Slashdot Mirror
Linux in Munich Followup
Rican writes "Wired has a story that details some of the difficulties that Project LiMux seems to be experiencing in Munich. Including financial and technical issues. On the positive side it looks like despite these setbacks they are continuing with the project and have a positive attitude about its completion. Let's keep our fingers crossed and do what we can to support this monumental effort that will benefit the whole Open Source Community."
If there is one place that will be the turning point for OSS and Linux, it will be Europe and Asia. Wait: I guess that isn't one place! :)
MOUNT TAPE U1439 ON B3, NO RING
Interesting that all the problems are interoperability with proprietary software. There haven't been any problems with the people using linux.
"Right now we are proceeding as planned, and we have no hints or signals that the city counsel is regretting or reconsidering their decision to move to Linux,"
Sounds like the normal hitches you'd expect when doing any large-scale migration. Something more detailed would have been nice rather than generalities about 'software compatibility' and 'security'.
--
This sig is inoffensive.
..i dont think there is the need to start making up big stories out of this. Of course this kind of migration takes a lot of time, specially for the training.
I really hope that everything proceeds as planned.. a project like this is important for the public opinion of Free Software and Linux...
I think part of the problem with migration is that in many instances many people who use linux and love did it because they were disenchanted with proprietary OSes for personal reasons, and these guys are trying to migrate for a multitude of reasons, including monetary ones. Add on to that the fact you're retraining thousands of people, and you've got one heck of a mess on your hands.
Nonetheless, hopefully they persevere.
Wired has a story that details some of the difficulties that Project LiMux seems to be experiencing in Munich. Including financial and technical issues.
What this experiment will have to do is prove that Linux can do it for less money and be more efficient than proprietary solutions such as Windows.
Studies on open-source security, desktop ergonomics and the software components' stability and compatibility with other applications will be included in the process.
For my money, I would have bet on OS X providing a better system from these perspectives.
IBM and Germany-based Linux distributor SuSE are expected to help offset the costs of the migration by supplying technical support and conducting some of the studies that the Munich city council has requested.
This will most likely be of huge importance in maintaining this transition, but more support may be needed in bringing custom applications from Windows to Linux.
Visit Jonesblog and say hello.
So migrating all the systems from Win98 and Windows NT to Windows 2003 Server and WinXP desktops would have run within schedule and under budget? I think not. This is an IT project for god's sake, they never get completed on time and within budget. You will always strike issues and problems once you get into the thick of it.
Why all the fuss? Why any fuss? Sounds like business as usual to me.
Jedidiah
Craft Beer Programming T-shirts
But instead of paying $23.7 million for the Microsoft solution, Munich's city council opted to spend roughly $35.7 million to switch to open source, saying that the higher price would be offset by lower costs and more flexibility in licensing fees and software choices over the long run.
That's an initial difference of $12.5 million, or $892 per system. They say that they'll make it up in the long run, but they must really mean long term.
I love linux as much as the next guy, but the duty of this agency is to provide the cheapest solution to server their constinuents. This whole project smells to me like the council had a beef with Microsoft, and allowed that bias to lead to a poor business decision. Now they're trying to justify it so they don't have to admit that they made a mistake.
There are probably organizations that are ready for a wholesale ms to linux migration, but this doesn't look to be one. All of their staff have to be retrained, the price is more expensive, and a considerable custom windows application base seem to make this a bad idea. The linux community can only be hurt by a square peg linux solution being shoved into a circle microsoft hole.
Microsoft is loving this, and preparing marketing material right now that shows that replacing Microsoft with Linux doesn't make business sense. Business realities, not propoganda, should dictate migration to linux. Forcing the issue (as it appears they are doing) only hurts linux in the long run.
Here in Canada, most government departments and agencys are either undertaking, or considering the migration from Windows NT 4.0 and Windows 95 to Windows 2000. I once added up the total cost of Windows NT 4 Workstation licences, NT 4 Server licenses, NT 4 CALs, and MS Office 2000 licenses for the government agency that I worked for (~20000 pcs served by ~500 NT 4 servers) and the figure wasn't pretty. A signifigant portion of the operating costs for many government offices goes to buying access to Microsoft's IP; I would like to see my tax dollars used in a more productive manner. Kudos to Munich and best wishes that their problems will be overcome shortly.
