Slashdot Mirror
Munich Votes for Linux Migration Plan
JoScherl writes "The German news site Heise reports (German, Babelfish version) that the city council of Munich (3rd biggest city in Germany, 1.3 million inhabitants) has voted for the detailed concept of the LiMux - Linux for Munich (German, Babelfish version) project with votes from all parties except the CSU (Christlich Soziale Union, christion social union). With this decision the 13,000 Desktops and Servers of the city administration will be migrated to Linux. CSU, which has just won the European elections, said they won't support Linux since its Feierabendprogrammierer ('leisure-time coders') would destroy Munich's IT-landscape (Microsoft Germany and other big companies are located in and around Munich) and they also fear that the personnel would have problems with learning how to use OpenOffice and other migrated systems. The migration plan has the following steps: This year the Windows NT desktops get OpenOffice and Mozilla as their default office and browsing suite. In 2005 and 2006 the systems will be migrated to Linux, with some applications running on Windows application servers. In 2008 all applications should run native on Linux."
...of those bargaining things where they are just trying to get a better deal from Microsoft?
Ok Linux is a good OS, but they're about to have to retrain approx 16,000 workers, many of whom never heard of Linux and some are total creampuffs in computers. They will be retraining to a platform the users may not like as well. The long run costs will probably be worth it as upgrades are free. However the short term costs of re-training I shudder to think about. At 16,000 workers they need a whole university's capacity of retraining. In fact any Linux guru looking for a job? Munich sounds like a good bet...
...in bed
Is IBM going to donate the services (as in lots of IT help for free, again) of a large crew of techs to assist in the transition like they did for the earlier part?
Linux fans had better hope that this goes well because if it doesn't you can guarantee that Microsoft will be hopping up and down screaming "I told you so".....or "Ich tolden youze sozen" (in German)
European companies buying for their European offices pay with (currently strong) Euro, so it's not neccessarily meaningful to convert it to USD. Also, consider that migration project budgets typically include hours and buffer - 2000 Euro per machine isn't that unreasonable, especially if you take for granted that it's worth it to escape the "MS tax."
So they are pretty much the republican party? Mouth the words but don't give a shit?
It's a nice idea, sometimes. However, bear in mind that the people involved are city administrators. So your comment about it (OO etc.) being extremely simple is optimistic. It's also notable that there are 13,000 desktops involved. How much time do you want to put into deskside support? If they get people used to the application on a familiar OS (i.e. something about the new environment feels the same) then they can cut across later with much less resistance.
Do you have any idea how government / business works when implementing / changing new technology? 4 years is actually a remarkably quick time to change ALL software over to Linux.
So they have decided to do this. Firstly, they have to determine what problems they will encounter. What apps might they need that they may have difficulties finding under Linux? Code may have to be migrated from ASP / whatever. Excel / word macros rewritten. And this is just the tip of the iceberg.
Implementing a radical change in a very large organisation goes beyond just deciding "hey, Ive got this really cool idea, lets just format all the hard drives and install Linux".
Even training, each hour the employees are in training is not only costing for the training, but also for lost productivity. The IT support has to be re-trained in the new software.
And on the server side, any code / app migration is no "simple" task.
So no, it is not "extremely simple".
> i don't even understand why there are religious
> parties in a democracy.
Because the demos (the electorate) is(/was)
religious?
-I like my women like I like my tea: green-
While this may be hard for you to understand, most people (strange as it is) could care less what operating system or browser they use. All they know is how to accomplish their jobs using the tools they've been trained on, and most of them spent hours upon hours learning those tools.
As a result, when they turn on their computer and their icons aren't in the same place, many people assume that the machine is broken and conclude that the best option is to call IT and open a trouble ticket.
That said, "babying" them, as you put it, can never be done enough. Switching to Linux isn't that simple for someone who couldn't tell the difference between Intel and InDesign. You can replace the word "Linux" with any operating system, application, or even desktop theme.
A smart IT plan never doubts the inability of the users to -not understand-, and the CTO/IT staff who remembers this keeps his/her job. On the other hand, the unemployed IT staffer "-just does it-, giving them courses for -maybe an hour a day for a couple months-".
Add to the fact that this is a government we're talking about, and taking 3-5 years to migrate an entire city IT infrastructure into as yet uncharted waters is probably being -too-optimistic.
bash-3.00$ uname -a
SunOS panda 5.10 Generic sun4u sparc SUNW,Ultra-2
So the hardware is possibly behind the times, this means a future upgrade will likely include hardware + new version of Linux distro, which may be more prone to introducing compatibility problems than Windows.
