the year of desktop Linux, far from it. You really can't motivate not buying office for a few bucks per month and expect to save money.
You CAN probably save a ton by ditching exchange/SQL server and so on, where there are plenty of good alternatives. But Ubuntu/OpenOffice instead of Win/Office? I doubt it, given reasonably low volume license costs.
The people using the software are being paid roughly €2,500 to €5,000 a month, so obviously Munich can't afford €2.5 to €5 a month or whatever Microsoft Office would cost them....
(Average gross salary in Germany was €3,391 in 3Q2013 so the cost of employing someone would be higher than that.)
Nobody cites Windows or Office as an impediment to growth. It is very unlikely that the Windows Web client share decline has ever hit a "productivity" speed bump.
Another way to look at it is that mobility turned out to be much more highly valued than Office or Windows compatibility.
That just says that web browsing is now a mobile activity. I think this project is about replacing desktop and laptop computers with linux. On desktop laptop windows is still dominating and losing mostly to MacOS
Office productivity is now a Web activity, in part because that's how modern tools can be used on otherwise locked-down desktops. Whether it's desktop Linux or people bringing their own iPads to workplaces with PCs running old versions of Windows, IE, and Office, nobody sees the supposed productivity of their desktop systems as a reason not to abandon them.
The cost of win (desktop) licenses is virtually zero, and office subscriptions are also dirt cheap. That's not where you save money. It's a few dollars per month. You can motivate that cost even for someone who uses office apps casually once a week.
And then we should remember that the big cross platform ones are used by an audience 20x larger than Linux only applications, so they should be the good ones. What's a good, large/complex OSS app that beats its proprietary competitors? It sure isn't gimp or eclipse. It's Chromium! Which is backed by a megacorp. See the pattern? I have yet to see an OSS project impress me with a large/complex desktop application. I'm sceptical it will ever happen.
Sorry, but it's not like arguing you can use "Gimp rather than Photoshop" at all.
You just identified a specialist application that's used in a specific use-case by a limited user-base who fully use the capabilities of the software. When a designer uses those tools their productivity is significantly increased or limited by the software. The focus of the role is using the computer as a tool, so if it performs poorly so do they.
Most 'general office' users, and certainly many administrators don't fully use the capabilities of the software they have because it's not the focus of their jobs. The software isn't the limiting factor in them doing the job. Take the example of someone working in job centre or similar, they'll likely use a Word processor for providing letters, e-mail for external/internal interaction and some custom internal applications (e.g. the jobs database). There are all well within the capabilities of Open Source software.
Framing it that way would certainly be an argument against replacing Photoshop for Gimp; but it doesn't tell you anything useful about replacing general Windows/MS Office with Linux and Open Source.
I think the same argument holds for OO vs MS Office as well. They are much closer than Gimp/PS obviously, but for power users they aren't even close to interchangeable.
A good middle ground would be migrating the 95 or 99% who are "casual" users to open source, but you'd then have a heterogeneous system, which isn't ideal. VM's could be used to have all-Linux desktops and still provide users of win-only software with what they need.
My point is: the big savings are in the 95% but the big cost is for migrating the last 5%.
The LiMux wiki page mentions a timeline of how the migration happened, so it looks like their phased rollout helped them 'convert' to Linux smoothly.
Quoting from the wiki..
September 6, 2005 - It is decided that the project needs an additional one year pilot test, and migration slips one year.[19]
September 22, 2006 - The "soft" migration begins, one year behind original schedule.[20]
November 2008: 1200 out of 14,000 have migrated to the LiMux environment (9%; March 2008: 1000=7%), in addition 12000 workstations use OpenOffice.org 2 installed on Windows (March 2008: 6000) and more 100% use Mozilla Firefox 1.5 and Mozilla Thunderbird 1.5 (March 2008: 90%). 18000 of 21000 macros, templates and forms are changed into Linux-enabled[21]
One thing bureaucracies are susceptible to is vendors requiring mandatory software updates on custom software. This costs much more than most people think, and with the city of City of Munich would dwarf any productivity overhead.
Sure it's the case with Photoshop as it's an off the shelf piece product, but it won't be with CustomVendorDirectXAccounting v13.5. I doubt City of Munich has teams of people running photoshop, and any designers likely have suite of macs anyway.
