Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
Space Government NASA United States

MuckRock Identifies The Oldest US Government Computer Still in Use (muckrock.com) 60

Slashdot reader v3rgEz writes: When MuckRock started using public records to find the oldest computer in use by the U.S. government, they scoured the country -- but it wasn't until a few tipsters that they set their sights a little higher and found that the oldest computer in use by the government might be among other planets entirely.
The oldest computer still in use by the U.S. government appears to be the on-board systems for the Voyager 1 and 2 space probes -- nearly 40 years old, and 12.47 billion miles away from earth. (Last year NASA put out a call for a FORTRAN programmer to upgrade the probes' software.) But an earlier MuckRock article identified their oldest software still in use on earth -- "the computers inside the IRS that makes sure everybody is paying their taxes". And it also identified their oldest hardware still in use -- "the machines running the nuclear defense system". (The launch commands are still stored on 8-inch floppy disks.)
This discussion has been archived. No new comments can be posted.

MuckRock Identifies The Oldest US Government Computer Still in Use

Comments Filter:
  • LOL! (Score:4, Funny)

    by DaMattster ( 977781 ) on Saturday October 08, 2016 @04:25PM (#53038483)
    Let's replace the computers controlling the nuclear launch system with ones from Lenovo! ROFLMAO! After all, just about every major computer component is made in China. Let's make it easy for our new overlords.
  • I'm curious if we still use any of the systems from the Minuteman-III from 1960 are still in use. I know they occasionally get upgrades, but I'd be willing to be some original stuff is still in use in some capacity.
  • by Blythe Bowman ( 4372095 ) on Saturday October 08, 2016 @04:58PM (#53038573)
    I imagine the old 8" floppies mentioned in one of the linked Slashdot articles were quite sturdy and reliable, at least compared to the crap 3 1/2" floppies I had the (dis)pleasure of using, right about the time CD-R drives started to become common on PCs. Read errors, "Track 0 bad -Disk unusable" and other such shit being fairly common no matter what floppy disk brand or how good the floppy drive in the computer was. Sometimes, after many tries, and some physical manupulation of the disk I could get the disk to finaly read the "bad" sector correctly, so I could copy the data off of it before it went into the garbage. I rarely had these problems with the 5 1/4" disks I used years before. Old computer equipmemt from 30+ years ago were built like tanks, designed to last at least a decade, much more reliable due to the primitive, simple OSes and software, the chips themselves had bigger/thicker transistors runing at very low speeds, and most used "hardwired" mask roms for the BIOS. Your tablet or phone probaly won't work 30 years from now due to the contents and BIOS data fading from the flash chips used for everything including system roms, but as long as the old antique hardware from the 80s back is kept in good condition, they will probaly still work fine centuries later, at least after the capacitors are replaced (caps are rather volitile, even modern ones).
    • I imagine the old 8" floppies ... were quite sturdy and reliable, at least compared to the crap 3 1/2" floppies I had the (dis)pleasure of using...... I rarely had these problems with the 5 1/4" disks I used years before.

      Old computer equipmemt from 30+ years ago were built like tanks,

      You are forgetting that 3 1/2" disks were also 30+ yo tech; I have never had any trouble with them. One of the reasons for their superseding 5.25" disks was that they were more robust. I still have 3 1/2" drives in two of my PCs and only last week pulled some data off a floppy that must have been written about 20 years ago.

      Your problem may have been that later floppies and drives were made with poorer quality control, it being assumed that anyone still using them was cheap and not very particular. The flo

      • and only last week pulled some data off a floppy that must have been written about 20 years ago.

        Yeah, I keep my ascii porn on 3.5" floppy disks too.

      • by guruevi ( 827432 )

        Not necessarily poorer quality control. The main issue with miniaturization is accuracy. On really old media (e.g. A vinyl record or even old floppy disks) a human can see and even control the position of the read header. As things started getting more dense, the accuracy had to go up to a point that even slight misalignments can cause data issues. Especially with the 3.5" floppys you sometimes need the exact same hardware to read a floppy because other manufacturers had different stepper motors with slight

        • I have a theory that after about 1995 when every PC came with a CD drive and most software including the OS came on CD, the quality control went down the toilet. I had a drive that punched tiny holes in disks and itself, but after that it was common to get a drive that only worked a few times and same for the disks. Lowest bidder don't give a shit checkbox manufacturing. The badness culminated in the early 00s when we were sold RAM and power supplies that don't work.