So they may also have to recompile some of the software they are using upon upgrading.
It seems to me more like they are sacrificing themselves to help pave the way for others, it is going to be a long time before they start to see the savings.
Saying Java is nice because it works on all OS's is like saying that anal sex is nice because it works on all genders.
While this may be hard for you to understand, most people (strange as it is) could care less what operating system or browser they use. All they know is how to accomplish their jobs using the tools they've been trained on, and most of them spent hours upon hours learning those tools.
Similarly true for graphic designers, photographers and other media folk (music, movie), however they tend to bring an attitude of what they use is best. This is commonly seen with the majority of these people that use Mac systems. Macs aren't bad, but they'll never accept they're not any better, or that they do something wrong. There's a certain elitist attitude, even though they hardly know anything about the system.
Disclaimer: I realize there are exceptions to this, but dealing with the media students at RIT, I can safely say that the demographic here largely supports this observation. So biznitchin!
Presently here, but not there.
As others have pointed out, the CSU is "Christian" in name only.
/.'ers hate religion in general, so it's not a surprise they would hate religious parties. A matter of opinion, I guess.
However, I don't understand why you think that religious parties don't belong in a democracy. People who are serious adherents to a religion tend to feel in a similar way about certain issues (death penalty, abortion come to mind), and thus it makes _sense_ for parties to come together under a religious guise. This does not mean they should be exclusionary, of course, but it's not at all unbelievable that the party would initially form under a religious core.
A religious party does not necessarily mean imposing your religion on everyone else, either. The strict Islamist party won in Turkey, yet Erdrogan hasn't rocked the boat like some people imagined he would. Obviously, in the more liberal European states, the idea of imposing a state religion is even more laughable.
Of course, some
-Erwos
Plausible conjecture should not be misrepresented as proof positive.
Do you mean Microsoft Deutschland Gmbh losing out to IBM Deutschland?
$30m divided by 13,000 machines = $2300/machine? Is this the reasonable cost companies should budget for to migration from Windows to Linux?
It is. Actually it is a bit more expensive than staying with MS for the moment, but the main criteria were stability, security and removal of the dependency of one company only (MS). The move is expected to pay of in the long term (>10 years). Cities are long-term planners, or at least should be.
That is why Balmer failed to convince them to stay with MS by offering better prices. The fundamental motivation was not the current prices, but strategic reasons.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
It sounds as if they're going from a Wintel fat-client/server architecture to a Lintel fat-client/server architecture. Whether or not you agree with me that this is a dubious decision, consider that deploying a true multiuser operating system in effectively single-user mode is a lot like deploying chainsaws to a bunch of chimpanzees.
In my experience *nix's strengths become apparent when you use it as it was meant to be used: a lot of terminals plus maybe a few high-powered standalone workstations. For many standalone machines it's no less of a headache than Windows and in some ways more of one.
I know, I know, thin-clients never took off, yadda-yadda. But I maintain that the biggest part of why they haven't is that deploying Office this way is prohibitively expensive. If you're moving to OO.o, it starts to look a lot better.
(One nice thing about a Linux thin-client setup is that legacy Windows machines can act as terminals with Cygwin/X, allowing Windows and Linux apps as to be deployed in parallel.)
Google confirms: Ruby is the world's most beloved programm
$30m divided by 13,000 machines = $2300/machine? Is this the reasonable cost companies should budget for to migration from Windows to Linux?
It's not that simple. This is an I/T plan covering the operational and support costs of the next several years. Keep in mind that if they didn't decide to do this, they were looking at having to pay, IIRC, some 23M Euro in licensing fees to Microsoft to stay on Windows, plus some hardware upgrade costs. So a better estimate of the Linux migration cost would be something less than 7M Euro divided by 13,000 machines, call it 300 or 400 Euro per machine. That's not so bad when you consider all of the migration labor and the retraining costs.
Whether or not it's worth 7M to get off the Microsoft treadmill is a pretty subjective and speculative question. I'm glad they think so.
Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
You missed the most important fact -- computers (hardware and software) age exponentially faster than the physical counterparts you compare them to. Sidewalks are always useful, so long as they're in good repair. Sure, you may occasionally need to widen a walk to handle more human bandwidth, but in general you could pour some concrete for sidewalks and then leave it alone for decades (but for a periodic cleaning and weeding), and never have a problem.