I'd wager the cost savings of switching to OpenSource likely won't come from initial costs. It will come from forcing custom software vendors to distribute software that inherently requires less ongoing costs.
I am not at all surprised that this worked. The biggest problem for mainstream linux is poor advertising. People with limited computer knowledge are just scared of change, and honestly, if no one knows how to pronounce Ubuntu then they are probably not going to pick a computer based on it.
I mostly use Windows 7 and OS X, but I have worked with Debian/Ubuntu as a casual user. Linux was much easier to work with than Windows 8.1, but 90% of the population doesn't know that. Most people would be as happy, or happier with Linux if there was just one big advertising campaign.
Well, Linux is now the most commonly used general purpose OS/kernel in the world. Every Android device, every Chromebook device (plus all the regular desktops and servers).
I can't help but feel that part of Android and Chromebook's success is in part due to the users not understanding/having-to-know they're running Linux. Android and Chromebook don't even mention Linux anywhere, and users are content. I think regular Joe's have a stigma about Linux, and just abstracting that aspect away has helped adoption.
Android may technically be Linux, but I am not sure I would count it as a step towards changing the desktop computer people get for their home. Chromebook is similar- as far as I can tell it just runs one browser window. This makes it more like a video game system than a PC.
In industrializing parts of the world, mobile devices are often the only access to the internet and software (and sometimes the first experience with a computing device).
Chromebook/ChromeOS is much more capable now than it used to be, with multi-tasking and multiple "apps" can be on-screen at the same time.
What about long-term compatibility of Microsoft (or Apple) products? How can proprietary, maybe even binary file-formats be accessed in 100 years from now? Cities usually keep their books forever. I have so many files from the nineties that are already very difficult to restore. I could imagine that open source formats have a huge advantage here?! #LongtermCosts
When a document stops being a document and becomes a record, it should be saved in a format such as PDF/A for longevity, and not a binary format. That would be regardless of if it was created in libreOffice or Microsoft Office.
Short term costs work the other way though: every company or organization needs to be able to read an office file in the latest format, edit it, and return it to the sender with full fidelity. If OpenOffice inserts a page break where MS office wouldn't have, that is a problem. The open programs read docx reasonably well, but implementing ms word clean room with all it's behaviors is near impossible, and every difference hides a cost for the users that communicate.
Within an organization you could agree to use some open format (regardless of word processor), and you could archive in some long term format. But you can't change the fact that incoming documents will be in the format that Microsofts latest word processor spits out, and you never want to have that nagging feeling that your word processor may have misread it ever so slightly.
There are legal problems with binary file formats. Let's say someone comes forward allegating corruption 20 years ago and starts suing the city. Can the city prove that it was all above board? Well no-one can read the files anymore....
It reminds me http://xkcd.com/934/ for most simple users. What a government officer do at his/her workstation? They probably have every service through web pages, almost nobody builds native programs anymore. And for the most cases LibreOffice is not perfect, but adequate to do simple jobs. So it is very logical to use something free for these tasks, to replace Windows+Office and save money.
Outside of the HN echo chamber there are thousands of companies building native applications, which are used by millions of organisations around the world. Not everything is delivered via the web.
What a strange view of what a "government officer" does.
If you include all civil servants (which the definition listed by wikipedia does) you will find police officers, national economists, sociologists, intelligence officers, surgeons, traders and many many more professionals there.
All the specialists will of course require domain specific software. In anything involving economics or statistics you will probably find SAS or R, someone tasked with planning mobile spectrum would have GIS applications etc.
Yes, but for every person who needs specialized software to do their job, there are many who don't.
For instance, even most police officers in the area where I live don't need special software. Some do, but most use a few web apps and java apps to do their paper work.
It's like a non-tech mid-sized business, really. All the management and many of the employees don't need anything more than word processing. And then there are people in technical roles whose needs are served on an individual basis (or, in good environments, their manager is handed a budget and told "have you people get what they need")
In my time working for financial firms (~4 years), I can tell you that there are a lot of proprietary applications being developed with web technologies.
I really wonder if LibreOffice is not totally adequate for the users in the city of Munich.
I know that it has its problems if you want to use it with Word generated documents, and it obviously has problems with add-ons developed for Word.
But I could really think that it works quite nice, if all your documents are generated for LibreOffice.
