Stories
Slash Boxes
Comments

SoylentNews is people

posted by martyb on Friday January 22 2016, @09:52AM   Printer-friendly
from the lock-in-is-expensive dept.

Munich still uses 41 proprietary apps that will only run under XP or 2000. The city has estimated it will cost $18M to replace them over a 4-year span.

Nick Heath at TechRepublic reports

Windows XP and 2000 are used by fewer than 1,500 of the more than 16,000 staff at the council, which relies on the aged Microsoft systems to run 41 applications.

[...] In order to stop using Windows XP and 2000, these 41 applications will either be migrated to a newer, supported operating system, replaced with more modern software, or phased out--as part of a four year project costing €16.6M ($18.03M).

[...] Munich carried on using XP and 2000 due to these 41 applications being used for crucial work in the city, from monitoring emissions for air pollution to flood protection.

To secure the OSes, Munich ran them on virtual machines and on standalone computers, as well as using what it calls "restrictive data interchange", quarantine systems, and additional protective measures.

The council has decided to stop using these older unsupported versions of Windows now as, not only are they a security risk, but according to a report [PDF, Deutsch] they have limited support for network and data security features the council wants to use.

[...] Often it can be the case that organisations can't update the application to run on a newer OS because the people with the necessary skills are gone or the company that originally wrote the software no longer exists.

[...] The project at Munich will be split into two phases: The first will assess the work needed and the second will carry it out. Work got underway at the end of [2015] and is expected to be complete by the end of September 2019.


Original Submission

Related Stories

India's Defense Services Are Switching to GNU/Linux 12 comments

The site It's FOSS is reporting that India's Defense Services are switching to GNU/Linux, ditching an insecure legacy operating system, with an August 15 deadline. Little is known about their home spun distro except that it seems to be based on Ubuntu.

What's Happening: According to a recent report, the Defence Ministry of India has decided to replace Windows with an in-house developed Linux distro called 'Maya' on all computers that are connected to the Internet.

Also reported at The Hindu, Defence Ministry to switch to locally built OS Maya amid threats, which explains that this move is a reaction to increasingly successful attacks against a certain, pervasive, desktop legacy operating system. x

Currently, Maya is being installed only in Defence Ministry systems and not on computers connected to the networks of the three Services. On this, the official said the three Services had also vetted it and would adopt it on service networks as well soon. The Navy had already cleared it and the Army and the Air Force were currently evaluating it, the official added.

Maya was developed by government development agencies within six months, the official said. Maya would prevent malware attacks and other cyberattacks which had seen a steep increase, the official noted.

However, the attacks in and of themselves are less of a problem than the fact that a large, and increasing, number of them are successful against that aging legacy desktop operating system.

For India to pull this off successfully, they must study how their opponent has maneuvered over the years against GNU/Linux deployments and in particular look at case studies like Kerala, Munich, Lower Saxony, Vaasa, and Turku. India's opponent in this move has had many programmes, years ago one was EDGI, and a long standing mandate that "under NO circumstances lose against Linux".

Previously:
(2018) German Documentary on Relations Between Microsoft and Public Administration Now Available in English
(2018) German State of Lower Saxony Plans to Switch From Linux to Windows
(2017) Munich Switching From Linux to Windows 10
(2017) Linux Champion Munich Takes Decisive Step Towards Returning to Windows
(2016) Draft Report Doesn't Say -Which- Software is Causing Problems in Munich
(2016) Munich: The High Cost of Having Committed to Closed-Source Software
(2014) Another German Town Says It Has Completed Its Switch To FOSS


Original Submission

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 0, Insightful) by Anonymous Coward on Friday January 22 2016, @10:00AM

    by Anonymous Coward on Friday January 22 2016, @10:00AM (#293056)

    But not BSD, and not GNU, and not any other open-source software.

    Only Linux.

    Mod me Insightful.

  • (Score: 0) by Anonymous Coward on Friday January 22 2016, @10:09AM

    by Anonymous Coward on Friday January 22 2016, @10:09AM (#293057)

    It's all awesome, so, whatever.

  • (Score: 5, Insightful) by canopic jug on Friday January 22 2016, @10:24AM

    by canopic jug (3949) Subscriber Badge on Friday January 22 2016, @10:24AM (#293061) Journal

    One of the ways that M$ has been able to squeak by in procurement is that they've avoided counting the exit costs in their calculations of total cost of ownership (TCO). All systems go away eventually. All systems will cost something to decommission. Thus it is essential that exit costs be included in TCO [europa.eu]. With crap designed for maximal lock-in, like anything from M$, the transition costs are going to be tremendous. Now that Munich is using FOSS and open standards, the transition to any other system will have a much lower cost, in addition to lower ongoing maintenance costs.