An eye for an eye... leaves the whole world blind.
The OSS movement can't affort to blow this. It's, IMO, that important to demonstrate to the city of Munich that they, without a doubt, made the right decision. Set a precident.
The fun and games---if they weren't over before, they are now.
So migrating a large governemtal user base from Windows to Linux is experiencing some difficulties. Gee...whodathunk? /. fanbois predicted an easy transition. "Just install Linux! All your problems will be solved!"
Only the (very vocal)
The real outcome will be in 3 or 4 years. Everything will have been transitioned and shakendown. Then, and only then can we see if has actually been beneficial or not. We can see the parts where a unified Linux base has worked, and where it has failed.
Volunteer to man an IRC channel.
Write some documentation.
Pretty up some already written documentation.
Answer questions on the newsgroups without griping or insulting people.
If you are German write the politicians praising them for their courage in choosing this solution and vote for them in the next election.
If you are not German then write to them anyway and see if it's legal for you to send them some money. Even five or ten dollars would be a highly effective symbolic gesture.
War is necrophilia.
Money seems to be the main problem here, and I think when things come to public administration money is not the most important thing.
:)
The most important thing to a public administration should be citizens, and there are a few issues that propietary software doesn't do for citizens. First of all, doesn't let to audit the code, and this is a problem when you manage sensible information. Another thing that Microsoft don't does for Munich citizens is promoting local software industry, here in Europe, apart from free vs. propietary software we have another problem. Import vs. local development. Linux es the only horse we can ride. And it's from european origin, also
More money, maybe, but why spend this money in a foreign industry when you can spend it developing the local software industry?
DON'T PANIC
--Mark
"It is nice to know that the computer understands the problem. But I would like to understand it too." --Eugene Wigner
This IS WHAT CAPITALISM IS ALL ABOUT -- COMPETITION. I don't care whose product we use, I just want it to work, and I want it to be fairly priced. And guess what -- we (the linux community) are the sole reason for getting microsoft to drop their prices. We are now COMPETITIVE. This is huge. This is what capitalism is all about. MS can't strongarm every company in the world for too much longer. You can now threaten them with the big L word or the big blue.
This is incredible. Others are going to see that MS can have lower prices, and we can now negotiate with the devil.
Berto
The goal of governments is not to do anything as cheap as possible, but to do the right thing.
The right thing, IMHO, is that no state may make itself dependant on a single external (commercial) entity. I would say, no matter what the cost, it is the obligation of any decent government to free themselves of any strangleholds that may exist.
As a side effect, it will be much much cheaper in the long term. Someone must make a first step; after that, the compatability issues shall diminish, and others to follow (e.g. other cities) shall have less problems. Once we are saved from the dictate of proprietary file formats (getting rid of which does cost some money initially) the savings are enormous.
Contrary to quoted companies, states do not only have to look at next quarters financial, i.e. be extremely short sighted, but have to think on timescales of up to 50 years.
The attempt by Microsoft to cut prices for one customer (preditory pricing against a competitor) is definitely anti-capitalist, since in a free-market situation you have to sell each product for the same price to all customers. Otherwise the system doesn't work efficiently and you end up with market failures such as monopoly.
In basic economics this is Microsoft attempting to eat into the consumer-surplus. It is anti-competitive and hence anti-capitalist.
You are right in that Linux is now competiting with Microsoft enough that they are having to lower their prices, which is a good thing. However, Microsofts response simply negates any competition Linux may pose, since they can lower prices for the few individuals who do threaten to swap until they promise not to and buy into the current lock-in situation. We'd have to wait until 10% of Microsoft's customers threaten to swap simultaniously for it to effect Microsofts bottom line.
and found I'd better clear things up a bit.
My point was that in Germany we have had a recent tendency to rush things. There are numoerous examples, the launch of the UMTS-Network debacle (highspeed mobile network), the Transrapid (a magnetic monorail) fiasco in China and, worst of all, the tollcollect (a sattelite based system to charge on traffic) desaster. All of these projects were ambitious and technically challenging and all of them have a long history of failure and mismanagement. I really love this city and I appreciate the brave decision of the city council to try the switch. But I have this fear that they will blow it and Munich will not be known as the "Linux-Capitol" but rather as the "bad example". Keep your fingers crossed that I am wrong!!
btw: here's a link to the english page of muenchen.de.
