Linux is great! All hail Linux!


  • Discourse touched me in a no-no place

    Click for bigger.

     

    Image Hosting by PhyreFile

    What has happened here is that Linux has detected that I'm still using first-generation Opterons, which can't even run vmware in 64bit mode because they're missing some stupid instruction or another. Obviously Linus and friends are reminding me to upgrade.



  • Jesus Christ.

    linux rant flame idiot fucking spam display:none



  • So,.. outdated OS has problems with outdated hardware?


  • Discourse touched me in a no-no place

    @Lord abletran said:

    So,.. outdated OS has problems with outdated hardware?
    This OS/system pair has been running happily unmolested for YEARS. The apparent cause is that I had the nerve to shut the shit down while I did some electrical work on their circuit. And naturally, a soft-reboot fixed the fucking problem, so it might have just been cosmic rays or some shit.

     

    But uh, i1586?



  • @Weng said:

    @Lord abletran said:

    So,.. outdated OS has problems with outdated hardware?
    This OS/system pair has been running happily unmolested for YEARS. The apparent cause is that I had the nerve to shut the shit down while I did some electrical work on their circuit. And naturally, a soft-reboot fixed the fucking problem, so it might have just been cosmic rays or some shit.

     

    But uh, i1586?

    Well, it does kind of make sense. In that not actually making any sense way.

    Edit: To clarify: It makes some sense that it might detect it as an i586 instead of an i686; for one thing, an AMD64 architecture wouldn't support the PAE of an i686, what with it already being 64-bit, however, it would still have the capabilities of at least an i586, so that makes a reasonable guess, I guess.

    Either that or your Opteron has the fdiv bug, and Debian detected that on startup somehow.

  • Discourse touched me in a no-no place

     i1586, not i586.

    There actually is a (really stupid) reason Linux devs did this. The P4 and up are still technically i686 (with sprinkles) in the engineering documentation, but Intel set the family on the P4's to 0xF instead of 0x6 (which everyone else copied) - and the Linuxfailures read the family to determine what number to put in that spot. They won't fix it, because "it's what the CPU is reporting itself as. Not our bug."



  • @Weng said:

     i1586, not i586.



    Sorry, I never learned how to read.
    @Weng said:

    There actually is a (really stupid) reason Linux devs did this. The P4 and up are still technically i686 (with sprinkles) in the engineering documentation, but Intel set the family on the P4's to 0xF instead of 0x6 (which everyone else copied) - and the Linuxfailures read the family to determine what number to put in that spot. They won't fix it, because "it's what the CPU is reporting itself as. Not our bug."


    Maybe I need to learn how to read, but I'm assuming that you're implying that AMD copied that? And that the devs, for no reason, used... printf("i%d86",vnumber); or something?


  • Discourse touched me in a no-no place

     Yes, that's exactly what I'm implying. And it wasn't just AMD, it was EVERYONE implementing an i686-plus-sprinkles CPU after the introduction of the P4. Essentially, at that very moment, the 80x86 numbering scheme became completely irrelevant and nobody cared because they were all differentiating their CPUs with different instruction sets during the post-Pentium-Pro era anyway.

     

    Oddly enough, Intel themselves stepped back down to a family ID of 6 after the P4. I don't know if AMD and the minor manufacturers went back, though.



  • We need rules for the sidebar. I'm thinking of something like:

    When you post a WTF in the sidebar that isn't entirely obvious, include a short paragraph of why you think it is a WTF. A WTF shall be considered not entirely obvious if it requires any knowledge that goes beyond basic computer using knowledge, including, but not limited to, SQL, Hardware stuff, any programming language (except for very simple snippets that can be understood by anyone who can read pseudocode), the circumstances that brought it forth, or the poster's mental health status or drug addictions.

  • Discourse touched me in a no-no place

     Yeah. I should have done a better job, but I'm worn out (FUCK DRUM BRAKES) and I was hugely pissed that I had to run down a stupid non-problem with a stupid erroneous error message just to get to the god damned maintenance records for my truck so I could update them with lots of curse words and angry bitching about the evils of 1900's era brakes on a product of the 1990's.



  • @Weng said:

     Yeah. I should have done a better job, but I'm worn out (FUCK DRUM BRAKES) and I was hugely pissed that I had to run down a stupid non-problem with a stupid erroneous error message just to get to the god damned maintenance records for my truck so I could update them with lots of curse words and angry bitching about the evils of 1900's era brakes on a product of the 1990's.

    I find it hilarious though how the first reaction to your anger was to post it here, instead of trying a reboot that later proved to fix it. That compensates for my initial confusion well enough.