I am using Open / LibreOffice since nearly 10 years exclusively at home and I don't have any problems using it - but my needs are VERY modest. Basically I am using it as an WYSIWYG typewriter.
I use LibreOffice in an environment where almost every other user uses Microsoft Office tools. I've had difficulty with some documents in the docx format. Most of the documents will turn out to be ok but when certain types of formatting are included problems start to appear.
Most of the problems come from, I think, unnecessary features. For example a footer for the author's name on each page.
I think this is really sad. As someone who went through university when the last wave of centralized unix died off, then the move to decentralized windows, and to decentralized linux (because they couldn't really figure it out mostly).
I found that move to decentralization and the availability of compilers and tools and dev. environments and easy ways to get admin privileges, well I credit that with my learning how to code/admin/... and to my succes later in life. These computers, I've heard from a friend, are LOCKED down. As in, try to compile anything against a GUI library is a complete non-starter. Compiling anything under linux without root is almost impossible.
Add to that that even the development environments that do exist on linux are ... well sorry to say but let's be honest here ... far inferior for learning to code and to writing simple games than the ones available on windows.
Nobody's going to learn how to use computers, how to program and how to improve their situation on these linux machines. And that sucks.
But hey, easier on the admins !
(not that it's anywhere near as bad as SAAS though)
Compiling anything under linux without root is almost impossible
This is bizarre, and very much not true. If you have a C compiler and sufficient disk space you can build whatever you want in your home directory. It might take a while, but most things that use autoconf will accept `--prefix=/home/user/stuff/` and target that location.
Not having a C compiler is rare as it tends to be deep in the requirements chain of other packages. For other languages, I believe Python is mandatory on Redhat and Perl on Debian, if you don't mind old versions.
FUD. Try to get anything done in a Windows group policy environment. Same problem, different OS. The serious IDEs are (mostly) available on all platforms, except for Visual Studio (naturally). The machine is installed to be used by a regular user, why would you want it to be developer-friendly? Regular civil servants need to compile applications?
The problem is usually interacting with other agencies; those don't have well-defined data formats for data interchange, instead they often work with forms in Word documents, and require the document to be edited with a very specific version of MS Word. Of course that makes interaction harder.
The technical lead of the LiMux project told a story once about a German federal agency that required documents to be submitted in a very specific, unusual font that basically everybody had to buy first.
I am sure that requirements like this aren't that unusual (the NIH requires grants to be submitted in one of 3 fonts), but an arbitrary rule shouldn't be confused with a compatibility issue.
As you note, these requirements tend to be with specific versions of Word, and so libreoffice and co present no more of an impediment that using other/current versions of Word.
Joke aside, it makes me happy to see the transition from Microsoft to Linux has been a success. I remember reading the news headline when they first announced it. I wonder how many of the 22,000 government officers that use Ubuntu at work are now also using Ubuntu at home?
The display drivers needs a lot of work (intel seems solid). Also, desktop itself needs to be super stable.
Also, the software (from companies) is treated like a second-class citizen. Chrome is horrible on linux, --disable-gpu --audio-buffer-size what user is going to bother with that?
I have to do all these tweaks and workarounds to get things to work. Regular users will not have the patience do this.
The year of LOD will only happen when everything is stable and works. Imagine if CentOS, Debian, or Ubuntu on a server were this flaky? Would you use it? This is why people jump ship to Mac.
Yep. I think it has to do with AMD drivers but I am not sure (I have seen some Nvidia users complain too). It started around Chrome 33.
By the way, I don't think ChromeOS is a good example. Can you use Firefox on ChromeOS? I consider ChromeOS like purpose-specific, dedicated hardware (like a router). It is a start though. By the way, Chromebooks use intel gpus - they are pretty solid.
LOD has to be better than alternatives by a significant margin in order to attract users.
My biggest gripe with Chrome on Linux is you can't change the default behavior of clicking in the address bar. In Firefox I can make it so clicking once selects everything. In Chrome, it drops a cursor and I have to click two or three times to get it to select everything. I understand some people want that behavior, but I don't. The Chrome team's response is that this is the Linux default and Chrome is a native Linux application, so that's how it behaves and it can't be changed.
So then they add that to the other 10,000 config options people have requested, and now it's a bloated interface and people will lament "Why can't they keep it simple?" until they strip out all of the pet features and people get mad for them "stupifying" the interface.