    --
    Money is not free speech. Elections should not be auctions.
    • (Score: 3, Insightful) by Immerman on Friday January 22 2016, @04:36PM

      by Immerman (3985) on Friday January 22 2016, @04:36PM (#293188)

      To be fair though, there's exit costs for any OS. Good luck getting binaries from one linux distro to run on another from a different "family", or even the same family several years later. It's hit or miss even getting source code to compile without modifications, assuming you have access to it at all.

      • (Score: 1, Troll) by kurenai.tsubasa on Friday January 22 2016, @04:58PM

        by kurenai.tsubasa (5227) on Friday January 22 2016, @04:58PM (#293201) Journal

        This is why it's important to have all the source. The binaries are merely a convenient way to get the computer to go, sort of an intermediate step in 1.) perform analysis 2.) instruct the computer to perform that with which we are now acquainted 3.) binaries 4.) the computer goes.

        The real danger of closed source software is that others will just flat-out assume that you can do things to it you can't without breaking the EULA and probably DMCA and TISA when that one's ratified.

        This can lead to the determination that you're just refusing to do what a woman is asking you to do because you hate women. Your attempts to explain why you're not going to reverse engineer the piece of shit and maintain a binary patch will be seen as mansplaining. This is why I swear I will never ever accept another job where I have to touch anything closed source again. Then again, I'll probably just be content to flip burgers, since nobody's called me sexist for making a good burger. Besides, when I make a burger, I have open access to everything I need from the beef (or bison/turkey/etc), condiments, lettuce, buns, bacon++, etc.

        (Fine, mod off topic, I don't care.)

        • (Score: 0) by Anonymous Coward on Friday January 22 2016, @05:52PM

          by Anonymous Coward on Friday January 22 2016, @05:52PM (#293236)

          (Fine, mod off topic, I don't care.)

          Off topic? No, not really.
          That was a nice comparison, yet it made me hungry.

          --
          “War… war never changes.”
          https://www.youtube.com/watch?v=vE25wHQPHIE [youtube.com]

      • (Score: 0) by Anonymous Coward on Friday January 22 2016, @08:04PM

        by Anonymous Coward on Friday January 22 2016, @08:04PM (#293298)
        You may have access to the source, but if you're not a good enough coder and nobody else is interested in figuring out why some long dead company's crappy driver doesn't work with the latest kernel, you still lose hardware support.

        Whereas with Windows XP you could have the same driver working for 15 years and many executables for even long (yeah some of them might be malware but hey backward compatibility ;) ). Of course Microsoft is changing that now - they're very aggressively trying to force people to Windows 10.

        Sad times ahead, since the Desktop Linux bunch have a track record of making their stuff even worse whenever Microsoft makes their stuff crappier. Breaking/fragmenting GUIs, breaking sound, picking stupid defaults (Microsoft picks stupid defaults too but they're the incumbent monopoly).
        • (Score: 1, Informative) by Anonymous Coward on Saturday January 23 2016, @04:05AM

          by Anonymous Coward on Saturday January 23 2016, @04:05AM (#293477)

          The Linux Driver Project has hundreds of guys waiting in line to make your stuff Linux-compatible.
          Look for the number 300. [googleusercontent.com] (orig) [lwn.net]

          Give them a sample of the stuff you want to be compatible with Linux and they'll have it figured out in no time.

          with Windows XP [...] yeah some of them might be malware

          I'm sold. Where do I sign up? 8-)

          the Desktop Linux bunch have a track record of making their stuff even worse

          Not familiar with Trinity (fork of KDE 3) or MATE (fork of GNOME 2) or Cinnamon (fork of GNOME Shell)?

          Hope you got all the bile out of your system.
          I'd say that mostly you are shooting blanks.

          -- OriginalOwner_ [soylentnews.org]

          • (Score: 0) by Anonymous Coward on Saturday January 23 2016, @07:31AM

            by Anonymous Coward on Saturday January 23 2016, @07:31AM (#293536)

            the Desktop Linux bunch have a track record of making their stuff even worse

            Not familiar with Trinity (fork of KDE 3) or MATE (fork of GNOME 2) or Cinnamon (fork of GNOME Shell)?

            Hope you got all the bile out of your system.
            I'd say that mostly you are shooting blanks.

            Are you that retarded? Why the fuck do you think they forked in the first place?

          • (Score: 0) by Anonymous Coward on Saturday January 23 2016, @10:41AM

            by Anonymous Coward on Saturday January 23 2016, @10:41AM (#293584)

            The Linux Driver Project has hundreds of guys waiting in line to make your stuff Linux-compatible.
            Look for the number 300. (orig)

            Give them a sample of the stuff you want to be compatible with Linux and they'll have it figured out in no time.