Try doing the same thing with computers. Go ahead, get some ancient computing hardware from the 70s, 80s, or even early 90s. Install the ancient software. Now try to use that effectively in a technologically-advancing world. Oops! You can't! At least, not as user desktops and such, if you want to keep productive and happy users.
Now let's flip it around. What you're suggesting already exists. How often did you hear about banks and financial institutions using 20-30+ year old software because it still worked during the lead up and fizzle of the whole Y2K crap? And what did those institutions do when a serious threat came around? They started hiring people to fix the current software (patch the sidewalk). Very few chose to upgrade their systems instead (rip up the sidewalk and repave).
Switching from Office 97 (what everybody really uses) to Office 200x is as traumatic as switching to OpenOffice. As Microsoft points out, OpenOffice is comparable to Office 97. And Office 97 is about as good as Office ever got. Beyond that, it's tons of features you don't need, and integration with stuff you don't want to integrate with.
And I've yet to find a manager who understands that REWRITING that fucking piece of ancient shit in Perl or something in five days or whatever is better than trying to keep the stupid thing running forever.
Managers are morons. That simple.
At City College of San Francisco, we run SCT Banner as the MIS system. This thing is a piece of crap from the mainframe days reworked to run on Oracle on HP/UX and PCs. The screens are obviously mainframe crap reworked in Oracle Forms (which is also crap) and the database design is brain dead. The program names are coded like SWOSCHB, the letters of which mean something if you can be bothered to try to find it in the documentation (such as it is).
It's a HUGE system - but it was designed for universities, not community colleges, so large sections of it aren't even used and other stuff must be custom built in house to handle the way a community college works. I found 400,000 records dating back to when the college put the system in sitting in a table which is never used and never cleaned out.
I've offered to rebuild the entire goddamn thing using PostgreSQL on Linux on commodity servers in a couple years for $100K, but nobody takes me seriously. Like it matters how long it would take - the college isn't going anywhere.
The problem is never the technical issues of moving from one system to another - it's the politics and idiot managers - always.
Richard Steven Hack - This sig is TOO GODDAMN SHORT TO DO ANYTHING USEFUL WITH! MORONS!
I will take you up on that bet. I will even tell you that none of those four applications: e-mail, browsing, word processing, and spreadsheets are enough to place a computer on a municipal employee's desk. The government killer app varies by department. But each department most likely has a client server application related to providing city services: Water billing, tax collection, property records, etc. None of these would need the above mentioned 4 apps.
The overall IT scenario is a mix of Telnet/Terminal emulation, Windows desktop apps (VB/VC++/FoxPro/Office VBA), Intranet applications, DOS (yea really!) apps, and maybe even some desktop java. Can you migrate most of these to _Insert OS Distribution Here_? Sure, Why not. That is your decision. However, some groups are going to have a specific piece of commercial software that just will not convert easily or work with VMWare, WINE, or your emulator of choice.
Have you Meta Moderated t
Not only is such a large migration more complicated than you may imagine, but think about the changes that will happen to Free Software/OSS in the intervening period, between now and 2008. By that time, Linux and the GNU accoutrements will be more mature, and will probably be just about ready for desktop use by government officials (who, at least here in the US, are typically anything other than power users). So by delaying, Munich is not only playing it safe, they are gaining a lot of usability that isn't there yet and might not arrive for a while.
That being said, I think a fast migration is perfect for either a small business or a business dominated by tech-savvy employees (e.g. a programming firm). Migrations just need to be tailored to the situation at hand, that's all.
If my answers frighten you, stop asking scary questions.
I've offered to rebuild the entire goddamn thing using PostgreSQL on Linux on commodity servers in a couple years for $100K, but nobody takes me seriously.
I don't blame them.
Haven't seen an office yet where almost everyone doesn't run at least some sort of custom app.
- A zillion Excel spreadsheet macros to be converted to OOo
Whatever their payroll system is.
- Custom reporting in Access out of the Oracle/SQLServer backend needs to be rebuilt
- The city engineers need some new CAD package to manage the sewer sytem. Oh, and all those existing files may need to be converted.
- All of their current Word/Powerpoint files need to be bumped against OOo for compatibility. It's not quite as seamless at it appears.
- All of their current development tools
Converting the one machine in your living room is one thing. Switching a whole business/city is quite another.
