• noaunda (unregistered)

    I used the text screensaver, which displayed FRIST throughout my screen...

  • Matthew (unregistered)

    There could be a whole category of TDWTF where a screensaver on a server is the problem. Definitely not the first one of these.

  • WowImpressed (unregistered)

    Screensaver on server was eating all the resources.

    Check.

    Wow. This made my day.

    I'm off to xkcd.com

  • Warren (unregistered)

    We are missing a vital detail. If this was Windows 95, did James spot any teapots while he was looking at the pipes?

  • Mainframe (unregistered) in reply to Matthew

    Agreed. This comes to mind: http://thedailywtf.com/Articles/A-Fat-Pipe.aspx

  • Black Bart (unregistered)

    TRWTF was Windows 95, that couldn't even set the priority of a screensaver correctly.

  • (cs) in reply to Mainframe
    Mainframe:
    Agreed. This comes to mind: http://thedailywtf.com/Articles/A-Fat-Pipe.aspx
    Yeah, I thought of that one, too. At least this time there's no Mario...
  • (cs) in reply to Black Bart
    Black Bart:
    TRWTF was Windows 95, that couldn't even set the priority of a screensaver correctly.
    Oh, it *could*. It just didn't bother.
  • foxyshadis (unregistered) in reply to Black Bart
    Black Bart:
    TRWTF was Windows 95, that couldn't even set the priority of a screensaver correctly.
    You appear to be laboring under the mistaken assumption that Windows 95 had anything like process or thread priority levels.

    Much like the middle of the story labored under the weight of terribad movie euphemisms.

  • Zathras (unregistered)

    Probably a stupid question, but why would a screensaver that's written in OpenGL - and therefore is presumably running on a GPU, rather than a software renderer - use up all the CPU time?

  • Nick (unregistered) in reply to Zathras
    Zathras:
    Probably a stupid question, but why would a screensaver that's written in OpenGL - and therefore is presumably running on a GPU, rather than a software renderer - use up all the CPU time?

    The CPU must still in part coordinate the access and other bits and bobs to the GPU that OpenGL and any other graphics API involve.

    Seeing as we're talking Windows 95 here, the CPU would not have been more than a potato chip...

  • Fritz, a.k.a. Fritzo (unregistered)

    The RWTF is that this wasn't terribly written and actually contained a funny line.

  • Shinobu (unregistered) in reply to Zathras
    Zathras:
    Probably a stupid question, but why would a screensaver that's written in OpenGL - and therefore is presumably running on a GPU, rather than a software renderer - use up all the CPU time?
    Because if you don't have a graphics card, the thing falls back on a software renderer. I'm thinking their server was a re-purposed 486.

    That said, this story is older than dirt.

  • Patrick (unregistered) in reply to Zathras

    GPU and Windows 95? Is this a story with another flux capacitor?

    No, really, the Voodoo graphics card was released at the end of 1996 and even from there on it was a long way to something you could call GPU. And even today there is no real reason to put a GPU into a network server.

  • Brompot (unregistered) in reply to Nick
    Nick:
    Zathras:
    Probably a stupid question, but why would a screensaver that's written in OpenGL - and therefore is presumably running on a GPU, rather than a software renderer - use up all the CPU time?

    The CPU must still in part coordinate the access and other bits and bobs to the GPU that OpenGL and any other graphics API involve.

    Seeing as we're talking Windows 95 here, the CPU would not have been more than a potato chip...

    Actually in the days of Windoze 95 you were lucky if you had 2D acceleration. 3D accelleration came mostly in add on cards such as a voodoo card, if at all. Not to be found on servers or even most workstations.

    As OpenGL is all 3D everything landed on the CPU, which was a Pentium I at best.

    Try playing the original version of Doom in a dos box, you'll get the idea of 3D graphics in that day and age.

  • (cs)

    I don't see how "The cleaner" is bad ass. Perhaps the author meant "bad as".

  • (cs)

    We had one of these moments at my previous job. One day, the network was totally sluggish and automated jobs usually running at night had failed. Nobody understood what had happened. Until we went to the desk of the IT-support guy.

    Turns out, they had rolled out a new fancy screensaver on all machines, which was nothing more than a movie of a spinning cube with the company logo. Servers included. Not only did it consume CPU, but our guy used a constant-on remote desktop connection to all of these machines, so the spinning logo-movies quickly saturated the network.

    Since this was screensaver for all countries ordered by corporate HQ, no-one had access to remove them either. It took almost a full day before everything was fixed server-side. However, it took months for the sales team to convince the right people that a looping .avi wasn't really the best choice for their laptop batteries.

  • Walky_one (unregistered)

    TRWTF was actually that "Pipe" screen saver. Many people had it active in those days and it regularly caused problems. I honestly don't know how anybody can be stupid enough to build a screen saver that needs a whole computer for himself to run.

  • Fritz, a.k.a. Fritzo (unregistered)

    The real TRWTF is that it wasn't Johnny Castaway.