            Waiting "in line" and "in no time" huh? So where are those non crappy working full featured 3D drivers for all AMD/ATI and NVidia's video cards?

            Stop drinking too much of the kool-aid?

            With Linux's approach whenever they break compatibility there is a cost to everyone else - whether it be $$$ or time contacting those hundreds of people you mentioned, or those hundreds themselves. Someone has to do additional work. Remember if the hardware company is long dead you may not get any documentation or source code.

            With Windows those drivers don't need to be changed. So fewer required to notice it it's not working anymore, fewer required to figure a way to get it working again. There's no need to realize that the company in Taiwan that used to sell the hardware is long gone and isn't going to help you or "the hundreds" to write that driver.

            But Microsoft's latest strategy with Windows seems to be towards a different path: http://www.computerworld.com/article/3023533/microsoft-windows/microsoft-support-windows-10-new-hardware-itbwcw.html [computerworld.com]

            Maybe a few large corps that need to stick to older versions of Windows should get together and make threatening noises about sponsoring something like ReactOS ;).

            • (Score: 1, Insightful) by Anonymous Coward on Saturday January 23 2016, @11:48PM

              by Anonymous Coward on Saturday January 23 2016, @11:48PM (#293753)

              It's obvious that you're talking about closed-source device drivers for crappy hardware that YOU bought from a crappy manufacturer who has crappy support.

              It is abundantly clear whom you -should- be blaming--but aren't.

              -- OriginalOwner_ [soylentnews.org]

              • (Score: 0) by Anonymous Coward on Sunday January 24 2016, @10:05AM

                by Anonymous Coward on Sunday January 24 2016, @10:05AM (#293885)

                You're the one who claimed:

                The Linux Driver Project has hundreds of guys waiting in line to make your stuff Linux-compatible.

                Thus it's more accurate to say they're only waiting in line to make certain things Linux-compatible and not other stuff.

                Thus you will lose support if your stuff doesn't happen to be one of those "certain things" when the kernel devs break compatibility.

                You can play the blame game all you want. The fact is even if AMD goes bust and stops supporting their drivers, you'd be able to use their video cards on Windows for a longer time than on Linux.

                That said nowadays video cards seem to die within 3 years, so won't really be a big problem in practice, since you should be budgeting for a new card...

                • (Score: 0) by Anonymous Coward on Sunday January 24 2016, @07:53PM

                  by Anonymous Coward on Sunday January 24 2016, @07:53PM (#294051)

                  When the software you use is open, you might not even need the big boys to solve your problem.
                  Why Software Openness Is Important [googleusercontent.com] (orig) [goodbyemicrosoft.net]
                  ...and, as he notes there:

                  If this were a Windows PC, I'd be stuck waiting for the manufacturer to issue an updated driver--not likely for my old video card!

                  -- OriginalOwner_ [soylentnews.org]

      • (Score: 2) by darkfeline on Saturday January 23 2016, @12:53AM

        by darkfeline (1030) on Saturday January 23 2016, @12:53AM (#293417) Homepage

        A Linux compiled binary will run on any and all Linux distros, and will even run on BSD as BSD has Linux emulation (within reason, no guarantees for a binary compiled for Linux 0.11, or for a modified Linux kernel).

        Now, of course there are packaging issues or dynamic library linking issues, but those exist even on the same OS distribution across installations.

        --
        Join the SDF Public Access UNIX System today!
      • (Score: 2) by canopic jug on Saturday January 23 2016, @09:49AM

        by canopic jug (3949) Subscriber Badge on Saturday January 23 2016, @09:49AM (#293576) Journal

        Good luck getting binaries from one linux distro to run on another from a different "family", or even the same family several years later.

        The binaries are irrelevant. It is the source that matters with Linux distros. That is the currency to track, and what enables portability and endurance. However, to take it a step more abstract, it is really FOSS's use of open standards for data and protocols that really reduce long term costs. You don't work with programs, they come and go, too, as much as you use the programs to work with data. It is the data that sticks around the longest and open standards enable independence from individual vendors or programs. The article blames "old applications", but I wager that it's really access to the data that these applications work with that is the bottleneck.

        --
        Money is not free speech. Elections should not be auctions.
        • (Score: 2) by Immerman on Sunday January 24 2016, @07:44PM

          by Immerman (3985) on Sunday January 24 2016, @07:44PM (#294047)

          That's great if you're using open source or internally developed software -but there's no guarantee that the software you need/want to use is available as such.