  • Discourse touched me in a no-no place

     Dude. I'd just turned the thing on. Why the fuck should I have had to reboot it? Operating systems should not arbitrarily decide to go on vacation and freign incompatability.

     

    Now you you want some SERIOUS wtfitude, I could start talking about my truck's wiring.

    Sample: parking brake light will only come on if the 4WD "indicator light" is illuminated. The 4WD indicator light is not an indicator light - it's just a backlight for the trim piece that says "4WD" next to the transfer case shifter. All the documentation says this is an indicator light. Presumably, it was made to be always-on after the documentation had already been written and printed because someone realized a mandatory safety feature had been wired through it.

     

    Lo, I need to get my truck inspected, and my 4WD light is out. So I now have to go digging through the dashboard and wire the stupid thing PROPERLY (or change the bulb, which is actually even harder)



  •  Flakey hardware is flakey. And, yes, the stuff you get out of flakey hardware regularly makes one go "What the..." 

    One of my rules in PC maintainence is "It could be bad memory." A few bad bits can cause any symptom you name.



  •  It's funny you're blaming the OS, while it could have had a number of causes, including hardware failures. Never seen Windows boot and then throw a blue screen, but working fine after a reboot?

    But I guess it's easy to just bash Linux, because everyone's doing it. 



  • @pbean said:

    But I guess it's easy to just bash Linux, because everyone's doing it. 

    Also, bashing Linux is great if you want to avoid getting trolled by blakeyrat.



  • @derula said:

    Also, bashing Linux is great if you want to avoid getting trolled by blakeyrat.

    I didn't think it was possible to avoid getting trolled by blakeyrat, no matter what you bash or dash!



  • @pbean said:

    But I guess it's easy to just bash Linux
     

     of course it is. in fact, most linux distros require you to bash at least some of the time, and if nothing else, learning to bash is useful.



  • @pbean said:

     It's funny you're blaming the OS, while it could have had a number of causes, including hardware failures. Never seen Windows boot and then throw a blue screen, but working fine after a reboot?


    No, but I had a case of a W98 machine sometimes locking up, where further attempts to reset/reboot led to a "General Protection Fault" message on a black screen (which, unitl then, I thought died with Win3.11). It needed a full power cutoff and wait for capacitor discharge to start up again.

    I guess it was some memory/registers on some PCI card that didn't get zeroed on reset, and were messing up everything else.


  • Discourse touched me in a no-no place

    @pbean said:

     It's funny you're blaming the OS, while it could have had a number of causes, including hardware failures. Never seen Windows boot and then throw a blue screen, but working fine after a reboot?
    Nope, ain't seen that since Win9x. Everything has a detectable cause.

     

    There are three classes of hardware failure that could be responsible here, all of which would be permanent and noncorrecting.

    1) Physical CPU damage.
    2) Out and out BIOS corruption.
    3) Flaky memory - which isn't the transient phenomenon common in desktop memory when you've got ECC.

     

    And really. [b]i1586[/b].



  • @bannedfromcoding said:

    @pbean said:

     It's funny you're blaming the OS, while it could have had a number of causes, including hardware failures. Never seen Windows boot and then throw a blue screen, but working fine after a reboot?

    No, but I had a case of a W98 machine sometimes locking up, where further attempts to reset/reboot led to a "General Protection Fault" message on a black screen (which, unitl then, I thought died with Win3.11). It needed a full power cutoff and wait for capacitor discharge to start up again.
    I guess it was some memory/registers on some PCI card that didn't get zeroed on reset, and were messing up everything else.

    A long time ago I saw this weird stuff happening with a  Windows NT machine that had Dialogic cards. The first time after a reboot the cards would not work, a shutdown or at least a second reboot was needed. And it was not a transient thing, it was all the time. Sadly it was necessary to reboot the machine at least once a week because the software was lousy, and booting was super slow, so each time it took +30 minutes to do a full reset.

     



  • @thistooshallpass said:

    Sadly it was necessary to reboot the machine at least once a week because the software was lousy, and booting was super slow, so each time it took +30 minutes to do a full reset.

    It took me longer than normal to understand what you meant by that, since the standard way to write it is "30+ minutes."



  • @boomzilla said:

    @thistooshallpass said:
    Sadly it was necessary to reboot the machine at least once a week because the software was lousy, and booting was super slow, so each time it took +30 minutes to do a full reset.

    It took me longer than normal to understand what you meant by that, since the standard way to write it is "at least 30 minutes."



  • @Lord abletran said:

    @boomzilla said:
    @thistooshallpass said:
    Sadly it was necessary to reboot the machine at least once a week because the software was lousy, and booting was super slow, so each time it took +30 minutes to do a full reset.