  • Norris (unregistered)

    I once worked for a VAR (Value Added Reseller) around about the time that the SETI@Home project was started. Yep, you've already guessed - one bright spark of a techie was ramping up his score by using the servers he was building for a customer... and forgot to remove the software before they were delivered. However, this isn't a story just about CPU utilisation, as this was back in the days of highly expensive ISDN links, and SETI@home needs to grab another set of data every few minutes...

  • faoileag (unregistered) in reply to Steve The Cynic
    Steve The Cynic:
    Black Bart:
    TRWTF was Windows 95, that couldn't even set the priority of a screensaver correctly.
    Oh, it *could*. It just didn't bother.
    Could it really? I seem to recall that early windows versions used "cooperative multitasking" instead of "preemptive multitasking", thus putting the onus of letting other applications room to breathe erm cycles to use on the application developer instead of on the operating system. But I might be wrong, perhaps someone can clarify?

    Captcha: populus - I played that a lot on windows 95 :-)

  • (cs) in reply to Walky_one
    Walky_one:
    TRWTF was actually that "Pipe" screen saver. ... I honestly don't know how anybody can be stupid enough to build a screen saver that needs a whole computer for himself to run.
    Common misconception. The pipe drawing only really needed about 2% of available CPU; it's purpose was to keep the humans distracted while the other 98% was used to try to achieve self-awareness.
  • (cs) in reply to Matthew

    Very similar to http://thedailywtf.com/Articles/A-Fat-Pipe.aspx, but with less deliberate stupidity.

  • ggy (unregistered)
    They turned it off and own, which fixed the problem. But only momentarily.
    Well, if they kept owning the system of course it was only okay momentarily!
  • QJo (unregistered)

    TRWTF was screensavers themselves. I lost count of the number of times some colleague would prattle on about this or that cool screensaver, which would frequently be the direct cause of computer problems. Many of these damn things contained deliberate malware.

    This entry is a worthy one for TDWTF purely because of, as the last line indicates, the 500-mile journey to fix it. Such was typical in those days.

    Takes me back to those bad old days. Bravo.

  • Adam (unregistered)

    TRWTF is that he killed a hooker.

  • ANON (unregistered) in reply to QJo
    QJo:
    TRWTF was screensavers themselves. I lost count of the number of times some colleague would prattle on about this or that cool screensaver, which would frequently be the direct cause of computer problems. Many of these damn things contained deliberate malware.

    This entry is a worthy one for TDWTF purely because of, as the last line indicates, the 500-mile journey to fix it. Such was typical in those days.

    Takes me back to those bad old days. Bravo.

    I agree. Good that LCDs and Plasma screens don't need them anymore.

  • Mike (unregistered) in reply to Matthew

    I once saw an issue where IF DaVinci Mail triggered a new mail alert on a computer that was currently displaying the Fireworks screen saver it would lock the system to the point that no computer on the network after that computer had network access.

    Old Arcnet - 286 boxes. It was a while ago.

  • huppenzuppen (unregistered) in reply to faoileag
    faoileag:
    Could it really? I seem to recall that early windows versions used "cooperative multitasking" instead of "preemptive multitasking", thus putting the onus of letting other applications room to breathe erm cycles to use on the application developer instead of on the operating system. But I might be wrong, perhaps someone can clarify?

    No, that was Windows 3.1

  • Daniel Migowski (unregistered)

    This is also an often forgotten problem with virtual machines running Windows 7. They almost always still have their screen savers on, and the default Windows 7 screensaver is very resource hungry and always grabs about 100Mhz for doing nothing on the VMs, especially not "saving" anything.

    If you wonder why thats a problem for us... at our site we have some virtualized Window 7's to run unit tests in the night.

  • Frank (unregistered) in reply to Matthew

    Indeed. I had one in the 90's in the UK for client/server app I had just written for a large music label.

  • (cs)

    Request Deletion: Reason - duplicate

  • (cs) in reply to faoileag
    faoileag:
    Steve The Cynic:
    Black Bart:
    TRWTF was Windows 95, that couldn't even set the priority of a screensaver correctly.
    Oh, it *could*. It just didn't bother.
    Could it really? I seem to recall that early windows versions used "cooperative multitasking" instead of "preemptive multitasking", thus putting the onus of letting other applications room to breathe erm cycles to use on the application developer instead of on the operating system. But I might be wrong, perhaps someone can clarify?

    Captcha: populus - I played that a lot on windows 95 :-)

    You are wrong, up to a point. Windows (non-NT) 3.* and earlier had strict cooperative multitasking betweenWindowsapplications. This collection then (in "386-enhanced" mode) preemptively multitasked against the mini-VMs that hosted native DOS programs. A great deal of the I/O subsystem was mutexed to make it sane, although 32-bit file access and disk access mitigated this somewhat.

    NT had a similar behaviour between all Win16 applications that shared an address space, but everything else was preemptive.