  • (Score: 3, Insightful) by Beige on Friday January 22 2016, @11:18AM

    by Beige (3989) on Friday January 22 2016, @11:18AM (#293064) Homepage

    Issues pertaining to legacy systems are by no means unique to closed source software.

    Since the "apps" only run on XP or 2000 I'm assuming they are 16 bit Dos programs. They were likely developed long before Windows XP came out, so it's plausible the city got 20+ years of use out of their initial investment. This is actually not too bad from an overall TCO perspective and it's hard to imagine the situation being much different had the software been written for some ancient Unix using some now-obsolete libraries, and data files with endianess issues etc. They'd be stuck getting new hardware, new software, new support contracts etc. either way.

    I'm actually curious about the premise of the headline. If the city went and bought new software which was e.g. built with Node JS, a HTML5 browser UI or whatever, could they realistically expect to get another 20 years out of that investment without having to pay developers to "upgrade" the software every few years just to keep it working?

    • (Score: 0) by Anonymous Coward on Friday January 22 2016, @11:41AM

      by Anonymous Coward on Friday January 22 2016, @11:41AM (#293067)

      I'm assuming you're wrong.
      Never heard of FreeDOS, DOSBox, DOSEMU (all FOSS)?
      Should be duck soup to get DOS apps running without any MICROS~1 code.

      -- OriginalOwner_ [soylentnews.org]

      • (Score: 2) by Beige on Friday January 22 2016, @09:58PM

        by Beige (3989) on Friday January 22 2016, @09:58PM (#293347) Homepage

        "Should be duck soup" is easy to say if you've not actually done that much legacy system porting :-) Things like legacy parallel port copy protection dongles, legacy hardware (e.g. barcode readers intended for AT or XT keyboard connectors), ancient network drive setups etc can pose interesting problems on modern OSes even if you can spin up the 16 bit client in an emulator. A lot of tweaking, some reverse engineering and even reassembly is usually needed to work around such issues, and it's frankly not worth the effort in most cases. However, there's always the odd "company is fubar unless we get this system from the 80s back online"-type emergency.

        • (Score: 0) by Anonymous Coward on Friday January 22 2016, @11:23PM

          by Anonymous Coward on Friday January 22 2016, @11:23PM (#293380)

          you've not actually done that much legacy system porting

          If -zero- qualifies as "not that much", you're right on the money.
          OTOH, I haven't heard of any problems taking apps that run under MICROS~1's DOS and running those under FreeDOS.
          In fact, I read that FreeDOS can get USB going where MICROS~1's DOS never did.

          dongles

          Dongleware is another area where I'm completely ignorant.
          I never saw any that I thought was worth the asking price (in bucks or in freedom).
          I remember a narrative years ago in sci.electronics.design by a guy who had a stack of dongles out the back of his PC.
          The cleaning crew moved his computer and CRUNCH; no usable stuff for a while.

          it's frankly not worth the effort in most cases

          We need to hear that more often--especially WRT closed proprietary stuff.
          Start fresh and make it open this time.

          -- OriginalOwner_ [soylentnews.org]

    • (Score: 0) by Anonymous Coward on Friday January 22 2016, @12:10PM

      by Anonymous Coward on Friday January 22 2016, @12:10PM (#293075)

      The key to legacy systems is to ensure the system/data/network requirements for them are well documented and reviewed often enough (and by enough people). Any software can be re-written - even if from scratch - with enough resources and project commitment (which Munich appears to have). But ...

      ... sometimes the devices used (e.g., for data gathering) no longer have drivers that are supported on current/newer/less-archaic platforms. In these instances outsource the development/re-engineering of the drivers to a specialty firm (but always retain rights to requirement & development documents, and the source code). The time and cost associated to hiring the right staff (with the specific skill set), bringing them up to speed, and developing/testing the replacement drivers is not practical. Outsourcing will look like it costs more up front, but you don't want someone who isn't a "driver person" developing a driver for you.

    • (Score: 4, Insightful) by GreatAuntAnesthesia on Friday January 22 2016, @12:15PM

      by GreatAuntAnesthesia (3275) on Friday January 22 2016, @12:15PM (#293077) Journal

      > They'd be stuck getting new hardware, new software, new support contracts etc. either way.

      Yeah, but at least if they have the source they can shop around for contractors:

      If the source is held by the people who made the tool then the client can only go to the provider, who is then in a position to set whatever outrageous prices they like. They do this because (a) they can and (b) they'd much rather be working on some big new project that futzing around trying to patch that over-priced rickety heap of shit they just barely got duct-taped together and out the door 2 years ago. So they jack the prices up to the point where the client decides to make do with workarounds and put the necessary work off for years or decades until they really an't ignore it any more. By then the people who knew the product are all retired / dead / living off grid in a yurt in Nepal and the client has no choice but to have the whole damn thing rewritten from scratch (usually by the same providers) which will of course cost an arm and leg because every big CEO knows that .gov / .mil contracts go over-schedule and over-budget and the contractors laugh all the way to the bank, and the whole cycle repeats.