    It took me longer than normal to understand what you meant by that, since the standard way to write it is "WAY TOO FUCKING LONG COMPUTERS ARE WHORES FUCK YOU"

    FTFY



  • I think the most annoying Stupid Windows 98 Trick I experienced was a cycle along the following lines:

    {boot}
    {scanreg runs}
    "Windows has detected a problem in your registry. Press Restore to load the last known good configuration."
    {scanreg restores old data}
    {reboot}
    {scanreg runs}
    "Windows has detected a problem in your registry. Press Restore to load the last known good configuration."
    {scanreg restores old data}
    {reboot}
    {scanreg runs}
    "Windows has detected a problem in your registry. Press Restore to load the last known good configuration."
    ...

    There was no available option other than to restore the last "known good" configuration, which after a couple of times through this clearly wasn't.

    TRWTF was that scanreg actually had a /fix option that you could use to repair the error instead of just recycling an old copy (and that's what I used to get the system running again) - but there was no way to use this in the UI; restore was the only option given. Had to boot to command line mode and do a scanreg /fix manually. How hard would it have been to add a "Repair" button?

    For sheer variety and randomness of errors, though, I have to agree with flaky memory as the winner.



  • Funniest part of scanreg is that Win98 makes a backup of the registry BEFORE checking if it's valid, and keeps only five most recent backups. So you have a very good chance to lose the good copy on a boot loop (or after rebooting to retry scanreg).



  • I also loved one BSOD a few years ago I had on a laptop , caused by a Windows XP registry key corruption.

    It appeared for 1/60th of a second before the machine rebooted. Had to video the screen and slow-motion replay it to get the error message. 



  • @mikedjames said:

    Had to video the screen and slow-motion replay it to get the error message.

    Hey, bub, we've got a special thread for stuff like this. But since you brought it up, what else did you find in the video?



  • @derula said:

    Jesus Christ.
    linux rant flame idiot fucking spam display:none

    Do you really think he's interested and is willing to help?



  • @Scarlet Manuka said:

    For sheer variety and randomness of errors, though, I have to agree with flaky memory as the winner.

    Disagree. PSU is the clear winner for me. Especially those cheap Shaw or other no-name brand one gets from MSY (you guys will love their site! Though they seem to have cleaned up their act recently wrt cheap-ass parts). My brother-in-law lives on a farm, so with pumps and other equipment causing surges on the line just kills electronics out there. His PC gets weird and wonderful errors until the PSU is replaced, then it is better for a few months before they start again. I think he has a ThermalTake PSU at the moment, which has held up longer than the others.

    Oh, and Linux is better than Windows with bad memory: this computer crashed constantly running Windows but didn't with the Linux, though it was still fairly unusable except playing music. :-) Back in the days of XMMS.



  • @mikedjames said:

    I also loved one BSOD a few years ago I had on a laptop , caused by a Windows XP registry key corruption.

    It appeared for 1/60th of a second before the machine rebooted. Had to video the screen and slow-motion replay it to get the error message. 

    I love this convergence of people who are smart enough to want to actually see the BSOD error, but dumb enough to not think of looking in the event logs. This is the second or third one I've seen on this forum... WTF people?

    You honestly video taped it? This has to be a troll.



  • @Zemm said:

    Oh, and Linux is better than Windows with bad memory: this computer crashed constantly running Windows but didn't with the Linux, though it was still fairly unusable except playing music. :-) Back in the days of XMMS.

    Looks like it was GPU memory at fault. Windows probably crashed because it actually has working drivers for the video card.



  • @Zemm said:

    Oh, and Linux is better than Windows with bad memory: this computer crashed constantly running Windows but didn't with the Linux, though it was still fairly unusable except playing music. :-) Back in the days of XMMS.

    That's pretty cool. Probably just a matter of where the OS happened to place its critical pieces around the bad memory.

    My XP lappy bluescreened about once every other month or so when I had it, and that was even after the patch to repair Microsoft's buggy SFU software from a bad interaction with an even buggier enterprise antivirus product. The work laptop was replaced in April-ish with a new Win 7 lappy. I've since had two or three hard crashes with it, including one with (repairable) data corruption. Both laptops were unnecessarily expensive enterprisey Dells. I'm not sure if it's shoddy hardware or crappy ventilation or just the untypical way I use the software, but I've still yet to get the reliability from work equipment that I can from a consumer-grade piecemeal-bought desktop filled with the last decade's hardware running Linux.