Good night.
Lispy
Despite the wet dreams of the pointy-hairs the best way to do a large system migration is not to make a project out of it but to set it as an objective.
I understand peoples fear of uncertainty and their inclination towards organizing everything to avoid "chaos" but making a project of that scale is really just a nice way of deluding yourself. It will be chaos regardless.
If you want to migrate a disrtibuted organization of 14,000 desktops and unkown amount of servers from one operating system to another you do it by setting an enterprise standard and then knocking it out one project, system, or group at a time. Hell even microsoft didnt do a mass migration from their old unix desktops to their own operating system en masse, they migrated slowly where it made sense and pushed the remainder.
I will not be suprised if this project partially fails.
But according to Computerwoche and other reports, the city lacks the funds to invest in the planned testing and development of an open-source solution.
They've got about $35 million budgeted for this migration, and they're out of money at the 'testing and development' phase? How did they come up with their replacement cost figures without doing some 'testing'?
One man's -1 Flamebait is another man's +5 Funny.
All the damn time I'm reading "Linux isn't ready for the desktop".
I'musing Linux on the desktop.
It works fine. Does this mean I'm doing something wrong? I mean, it's not meant to work, is it?
Linux is ready for the desktop, as shown by the countless people using it on the desktop. Hell, you're at Slashdot, look around you...
The real issue is that Linux isn't ready for mass consumption in the same way that Windows is, but that's largely because people have had Microsoft stuffed down their throats for... ooh... 20 years? Something like that. People have issues with Linux because they're so used to doing everything the Microsoft way. People are taught that computers run Windows, and for 90% of people there is no other OS. Most people wouldn't know what an OS even is! It's inconveivable for computers to run any other way.
There are distro's combatting this though. I mean, first up there's obvious candidate's like Lindows; but things like Redhat (Fedora, whatever) and Mandrake are getting extremely user friendly. For the average end user a decent RedHat install will do most things - e-mail, internet, Office type stuff. It's only the hardcore minority among us that need more complex stuff and thus spend sleepless nights tweaking the kernel, trying to get Wine to run properly, trawling the net for those obscure drivers... But most people never even think about that type of stuff
Long story short - Linux is ready for the desktop, but people just need a bit of re-education to get their head round it, same as a lot of people need education to use Windows in the first place. A lot of end users need training to use Windows in the first place, and then need training with each new version of Windows, so training in Linux shouldn't be too much harder.
Here endeth the lesson.Linux is a pretty tough change. Gradual migration is necessary for continuty of operation, so you must work with interoperability.
I have set up in the past a Linux intranet servce in a Windows IT environment, precisely because I thought that it would be "better" in the long run to work with Tomcat/Apache/Linux than WebLogic/WebSphere/Windows.
The basic set-up was very easy, as always. But soon we got into things related to security & authorizations, for which we needed to interface with Active Directory... I'm glad we had some time to do this right. "Active Directory access from Java over LDAP with Kerberos authentication to a Win2K domain controller" is very sparsely and partially documented, and then what you can find on the net relates to earlier versions of software, other distros... A lot of testing and trying with some very good people on the task, and finally we got it off the ground. It's not the kiddie-script grade stuff or burn-a-distro-and-enjoy story we like to hear.
This was in a top-tier R&D lab, with research-grade time on our hands. Basically our core business. Now I'm not sure the municipal office of Munich can do these kinds of things themselves... And if they hire consultants to do it for them, you can be pretty sure they'll take advantage of their unawareness.
In that particular case, we got for a short period of time a MS Consulting dude to help us. The poor guy knew less about MS' own products than us! Now imagine the same guy "advising" the Munich city office on how to better interoperate with Microsoft's products.
Here's the timeline
1- City of Munich goes Linux
2- City of Munich realizes during the migration it will need to interoperate with Windows
3- City of Munich to MS: one more thing, guys... before we ditch you, how do we interoperate with your products exactly? No, not for us (we don't do, we make others do), but we need to tell our IT subcontractors.