      With open source you can make small, periodic upgrades as necessary. You can change contractors at any time, because nobody is holding your source code and data hostage, which means you acn keep control of prices.

      • (Score: 0) by Anonymous Coward on Friday January 22 2016, @04:57PM

        by Anonymous Coward on Friday January 22 2016, @04:57PM (#293200)

        The city could have had a term in their contact saying they'd get the source code along with every release. That doesn't make the code open source but it would let them shop around. Far too many people push open source for it's own sake. They stick fingers in their ears and yell open-source, open-source over and over again, like the article title. Everything that's happening with Munich right now could still be happening even if the code was freely viewable online. Their issues have nothing to do with the software being closed-source.

        • (Score: 0) by Anonymous Coward on Friday January 22 2016, @07:22PM

          by Anonymous Coward on Friday January 22 2016, @07:22PM (#293279)

          freely viewable

          What Munich is going for is CONTROL of their software ecosystem.
          You don't get that with "viewable".
          What they need is not some EULA-restricted[1] openwashing, but rather license terms that give them **actual** freedom. [googleusercontent.com] (orig) [gnu.org]

          [1] See the recent dust-up with Remix OS.

          -- OriginalOwner_ [soylentnews.org]

    • (Score: 2) by requerdanos on Friday January 22 2016, @01:13PM

      by requerdanos (5997) Subscriber Badge on Friday January 22 2016, @01:13PM (#293092) Journal

      Issues pertaining to legacy systems are by no means unique to closed source software.

      This. I worked a contract in the late 1990s for a major manufacturer [cummins.com], porting some of their legacy in-house data acquisition and reporting software at one of their U.S. factories from 16-bit to 32-bit. They were all Visual Studio projects under Microsoft's Windows. Even though I had their source, and I was going from an older version of Visual Studio to a newer one, the project was not trivial.

      • (Score: 0) by Anonymous Coward on Friday January 22 2016, @03:37PM

        by Anonymous Coward on Friday January 22 2016, @03:37PM (#293157)

        But what is unique to proprietary software is that it does not respect your freedoms. Want to make some changes to the software yourself, or hire some independent third party to do so? Too bad; you're a slave to the developers of the proprietary software. Not only that, but slimy companies like Microsoft often intentionally make their software incompatible with other, similar software so as to increase the costs of someone moving to competitors. Their software is inherently and intentionally defective. You don't find that with Free Software.

        • (Score: 2, Informative) by Anonymous Coward on Saturday January 23 2016, @05:12AM

          by Anonymous Coward on Saturday January 23 2016, @05:12AM (#293495)

          But both gcc and glibc were intentionally broken multiple times in order to avoid leaving stable interfaces proprietary software could use to hook into the toolchain, while still claiming it didn't fall under the GPL.

          It should be trivial to google and find the supporting links. But it was a major problem for pre-millenium linux development outside of the linux specific moving targets of that era, and was considered a major hurdle to wider adoption of gcc as a backend to other compiler frontends, outside of the normal GPL restrictions (since it had made even GPLed developments non-trivial to implement.) It is also part of the reason for the massive code generation incompatibilities between versions, since unnecessary changes were being made that broke things in subtle ways.

          LLVM's popularity combined with the switch to C++ is only going to make things worse, as well.

  • (Score: 3, Insightful) by stormreaver on Friday January 22 2016, @12:51PM

    by stormreaver (5101) on Friday January 22 2016, @12:51PM (#293086)

    The cost of replacing all Windows apps with Free apps for the entire city of Munich is a one time cost of $18M. Munich will recoup that cost in a few years, and then be free of Microsoft's (and others') iron grip. The city will then see ongoing savings and ongoing independence, the latter of which is priceless.

    • (Score: 2, Interesting) by Francis on Friday January 22 2016, @03:33PM

      by Francis (5544) on Friday January 22 2016, @03:33PM (#293155)

      Exactly. The library here replaced the proprietary software they had been using for their website and tracking software a few years ago. There was up front expenses and some glitches, but now they have a lot more control over what their software does and can add or remove features as needed. They can also hire whomever they like to fix things rather than the previous vendor.

      $18m for an entire city isn't even that much money. That comes out to less than $20 per man, woman and child in the city. As a one time cost that's hardly unreasonable. If they handle things correctly from there on out the cost shouldn't be any higher than what they were paying and might well be lower. Plus, they have a lot more control over what happens in the future.