  • @mikedjames said:

    It appeared for 1/60th of a second before the machine rebooted. Had to video the screen and slow-motion replay it to get the error message. 
    ...another one? Hint: keep tapping F8 before the Windows bootup screen appears, and select "Disable automatic restart on system failure".



  • Ah... Drum brakes are TRWTF.

    Protip: pull both drums, but only take one side apart at a time. That way you have a reference point (albeit one that is somewhat mirrored from the other one!)



  • @Xyro said:

    @Zemm said:
    Oh, and Linux is better than Windows with bad memory: this computer crashed constantly running Windows but didn't with the Linux, though it was still fairly unusable except playing music. :-) Back in the days of XMMS.

    That's pretty cool. Probably just a matter of where the OS happened to place its critical pieces around the bad memory.

    My XP lappy bluescreened about once every other month or so when I had it, and that was even after the patch to repair Microsoft's buggy SFU software from a bad interaction with an even buggier enterprise antivirus product. The work laptop was replaced in April-ish with a new Win 7 lappy. I've since had two or three hard crashes with it, including one with (repairable) data corruption. Both laptops were unnecessarily expensive enterprisey Dells. I'm not sure if it's shoddy hardware or crappy ventilation or just the untypical way I use the software, but I've still yet to get the reliability from work equipment that I can from a consumer-grade piecemeal-bought desktop filled with the last decade's hardware running Linux.

     

    i had this on a hp compaq laptop, 4 hard drives in 2 years.  I blame the hdd manufacturer, although I didn't check at the time i suspect hp swap it for (more or less) the same make / model hdd? From that experience no matter if it is a work or home lappy that I plan to use regularly I buy a ssd.... though the work one i put into expenses as it will prevent an issue that stalls me in my tracks from about a day.


     



  • @Helix said:

    i had this on a hp compaq laptop, 4 hard drives in 2 years.  I blame the hdd manufacturer, although I didn't check at the time i suspect hp swap it for (more or less) the same make / model hdd? From that experience no matter if it is a work or home lappy that I plan to use regularly I buy a ssd.... though the work one i put into expenses as it will prevent an issue that stalls me in my tracks from about a day.

    Yeah, we're having a lot of dead Hitachi 2.5" drives in HP laptops (and other-brand touchscreen devices). Warranty always replaces them with the same model.

    Back when the HDD business was Hitachi/IBM, they were pretty OK... not great but OK... ever since it becoming Hitachi alone, they're worse than Maxtor and Samsung were at their worst times.



  • @bannedfromcoding said:

    ever since it becoming Hitachi alone, they're worse than Maxtor and Samsung were at their worst times.
     

    I am not, nor have ever been, able to support the notion that one drive manufacturer produces consistently poorer quality drives. They have all been consistently good.

    My data point can beat up your data point.



  • @dhromed said:

    @bannedfromcoding said:

    ever since it becoming Hitachi alone, they're worse than Maxtor and Samsung were at their worst times.
     

    I am not, nor have ever been, able to support the notion that one drive manufacturer produces consistently poorer quality drives. They have all been consistently good.

    My data point can beat up your data point.


    Well, take geographical location as a factor; I wouldn't be surprised if the Netherlands would get very different quality of the "same" product than Poland does.

    Of course, the laptops have been purchased in a country from the "old good West" part of EU, so I'm kinda shooting down my own theory here.



  • @bannedfromcoding said:

    Well, take geographical location as a factor; I wouldn't be surprised if the Netherlands would get very different quality of the "same" product than Poland does.

    My father in law loves to bitch about Toyota, Nissan, et.al., which have had a pretty good reputation for quality for some time her in the US. Back in the 70s and 80s, he traveled to southeast Asia a lot, and apparently the seconds, lemons and rejects all went there, with the best output coming to places like the US.



  • Automatic restart being enabled by default is TRWTF.



  • @lolwtf said:

    Automatic restart being enabled by default is TRWTF.

    Why? It's not like you can fix the issue in real-time, the only real option is to reboot. Also, the error is nicely logged and stored away for later so you can still research it.

    Edit: plus it's possible the computer is running important services that should be restarted ASAP.



  • @Scarlet Manuka said:

    I think the most annoying Stupid Windows 98 Trick I experienced was a cycle along the following lines:

    {boot}
    {scanreg runs}
    "Windows has detected a problem in your registry. Press Restore to load the last known good configuration."
    {scanreg restores old data}
    {reboot}
    {scanreg runs}
    "Windows has detected a problem in your registry. Press Restore to load the last known good configuration."
    {scanreg restores old data}
    {reboot}
    {scanreg runs}
    "Windows has detected a problem in your registry. Press Restore to load the last known good configuration."
    ...