          On the other hand when I tried 15-20 year

          • by tlhIngan ( 30335 )

            I have a theory that after about 1995 when every PC came with a CD drive and most software including the OS came on CD, the quality control went down the toilet. I had a drive that punched tiny holes in disks and itself, but after that it was common to get a drive that only worked a few times and same for the disks. Lowest bidder don't give a shit checkbox manufacturing

            Of course, because the floppy became a useless appendage - good only to boot computers and a few other things (like store files on for trans

    • by Hylandr ( 813770 )

      "Track 0 bad -Disk unusable"

      This was usually a symptom of your anti-virus denying access to the boot sector of the drive because it detected badness. Wiping the disks with the library's anti-theft magnetizer would usually do the trick and the drive was usable again.

    • I imagine the old 8" floppies mentioned in one of the linked Slashdot articles were quite sturdy and reliable, at least compared to the crap 3 1/2" floppies I had the (dis)pleasure of using, right about the time CD-R drives started to become common on PCs.

      I googled around a bit looking for a chart I've seen several times, but I can't find it; it showed that 8" floppies are IIRC the second-best magnetic media in terms of the percentage of their potential data storage capacity which is actually used. The lower this percentage, the more resistant they are to damage, for a variety of reasons which are boring and which I'm not an expert on anyway. The best is the 160/320kB 5.25" floppy; they will last hilariously long. They were especially good towards the end of

    • by eionmac ( 949755 )

      The 8 inch discs still work well on some old machine room equipment I recently saw in UK. Very old machines used for repairing (oldish) rail trains. The main problem is that they be dusted off before use, as even inside a wooden box store container they accumulate dust. I was unable with my modern stuff to read them, so when they die the machine shop stuff dies as well. Then workshop has to work out new way to make copies of old parts, I assume 'printing' (additive manufacture?) by then will work.

    • The 8" (and 5 1/4") floppies were very fragile mechanically, but reliable if you were careful with them. Writing on them with a ballpoint pen or pencil would ruin them. Likewise, there was no cover for the media access hole when they were out of the drive; so, it was easy to contaminate them.

  • ...and it got few comments.

    There's just no pleasing this crowd.

    FWIW, I had trouble with 8" drives, not the diskettes themselves. The 5.25" and 3.5" stuff never failed on me, not once.

    • Re: (Score:3, Interesting)

      by Hylandr ( 813770 )

      We had a customer call us to fix their billing system that ran on an old 8088 in the mid 90s. It was in a welding shop and never cleaned. When we opened it there was metal power-laced dust-bunnies all over everything and the system was fried.

      We built them a new machine and went to restore their data. They had kept their backups on a 5.25 floppy and sometime they turned off the write verify because 'it was taking too long'. Well at some point years ago the drive stopped writing data to the disk and they had

    • by clovis ( 4684 )

      ...and it got few comments.

      There's just no pleasing this crowd.

      FWIW, I had trouble with 8" drives, not the diskettes themselves. The 5.25" and 3.5" stuff never failed on me, not once.

      I only a few 8-inch drives (microcode boot) on machines I supported, and they had more failures than anything else. I didn't have that many, but every single 8-inch drive I had to maintain failed in some way or another. Admittedly these were all non-IBM devices. I did not maintain the lone IBM Series/1 that we had on-site, but AFAIK it pretty much ran like a rock in a box.

      • Those drives were essentially identical to the drives for the DisplayWriter. Even if a tech got lazy with their feeler gauge and hosed the head alignment you could take a bit of time with a scope and dial in a drive to those badly written floppies. Drives were battleship-quality, indestructible. I replaced a few heads for wear, then just bought used off eBay. I think I last serviced a DisplayWriter system around 1999.