    • (Score: 0) by Anonymous Coward on Friday January 22 2016, @07:55PM

      by Anonymous Coward on Friday January 22 2016, @07:55PM (#293294)

      The tense of your verb may already be incorrect.
      Back in 2012, the then-mayor estimated a €4 million savings.
      He grossly underestimated the more than €10 million [googleusercontent.com] (orig) [h-online.com] they had saved up to that point.

      ...and, once again, saving money was not the goal of the transition to Free Software--that was just a happy coincidence.
      The goal was to gain CONTROL of their software ecosystem.
      That same then-mayor has one of the great lines of all time (which left the world's richest man speechless).
      "Freedom from you, Mr. Gates." [googleusercontent.com] (orig) [europa.eu]

      -- OriginalOwner_ [soylentnews.org]

  • (Score: 2) by bradley13 on Friday January 22 2016, @01:48PM

    by bradley13 (3053) on Friday January 22 2016, @01:48PM (#293102) Homepage Journal

    $9 million doesn't strike me as expensive at all. They have legacy software that needs replaced or updated. Forty one pieces of legacy software. Those will need to be individually evaluated, and likely most of them will need to be replaced. Maybe some can be combined, maybe not. In the worst case, you are looking at 41 instances of software development.

    $9 million / 41 = an average cost of $200k per application. If these applications are anything but trivial, that is not at all expensive.

    FWIW I don't see any advantage of OSS here. Even if they had the code (and maybe they do, in some cases), it is at least 15 years old, likely more. It is out of date. It is written in dialects of languages that are no longer supported. You don't want to port it, you want to replace it.

    --
    Everyone is somebody else's weirdo.
    • (Score: 2) by stormreaver on Friday January 22 2016, @02:59PM

      by stormreaver (5101) on Friday January 22 2016, @02:59PM (#293137)

      FWIW I don't see any advantage of OSS here. Even if they had the code (and maybe they do, in some cases), it is at least 15 years old, likely more. It is out of date. It is written in dialects of languages that are no longer supported. You don't want to port it, you want to replace it.

      Having the source code for the old software, even in the case of a complete rewrite under the circumstances you outlined, provides a huge time and cost advantage. I had to do that very thing for my organization: a complete (and I mean from the first byte to the last) rewrite of a large, obsolete system that was written in-house by a sole, now-retired programmer 25 years ago in a horrible language/environment that few use anymore (UniVerse). Having the old source code was invaluable, and saved me a ton of time trying to figure out the business logic that nobody fully understood.

    • (Score: 2) by theluggage on Friday January 22 2016, @03:04PM

      by theluggage (1797) on Friday January 22 2016, @03:04PM (#293139)

      FWIW I don't see any advantage of OSS here.

      This. Software maintenance is a cost of doing business. If you're a big company or institution, getting stuck with legacy systems is a planning failure on your part. That $18M isn't a cost - its just money that should have been spent on IT support over the last 10 years (plus interest) that was probably recorded as 'efficiency savings' by the previous management before they moved on.

      Sure, any human being could be caught out like this, but its still their responsibility.

      From TFA: "When you're a big company or a big organisation, projects don't happen quickly. You may have seen it [the end of support for an OS] coming since 2013 and said 'We're going to shut this application down because it's not worth updating it to the new operating system but the quickest we can shut it down is 2016," he said.

      Nope, the end of support for Windows XP has been coming since it was launched in 2001, and odds are that, since these applications won't run after XP, they were already relying on 16-bit "legacy support" then. The process of replacing the applications should have started six months before the end of whatever support contract they had with the developer. If it was a kludgey, skunkworks solution brewed up a long-departed employee, then the cost of replacement should be set against the $$$ that employee probably saved the company by implementing something that did the job for 10 years.

      The important factor isn't open source: its open data. Hard to migrate if your data is tied up in a proprietary format...

    • (Score: 2) by bill_mcgonigle on Friday January 22 2016, @03:14PM

      by bill_mcgonigle (1105) on Friday January 22 2016, @03:14PM (#293144)

      Right - the apps are being replaced because they're not featureful, not because security is a nightmare for the IT staff. Does anybody actually think that IT wasn't recommending that these all be replaced before the XP EOL deadline? These changes are being driven by the politicians, unfortunately.

    • (Score: 0) by Anonymous Coward on Friday January 22 2016, @04:43PM

      by Anonymous Coward on Friday January 22 2016, @04:43PM (#293191)

      FWIW I don't see any advantage of OSS here. Even if they had the code (and maybe they do, in some cases), it is at least 15 years old, likely more. It is out of date. It is written in dialects of languages that are no longer supported. You don't want to port it, you want to replace it.