    The 9x line was a strange sort of hybrid between the two worlds. DOS programs still preemptively multitasked against each other and against the Windows "subsystem", but Win32 programs also preemptively multitasked / multithreaded against each other, the DOS programs, and the Win16 programs that were running. But those Win16 programs had to cooperate with each other.

    And 9x had process and thread priority - if you ran SETI@Home (pre-BOINC) on 9x, it took low priority to prevent impact on your normal activity.

  • Hasse (unregistered)

    Did he see the theepot junction?

  • EvilSnack (unregistered)

    If he didn't take the time to uninstall the non-trivial screen savers, this problem will probably return.

    CAPTCHA venio: I came, I saw, I did something that a noob will be able to undo five minutes after I leave.

  • Patrick (unregistered) in reply to WowImpressed

    But it's Thursday, no new content today at xkcd. :-(

  • Thanatos Complex (unregistered) in reply to Adam
    Adam:
    TRWTF is that he killed a hooker.
    Glad I'm not the only one who noticed that.
  • faoileag (unregistered) in reply to Steve The Cynic
    Steve The Cynic:
    faoileag:
    ...But I might be wrong, perhaps someone can clarify?
    You are wrong, up to a point. Windows (non-NT) 3.* ... a lot of windows internals ... prevent impact on your normal activity.
    Wow, that certainly was some comprehensive clarification!

    That's what I like about the TDWTF forum - a lot of banter, but great background information as well.

  • Captcha: oppeto (unregistered)

    Wall of text. DNR.

  • belzebub (unregistered)

    I'm missing the last part - where upon his return, James is told that the system is down again and the customer has filed a complaint against him for breaking their screensaver, which - thank you very much - was working thr whole time just fine. They were fortunately able to restore the full glory of the screensaver in a few hours. Incidentaly the system failed just about 10 minutes after screensaver was fixed, which proves that James didn't fix anything, has meddled with their state-of-the-art screensaver and probably just got lucky. After replacing VGA card with OpenGL-capable one, James was fired for improper sexual behaviour (the hooker called him gay) and was killed by meteorite from mars. The (proper) End.

  • (cs)

    From those old ages this is the best screensaver ever: http://web.onetel.net.uk/~gnudawn/johnny/

    I remember that being deployed to all computers in a governmental institution, and people watching it when they got bored playing Solitaire.

  • (cs) in reply to Matthew

    THE PIPES STRIKE AGAIN1!!!!one11!!!eleven

  • TRWTF is... (unregistered)

    Clearly TRWTF is that someone was wasting CPU cycles on some networked ticketing crap when they were all clearly needed for the pipes.

  • Cisco (unregistered)

    Sadly most energy on this planet is probably used to run some stupid Windows screensaver on every unattended laptop.

    Especially the 3D ones seem to run with the highest frame rate that they can achieve, burning as much CPU/GPU power as they can.

    Of course any screen saver should do the contrary and use as little CPU/GPU as possible with a very low priority AND turn itself off completely when the screen goes into standby after some minutes.

  • faoileag (unregistered) in reply to belzebub
    belzebub:
    I'm missing the last part - where upon his return, James is told that the system is down again and the customer has filed a complaint against him for breaking their screensaver, which - thank you very much - was working thr whole time just fine. They were fortunately able to restore the full glory of the screensaver in a few hours. Incidentaly the system failed just about 10 minutes after screensaver was fixed, which proves that James didn't fix anything, has meddled with their state-of-the-art screensaver and probably just got lucky. After replacing VGA card with OpenGL-capable one, James was fired for improper sexual behaviour (the hooker called him gay) and was killed by meteorite from mars. The (proper) End.
    Nice try. But you forgot the president's daughter.
  • The Fury (unregistered)

    Why would you run Windows 95 on a server! Windows NT surely. Not sure it would have avoided this issue.

  • faoileag (unregistered) in reply to Cisco
    Cisco:
    Of course any screen saver should do the contrary and use as little CPU/GPU as possible with a very low priority AND turn itself off completely when the screen goes into standby after some minutes.
    Exactly. If not the framerate in my browser's HTML5 fishbowl (http://ie.microsoft.com/testdrive/performance/fishbowl/") might go down and my 250+ fish will get too little exercise.
  • Krunt (unregistered) in reply to foxyshadis
    foxyshadis:
    You appear to be laboring under the mistaken assumption that Windows 95 had anything like process or thread priority levels.

    It did have process and thread priority levels.

  • heh (unregistered)

    the wtf is having desktop on server... that and monitor...

  • JW (unregistered)

    This link (among others easily findablethrough Google), briefly discusses why using "ghetto" to mean "low calss, dangerous" is problematic.

    http://www.huffingtonpost.com/2013/10/24/offensive-words-_n_4144472.html

    (Briefly, since the original definition of "ghetto" was "part of a city where a minority dwells", then using it to mean "low class" is the same as saying "minorities are low class".)

Leave a comment on “I Love the Smell of Popcorn in the Morning”

Log In or post as a guest

Replying to comment #:

« Return to Article