> 1,000 Euro will get you a name brand PC with monitor and MS operating system and > Office licenses.
But that does not include installation, administration and infrastructure costs.
> You have to really build up a lot of hatred for a vendor to consider paying
> maybe 14,000,000 Euro over the top to oust said vendor
Certainly you're right, and some people hate MS, but from where I'm standing, most people just hate depending on one thing and paying out the nose for it when they KNOW there are better alternatives. Ask any avid linux user who admins both platforms - if he's worth his salt, he'll tell you that eliminating Microsoft at any cost is not practical or even desirable - but allowing better alternatives in, where they exist, is crucial to saving time, money and aggravation.
Also, you're not considering that those 14,000,000 extra Euro are seen as an investment for long term savings. Corporate licensing for MS is even worse then retail. At least you know in the short run what you're paying for and OTS application - the corporate licensing schemes have totally arbitrary terms (despite what is advertised on the MS site) and have you chasing your tail trying to figure out what you're entitled to. Forget explaining it to a CFO once you've got it nailed down. Then the whole paradigm changes with the next wave of releases. When using mostly free software, you have relative certainty in the form of more-or-less predictable labor and hardware costs - planning is much easier, a little less air needs to be added to the budget, and you have a better chance of staying within budget.
I'm personally counting on this to pressure MS and other big software vendors to either drop prices or increase quality, as well as provide me with the odd chance to roll out something that works instead of dealing with certain packages I know are not worth trying to support.
I think it's turned out OK in cases where MS has been forced to compete. IE was free and improved for a while (at least until they had a dominant lead, IE6+ has been a nightmare) and Exchange got better by version 5 in order to compete with Lotus on the groupware front and sendmail on the MTA front... meanwhile, MS cut server package deals that basically gave it away. Their OS has gotten a bit more stable, probably in response to the perception of Linux kernel as being rock-solid. Maybe MS Office or Citrix will get cheaper faced with the prospect of Linux desktops running centrally-managed open source office productivity software over X.
Factor in the costs of getting locked-in to a single software technology provider. Then you must use Word, Excel, whatever. Factor in the costs of non-interoperability of your files for future revisions of the software. And I mean like on the scale of 30 years. Factor in the cost of lost man-hours due to the Worm of the Day. Factor in the cost of lost man-hours due to viruses. Factor in the cost of constantly rebooting the machines. Factor in the cost of system administration for Windows. Etc. Etc.
There are so many hidden costs when dealing with Microsoft.
Besides, why should Munich buy from Microsoft, when there's a better alternative?
The answer is, the government will spend that money on something else, or (better) leave it with taxpayers so they can spend it on something else. The money will then flow to other jobs, in businesses and industries that are more competitive, where the government should be encouraging capital concentration and job growth.
That answer goes to software publishers, fruit farmers, coal miners, steel makers, missile manufacturers, and any other interest that thinks it should be paid, not for the value of its goods or services, but because such a fat pig is entitled to its place at the public trough. Off to slaughter, piggies!
or a tremendous PR disaster. If they fail, you can bet your ass MSFT will not get tired pointing this out. If they succeed, Novel/IBM/RHAT and everyone else will be touting the precedent.
If you view the migration away from MS-Windows to anything else to be inevitable, then the migration costs should be largely accounted as removing-MS-Windows costs rather than buying-Linux costs. In which case Linux costs an awful lot less than MS-Windows.
You also have to figure in the ongoing cost of maintenance, along with a number of so-called "extraordinary" items like cleaning up after the next CodeRed or MSBlast hits you. Linux is extremely unlikely to ever raise such costs.
But the big reason is that Germany really, really hates being run by foreigners, particularly Americans (but they have other pet peeves too), in any way.
Got time? Spend some of it coding or testing
don't forget, significant part of these 13M will stay in germany, may help to create new jobs there, etc ...
so paying 7M for windows can be more expensive (for germany at least) than paying 13M for linux
that a larger project in Extremadura, Spain, doesn't get this kind of attention (Some background for the spanish-impaired). It's already working (I thinks it's a little over two years now), it's been distributed to hundreds of thousands (including every desktop in the schools, one computer for every two students, mind you)... it even has inspired at least one already working project in Andalucía, Spain (and seeds of several others, as in Madrid, Zaragoza or Valencia; it seems all education in Spain is migrating to linux in the next few years).
My journal. Mainly about freedom.