      Why? Only proprietary source code stays antiquated and untouched. If they wanted to make improvements to the code base they could. 15 years is a pretty long time for anything to stay constant.

      And you mean ancient languages, like C, ada or lisp?

      It's cute to be an apologist for proprietary software but please at least come up with reasons that make sense. Which I realize is hard.

    • (Score: 0) by Anonymous Coward on Friday January 22 2016, @05:54PM

      by Anonymous Coward on Friday January 22 2016, @05:54PM (#293238)

      18000000/4/41/1500 = $73 per year per application per employee who uses said application.

  • (Score: 2) by choose another one on Friday January 22 2016, @03:25PM

    by choose another one (515) Subscriber Badge on Friday January 22 2016, @03:25PM (#293153)

    TFA is (possibly deliberately?) vague about what it means by "only run on XP / 2000".

    Munich moved from XP/2000 to Linux, so it may be that it means "does not run on Linux" or it may be that it means "no longer developed, no upgrade path at all".

    In the former case it _is_ a cost resulting from the decision to migrate to Linux, but in the latter case it is nothing to do with it - just a natural cost of replacing anything end-of-life (might be more expensive to get a replacement on Linux in some cases I guess).

    • (Score: 2) by opinionated_science on Friday January 22 2016, @03:50PM

      by opinionated_science (4031) on Friday January 22 2016, @03:50PM (#293162)

      Yes, but this is tax dollars being spent for the public good. The inertial argument "well it will cost to move from XP based software" is null and void.

      Chalk it up to "paying for bad former decisions".

      Software paid for with public money should be on FOSS except in narrow, and individually contested cases.

      Oh, and there are companies that will Linux-ize software for you, so I call FUD.

      The money saved in the long run by improving the stability(!), is an experiment worth watching...

      • (Score: 2) by dx3bydt3 on Friday January 22 2016, @04:43PM

        by dx3bydt3 (82) on Friday January 22 2016, @04:43PM (#293192)

        I expect what they mean is, that xp or 2000 are the newest windows systems that could run the software. Where I work we have one last program that can't be run on anything newer than xp, specifically it is a plugin that requires Autocad 2000, which can't be installed on systems that won't run the 16 bit installer. I've got a couple of machines still running xp, and a VM. I'll eventually get the whole thing working under Wine, but for now this works.

        • (Score: 2) by opinionated_science on Friday January 22 2016, @04:50PM

          by opinionated_science (4031) on Friday January 22 2016, @04:50PM (#293194)

          My experience is that older stuff is easier to get running than newer stuff - depending on what Win32 bits it used.

          Buy my point stands. Virtualization is now mature. ALL machines should be linux, and any legacy machines should be Virtual.

          Hence, the reproduction of functionality can be pursued in parallel.

          Of course, you would have to be careful not to get caught in an Enterprise/VMware cash-milking project, but it is totally feasible within existing FOSS systems.

          Many of ancient programs I have used (for low-level things) might even be in DOS (freedos), win95-win2009 (wine), and if you have $$ pay Crossover and they'll port it for you.

          We need to stop paying massive corporations to run our local governments - it should be *local*.

          • (Score: 0) by Anonymous Coward on Friday January 22 2016, @08:20PM

            by Anonymous Coward on Friday January 22 2016, @08:20PM (#293305)

            That's the point everyone seems to be missing.
            Why export cash when you can keep it local and get the Multiplier Effect working for your own community|region|country|economic union?

            With much of the world's economy in the crapper, Germany has figured out how to make the giant downturn work in its favor.
            (Radio host and author Thom Hartmann likes to say that Germany has ultimately won WWII.)
            I'll argue that Munich has taken that success 1 step further by going with FOSS and creating jobs in their own area.

            -- OriginalOwner_ [soylentnews.org]

            • (Score: 2) by linuxrocks123 on Friday January 22 2016, @09:25PM

              by linuxrocks123 (2557) on Friday January 22 2016, @09:25PM (#293328) Journal

              Because you're misunderstanding the "multiplier effect" and it's not an important consideration for these decisions?

              The multiplier effect means that government spending stimulates the economy by some multiple of the cash spent. Stimulating the economy sounds good, but, unless you're in a recession, it's really not; it just increases inflation. This is why we have central banks and monetary policy -- to balance the competing goals of growth and inflation.

              Keeping cash local is a silly goal. If Munich hires programmers it otherwise wouldn't need to manage its infrastructure, it's hiring those programmers away from jobs in the private sector of Munich, or jobs outside of Munich where the programmers would commute and then return home at the end of the day, thus bringing in cash to the local economy.