4- IT subcontractors blame bad interop on MS, who blames it back on these sloppy-Linux-hackers--and meanwhile, computers kaputt
5- ???
6- One good "TCO" sell point for MS?
Over 75% of all IT projects fail totally, is late or over budget. Most of them are not Linux projects.
So, Microsoft is in no position to gloat over any Linux setbacks in Munich.
This is a problem that probably has more to do with leadership, management methods than with any specific technology like Linux or Windows.
Even if the up front cost for switching to Linux turns out to be higher than expected, it will probably be cheaper in the long run. Using Linux they will be in much better control of their future upgrade costs.
Not to mention that, money spent on Linux stays in the local economy instead of feeding a foreign company.
God is REAL! Unless explicitly declared INTEGER
But you forget that they have the support contract from SuSE and IBM. These two companies knows the high profile of Munich and I am confident they won't let it slip. Even if the Munich IT people are not the top of the crop, they can call IBM and SuSE when things get tricky, two reliable computer companies that I would have much more confidence in when making a support call than MS any day.
--- guns don't kill people, people with guns kill people ---
The biggest problem isn't the initial Linux installs, it's that every middle manager has their own pet software usually Excel sheets with crappy macros, or commercial canned Access programs....All that stuff is most difficult to get working. Sure you can use Wine, but the point is that you don't know till you get there what they were using. On top of that Mid-managers espically are resistant to change...or redoing their work.[unless of course they deem it ememgency!] even if it saves the department money in the long run!
Germany is right in the middle of this. Strategicly the OSS/Linux vendors are in for a bonus few years.
You can guarantee that IBM, Novell and SuSE will all want to make this work out fine because they would then sell this solution on.
You have to ask youself - why spend hard earned EU taxpayer money on Microsoft licenses when it can be spent on local service providers integrating freely available OSS software ?. IBM Global Service are asking that same question - pity SCO are not getting stuck in too as their traditional strength in vertical integration would be perfect for the new EU countries, but now no-one will touch them with a barge pole.
To my considerable surprise, many people seem to think that a free software/open source solution will be or will have to be cheaper than the proprietary solutions. That's a misconception.
Free software doesn't compete with proprietary software on cost, reliability, performance or features. It's not even really appropriate to use the word compete, since the metric of success is so different for free software than for proprietary products. A proprietary product succeeds when it makes money. But when does free software succeed? When it attracts a lot of users, maybe. But what does that accomplish?
Well, it liberates users. The goal of proprietary software is to make money. The goal of free software is to liberate users. Not to make cheaper software, or even better software (although we try), but to make users more free. Any other motive just does not survive contact with reality. Nobody starts their project with the idea of making a cheaper widget. Most people start their project with the idea of making a better widget, but this seems to be a very elusive and highly subjective quality. The truth is we don't know how to make "better" software any more than the proprietary people do.
The benefit of free software is not that it's cheaper or better, since these are after the fact rationalizations at best and misleading fantasies at worst. The benefit is that it frees users from mindlessly protectionist policy, draconian restrictions on use and distribution, and a whole universe of demeaning do-not-trust-the-customer attitudes.
This freedom might be more expensive than the alternative, or it might be less expensive in the long run, or it might lead to better software, or not: these are issues that are infinitely malleable in the hands of a skilled advocate. But it all derives from the motive of freedom -- not the mere availability of source code.
With short termism prevalent in the corporate world, with people achieving success by meeting expected profits per quarter, and failed if expectations are not met (hint, profiuts were still made), it comes as no surprise that people can only measure success of a project merely on TCO or migration costs. The brainwashing of powerful IT companies is working wonderfully.
I want to be the owner of my IT infrastructure, both at home and at work (that is my compny or employer), I want to make the decisions of how, when and who does maintenance to my IT infrastrucutre, I don;t want to migrate or upgrade because it is convenient to the provider and when the provider asks me to upgrade or migrate I want to be able to shopr around for possible options.
I could not care less if going OSS will cost me a few bucks more in the short term if in the long term I will regain the control of my IT infrastrucutre which should be reflected in saner decision when spending money.
Shortermists should jump from a clift and marvel at the tremendous speed they achieve just one second before they come to a complete halt once they hit the floor.
IANAL but write like a drunk one.