Who says it has to be a config option in the interface? In Firefox it's an option buried in about:config under an obscure name, not user-facing unless someone really wants to find it.
My peeve for a long time was not having the bookmark bar work (drag the bookmark over a folder and popout the submenu). It worked on Mac and Windows. The just fixed it a few months ago (it's a little clunky though but I will take it)! Yay!
I'm currently running Chrome 35.0.1916.153 on Ubuntu 14.04 with XFCE and clicking once in the address bar highlights everything all. Don't know why mine is acting different than yours, maybe they changed that since you last tried it.
I last tried it years ago, so it's possible they changed it since then. When I read that response on their forum, I completely gave up on Chrome and haven't touched it since then. I don't care if "it's not the Linux way" [1], it's my way and who is Google to tell me otherwise?
They are not, unless there are Ubuntu computers available in stores. I actually know there are some German laws requiring computers be made available with varying or no OS, so it is much more likely than it is stateside to find an Ubuntu machine in store.
Unfortunately this is not the case. I only new 2 Online-Stores in Germany that sell Ubuntu maschines - one is my own: http://rockiger.com/
On the other hand, System76 computers are not available in Germany.
Best bet is probably is Sputnik from Dell, although their Ubuntu configuration is not that good.
Except you aren't giving Microsoft money while buying a Windows laptop. Nor can I see what you'd benefit from spending extra money to not give Microsoft money.
Not giving Microsoft money means you are not funding the exact reason why Linux on the desktop is obscure in the first place.
If MS did not have such comprehensive control of the consumer computer space, they would not have pervasive "deals" with every vendor to lock them in to shipping Windows across the board. Because if you, as a hardware manufacturer, install Ubuntu and make sure all the drivers work, out of the box it is no more buggy than Windows is at this point - probably less, and much more usable than Windows 8 for the average joe.
But Microsoft has a firm grasp across the entire industry on all the vendors, and their ubiquity for 20 years has given them mindshare you cannot easily unravel. And giving them more money is not helping that.
I believe it's been well documented by many writers that Linux on desktop (LOD) won't be a thing in the foreseeable future.
If you think about it, Android's problems with open source allowing manufacturers to have their own flavours of Android, is one of the reasons why the average user might think TouchWiz (or other skins) is Android and that's why it sucks.
Note that this is just _one_ reason I believe LOD won't be a thing, there surely are more.
"If you think about it, Android's problems with open source allowing manufacturers to have their own flavours of Android..."
One is not really dependent on the other. You could use open source and not allow modifications or you could use closed source software and allow endless customization.
I also don't consider it an "android problem" that manufacturers do this. Most of the good stuff ends up in standard android after a while, which wouldn't have happened if no one was allowed to make the changes in the first place.
> One is not really dependent on the other. You could use open source and not allow modifications or you could use closed source software and allow endless customization.
Good point.
> Most of the good stuff ends up in standard android after a while, which wouldn't have happened if no one was allowed to make the changes in the first place.
This was mainly due to Google pushing more stuff through Play Services. It would be nice if there was some more control/constraints on Linux flavours.
The numbers don't make sense to me. €34 million across 14,000 computers is about €2500 per computer. That seems a bit on the high side for the sort of computers a municipality requires. Particularly when an older computer provides suitable performance (otherwise the Linux boxes would need comparable upgrading).
I'll add that a better comparison might be to the cost of Windows machines running open source productivity software since the comparison is closer in terms of specification.
Except it isn't - you need to consider the costs of not just Windows, but all the software surrounding it. Photoshop licenses, Visual Studio licenses, Office licenses, etc. Even at business rates. Per workstation, $3000 can be an underestimate.
Meanwhile, using Gimp / Krita / LibreOffice / kdev etc and, at the scale of a city like Munich, just having a dev team to fix the localized problems the city has in the software, or having a fund to bounty bugs and features the city wants, would be cheaper in the long run, and leave the city in control of its own software infrastructure.
The humans being vastly more expensive than either the software or the hardware... which is why Munich is most likely in a deep financial hole over this project, due to the huge waste of staff time
I've only used Sharepoint very cursorily and that years ago, so my experience isn't the best or understanding complete.
As a first approximation: a Wiki.
Tools such as Atlassian's Trac and Jira (not themselves open sourced) offer a fair bit of the equivalent features.
For setting up Internet portals, a CMS such as Drupal or Wordpress could fit the bill.