        DisplayWiters had MS-DOS 2.11 (and maybe 2.2x) available, ran a few apps such as a time

  • There is a GAO report covering legacy US Government systems. It has somewhat more detail.
    http://www.gao.gov/assets/680/... [gao.gov]

    It has this to say about the nuke system:

    The Strategic Automated Command and Control System is the Department of Defense’s (Defense) dedicated high-speed data transmission, processing, and display system. The system coordinates the operational functions of the United States’ nuclear forces, such as intercontinental ballistic missiles, nuclear bombers, and tanker support aircrafts, among others. For those in the nuclear command area, the system’s primary function is to send and receive emergency action messages to nuclear forces.
    According to Defense officials, the system is made up of technologies and equipment that are at the end of their useful lives. For example, the system is still running on an IBM Series/1 Computer, which is a 1970s computing system, and written in assembly language code. It also uses 8-inch floppy disks, which are a 1970s-era storage device; and assembly programming code typically used in mainframes. Replacement parts for the system are difficult to find because they are now obsolete.

    and regarding the IRS system:

    The Internal Revenue Service’s (IRS), Individual Master File (IMF) is the authoritative data source for individual taxpayer accounts. Within IMF, accounts are updated, taxes are assessed, and refunds are generated as required during each tax filing period. Virtually all IRS information system applications and processes depend on output, directly or indirectly, from this data source.
    IMF was written in an outdated assembly language code and operates on a 2010 IBM z196/2817-m32 mainframe. This has resulted in difficulty delivering technical capabilities addressing identify theft and refund fraud, among other things. In addition, there is a risk of inaccuracies and system failures due to complexity of managing dozens of systems synchronizing individual taxpayer data across multiple data files and databases, limitations in meeting normal financial requirements and security controls, and keeping pace with modern financial institutions.

    • IMF was written in an outdated assembly language code and operates on a 2010 IBM z196/2817-m32 mainframe.

      If only we had a simple tax code without a bunch of bullshit loopholes, perhaps we could just re-code it from scratch.

      • I'm betting the IMF is actually data storage/retrieval. The business logic is elsewhere.

        And it cannot be impossible to slip in a modern system, in parallel, to read/write the data, be validated, and move on. Then build the logical systems around it.

        But there is a reason why so many federal government systems are allowed to persist well beyond their end of life. Figure that out, and you will then understand what it will take to modernize.

  • On the flip side, we have the modern IBM System i. All software written for the very first System/38 in the 1970s runs unchanged and without recompilation on its followup, the AS/400.

    This same software also continues to run on today's IBM System i running the very latest hardware.

    A similar situation exists on the System/360 line, which also continues to run on modern hardware today.

    Think of all the many rehosting failures that could have been avoided.

    • The AS/400 is more the follow on for the S/38.

      Which was the end of the S/38-S/36-S/34-S/32, stated with a similar instruction set to the System/3, possibly actually the first 'personal computer' despite the physical size. The System/3 Mod 6 was intended for interactive mode, promoted BASIC as the programming language of choice despite true lack of support from IBM, and even had a cassette drive for updates/diagnostics. See a picture of the CE panel almost makes me shed a tear.

      System/i is an AS/400 rebrand,

      • by kriston ( 7886 )

        Yeah, the Rochester facility was always playing second fiddle even after the AS/400 was a proven success.

        A good book treatment is in "Fortress Rochester: The Inside Story of the IBM iSeries" by Frank G. Soltis.

  • I'm not so sure it's a good idea to upgrade to something more modern and hackable so we can all get killed faster through big data, crowdsourcing, the cloud and an app (iOS and Android - $1.99).

  • Just curious - how secure do you suppose the IRS computers are against eFiled tax returns using buffer overflow attacks?
    • The buffer overflow gets as far as the dot-matrix A4 printer and bowfs because it was expecting an 8.5x11in printer. When the pages are fed into the scanner to get the data into the main system from the e-System, the formatting changes break the attack code, at which point the misalignment code for handling the scannner thows out those pages and the remainder of the record is processed correctly.

      How else did you think it would work?

      Incidentally, it's a near certainty that the actual answer in the report i

  • Guess he will not be coming back yet....oh well.

  • The MuckRock article makes a bit too much of the situation.

    For the SAC C&C, the Series/1 is a well-documented piece of hardware that shouldn't be difficult to keep running pretty much in perpetuity. Even if we ran out of 8" floppies, it wouldn't be hard to emulate the device with a more modern storage medium.

    As for the IRS, why would it suffer "catastrophic systems failure"? They've updated the hardware, and IBM is not going to stop producing System z machines any time soon. 370 assembly language (presu

  • so they think. Anybody tried to read them lately?

We can predict everything, except the future.

Working...