              Unless those programmers can't find work. In which case, the area is in a recession, and the central bank should act. For this reason, the multiplier effect is typically defined for an entire economy, not a piece of it. "Entire economy" usually means "the area in question and all areas in the same currency union".

              • (Score: 0) by Anonymous Coward on Friday January 22 2016, @10:11PM

                by Anonymous Coward on Friday January 22 2016, @10:11PM (#293352)

                That is true.
                I understand the Multiplier Effect at least as well as you do.

                unless you're in a recession

                Nonsense.
                The Multiplier Effect is about getting money circulating.
                N.B. Some guy who already has a bunch of wealth and doesn't spend that on goods and services does nothing for the economy.
                (Buying art that was produced 500 years ago doesn't count; neither does simply "buying" other rich guys' homes at ever-inflated prices in a round-robin fashion.)

                Additionally, the stock market is NOT The Economy.
                This is very apparent when over 23 percent of the population want a job and can't get one.
                The vast majority of an economy is ordinary folks buying ordinary stuff.

                Keeping cash local is a silly goal

                Not when you can produce all the stuff you need locally.
                Trade|outsourcing is for stuff you can't -already- produce.
                ...and you second-guessing the phenomenally successful Germans is just comical.

                the central bank[...]

                ...should die in a fire.
                It's the greatest source of misery in Europe.

                -- OriginalOwner_ [soylentnews.org]

    • (Score: 4, Informative) by sudo rm -rf on Friday January 22 2016, @04:24PM

      by sudo rm -rf (2357) on Friday January 22 2016, @04:24PM (#293184) Journal

      Even for a native speaker, the linked PDF is ardous to read, but what I get from it is that they are running VMs with Win 2000 and Win XP resp. on Linux machines, but also native Windows machines. The reason for EOL is discontinued support for the applications and OSes.
      BUT
      They want to replace this mess with an application server, ie web applications accessed by a Citrix ICAClient.
      Last time I checked, ICA was a proprietary protocol by Citrix, a company that (quoting wikipedia) "licensed source code from Microsoft and has been in partnership with the company throughout its history".

      As a side note from my personal experience: Have fun trying to get that citrix client (java) to run on linux.

      • (Score: 2) by linuxrocks123 on Friday January 22 2016, @09:14PM

        by linuxrocks123 (2557) on Friday January 22 2016, @09:14PM (#293326) Journal

        When I last used Citrix, Linux had a native NPAPI plugin. It was hard to get working, but it eventually did work.

    • (Score: 2) by sjames on Friday January 22 2016, @04:56PM

      by sjames (2882) on Friday January 22 2016, @04:56PM (#293197) Journal

      It means the latter. They kept a few XP machines around to run those 41 apps when they switched to Linux, but it's increasingly difficult to maintain that setup in spite of moving XP to a VM.

  • (Score: 2) by drussell on Friday January 22 2016, @09:09PM

    by drussell (2678) on Friday January 22 2016, @09:09PM (#293325) Journal

    They would probably run just fine under WINE also....

    • (Score: 0) by Anonymous Coward on Friday January 22 2016, @10:56PM

      by Anonymous Coward on Friday January 22 2016, @10:56PM (#293372)

      It amazes me that these decade-old apps apparently aren't WINE-compatible.
      They must use some really niche APIs.

      I used to have a bookmark for a page that used red and green cells to show WINE's API compatibilities.
      Can't find that now and my Google Fu is failing me.

      -- OriginalOwner_ [soylentnews.org]

      • (Score: 2) by PinkyGigglebrain on Saturday January 23 2016, @07:20AM

        by PinkyGigglebrain (4458) on Saturday January 23 2016, @07:20AM (#293528)

        Some of the apps might have some kind of hardware associated with them like dongles or I/O cards that WINE or Linux doesn't support or the software might be making some direct calls to the hardware or OS's I/O libraries that rely on the Windows XP driver model.

        It would be a big help to the WINE project if Munich could spend some of those millions improving WINE but it really depends on why the software is OS specific.

        --
        "Beware those who would deny you Knowledge, For in their hearts they dream themselves your Master."
        • (Score: 0) by Anonymous Coward on Saturday January 23 2016, @09:26AM

          by Anonymous Coward on Saturday January 23 2016, @09:26AM (#293569)

          There are a number of insightful possibilities in your additional thoughts.
          Once Munich gets this sorted, I hope they put up a page with details of the 41 and the solutions found.

          Beige, up the (meta)thread, mentioned dongles.
          I know that there was (and, I suspect, still is) an industry to provide replacements for damaged dongles.
          As I am philosophically opposed to dongleware (having my tools unavailable while I get things unfucked never appealed to me), my only experiences with the paradigm are observations of the suffering of others.

          -- OriginalOwner_ [soylentnews.org]