    There was no available option other than to restore the last "known good" configuration, which after a couple of times through this clearly wasn't.

    TRWTF was that scanreg actually had a /fix option that you could use to repair the error instead of just recycling an old copy (and that's what I used to get the system running again) - but there was no way to use this in the UI; restore was the only option given. Had to boot to command line mode and do a scanreg /fix manually. How hard would it have been to add a "Repair" button?

    For sheer variety and randomness of errors, though, I have to agree with flaky memory as the winner.

     

    Power supply problems at least for the tie.

    Had a machine one time that was crashing regularly.  I started looking at it and was enjoying the variety of BSOD error codes I'd never seen before.  Went into the BIOS and found the +12V line fluctuating between about +5V and +12V, causing the hard drive to spin down whenever the voltage got too low.

    I was IT support for our department, but I wasn't in the official IT group for our organization.  Most of the time, I could get things RMA'd readily by calling in an official tech (they just said "Okay, it's broken, what do you need?" -- I knew them, they knew me, etc.) but TRWTF was that this time they insisted an official tech come down and look at it.  I gave him my diagnosis (bad power supply), the reasons (constantly crashing + BIOS values fluctuating), and he told me "Well, I've seen the BIOS give bad values before.  It's probably a virus.  I'll look at it."

    90 minutes later, he comes out, tells me he can't keep the machine stable enough to know for sure if it's a virus.  He also said there was something in the office that was aggravating his allergies (still don't know what it was) and he was willing to go with my theory.  He would order a new power supply.  He also asked if I would mind installing it when it came in.

    A tip I learned from being an EMT:  treat the patient, not the monitor.  If I had seen that the BIOS values were bobbing up and down but the machine was otherwise acting fine, I'd agree that the BIOS may not have been measuring voltages correctly.  In this case, I had a monitor (the BIOS) that was showing incorrect values, plus a symptomatic patient (an unstable/unusuable PC).  So  I was certain that I was looking at a bad power supply here.

     



  • @bannedfromcoding said:

    @dhromed said:

    @bannedfromcoding said:

    ever since it becoming Hitachi alone, they're worse than Maxtor and Samsung were at their worst times.
     

    I am not, nor have ever been, able to support the notion that one drive manufacturer produces consistently poorer quality drives. They have all been consistently good.

    My data point can beat up your data point.


    Well, take geographical location as a factor; I wouldn't be surprised if the Netherlands would get very different quality of the "same" product than Poland does.

    Of course, the laptops have been purchased in a country from the "old good West" part of EU, so I'm kinda shooting down my own theory here.
     

    What if it's not product quality per se, rather the application of that model (and it's specifications) into the laptop and the total assembly application use by you?  Meaning that a fringe issue which could be exacerbated by another characteristic and maybe a third... making the problem somewhat unique.


     



  • @nonpartisan said:

    [

    Power supply problems at least for the tie.

    Had a machine one time that was crashing regularly.  I started looking at it and was enjoying the variety of BSOD error codes I'd never seen before.  Went into the BIOS and found the +12V line fluctuating between about +5V and +12V, causing the hard drive to spin down whenever the voltage got too low.

    I was IT support for our department, but I wasn't in the official IT group for our organization.  Most of the time, I could get things RMA'd readily by calling in an official tech (they just said "Okay, it's broken, what do you need?" -- I knew them, they knew me, etc.) but TRWTF was that this time they insisted an official tech come down and look at it.  I gave him my diagnosis (bad power supply), the reasons (constantly crashing + BIOS values fluctuating), and he told me "Well, I've seen the BIOS give bad values before.  It's probably a virus.  I'll look at it."

    90 minutes later, he comes out, tells me he can't keep the machine stable enough to know for sure if it's a virus.  He also said there was something in the office that was aggravating his allergies (still don't know what it was) and he was willing to go with my theory.  He would order a new power supply.  He also asked if I would mind installing it when it came in.

    A tip I learned from being an EMT:  treat the patient, not the monitor.  If I had seen that the BIOS values were bobbing up and down but the machine was otherwise acting fine, I'd agree that the BIOS may not have been measuring voltages correctly.  In this case, I had a monitor (the BIOS) that was showing incorrect values, plus a symptomatic patient (an unstable/unusuable PC).  So  I was certain that I was looking at a bad power supply here.

     

     

    Did you use a stethoscope on the hard disk too?    

     



  • @Helix said:

    Did you use a stethoscope on the hard disk too?    

     

    Not in this instance, but I have before . . .

    And to clarify, yes, the new power supply fixed the problem.  Imagine that.

     


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.