One of the features of Sharepoint sites that I've seen is that they're used effectively to "hang" MS Office documents off of them -- as Office docs. This grates strongly with my preference for either HTML or PDF formats for documents, plus some markup style. I'm growing increasingly strongly partial to using either Markdown (for simple documents/pages) or LaTeX (for more complex ones) for document preparation. Yes, there's a touch of up-front training involved, but once you've crossed that bridge, the world opens up, and the basics can be gotten in an hour or so.
The document management pieces are actually something I'd like a solution for more generally, and it's a space that's hugely frustrating. There's zotero, a desktop application, there's git, there are a number of other tools, each of which effectively only address part of the larger picture.
You can write down the list of sharepoint features you need, and then find individual OSS projects that fill the gap. It's not a one-stop thing, but it doesn't have to be a deal breaker either.
SharePoint is a Jack of all trades, master of none so based on you requirements you have to choose the right tool most of which are OSS or at the very least non depended on Windows
Definitely true. A wiki provides a good chunk of what I've seen folks in academia use SP for but I've seen other companies use primarily the office-type features (calendar, contacts, etc). But if you setup a wiki along with something like Zimbra and that could provide most of what SP offers and some extras.
It really depends on your requirements. What is it that you need? Like any software SP is a tool, just like home improvement you don't go out buy a tool and than try to use it for what ever task is in front of you (you might end up with a screw driver while you really needed a hammer). So I could make a list of Document Management, Content Management, Intranet, etc software but the correct answer depends on your needs since there are lots of choose and all (some) have there own strong and weak points.
SP is a mesh up of several software suits slowly coming to getter, it is getting better every release. SP could also be the right choice depending in your needs. Mainly work group portals but those tend to fail (as in not being used) since most of the time there is no proper business case nor business processes in place to support the grand idea
If you respond with specific requirements I'll be glad to give you a rundown of options for each requirement, as well as good "package deals" of software that works well in concert.
As I mentioned common open source productivity software runs on Windows. That includes Gimp, Emacs etc. - not that I believe that Munich is rolling out 14000 development/graphics boxes. The comparison does not appear to be between equivalent specifications. It's assuming that Windows requires Office. That's simply false.
Retraining costs, mostly. The crowd on HN can sometimes forget that your average computer user, even one that uses it professionally - database entry, or just using some enterprise software daily - is literally hard coded trained to click on specific buttons in specific places on Windows, with nill comprehension of what they do, and if given any alternative layout would immediately stop and be unable to function.
Why do you think Windows 8 has seen such enterprise backlash?
Office not being available for Linux does not imply that Office is required by Windows. The difference between platforms is the availability of Office and it's cheap and easy to use something other than Office on a Windows machine, don't buy Office and install the other thing instead.
AutoCad isn't available for Linux either, but it no less sense rolling it into the price of a Windows license than Office.
I'm very impressed with the scale of the project. I can definitely see both the short term and long term ramifications and huge huge huge potential savings.
I'm a bit curious to know what the shelf life on the OS choice is and how well it works. I believe redhat ran into this issue and adopted the "extended support" versions in order to address the concerns (the linux space tends to move _very_ quickly in directions and you can definitely be left out in the cold if you don't stay relatively current).
All in all I doubt it's worse than the microsoft "you need to upgrade _now_" philosphy and the lack of "we don't break compatability" that used to be enjoyed in the heyday of the "Full M$ Stack".
Back with windows 95 Microsoft actually took backwards compatability very very seriously:
An emphasis has been made by Microsoft on maintaining software backwards compatibility. To achieve this, when developing a new version of Windows, Microsoft sometimes had to implement workarounds to allow compatibility with third-party software that used the previous version's API in an undocumented or even (programmatically) illegal way. Raymond Chen, a Microsoft developer who works on the Windows API, has said: "I could probably write for months solely about bad things apps do and what we had to do to get them to work again (often in spite of themselves). Which is why I get particularly furious when people accuse Microsoft of maliciously breaking applications during OS upgrades. If any application failed to run on Windows 95, I took it as a personal failure."[19]
The whole point of software is that it increases productivity (word processor rather than typewriter). While it may be the case that Linux and open source alternatives saves a bunch on licenses, I don't buy the argument that it is an amount "saved" until thurough research is made in the productivity of the users.