Back in the day, the blue screen hex codes were actually insanely useful for debugging. Honestly? I think most (all?) linux distros could learn a lot from ACTUALLY showing the error when the machine dies a horrible death rather than assuming people will dig through the right log files before they rotate.
But I know a few years back (… so probably over a decade) they started having QR codes instead. Which sounded awesome until you try to use one and realize they all go to the same generic help page. Not sure if the hex codes were still there.
My favorite experience was coming to work and seeing a dedicated scientific instrument computer with the blue screen, error message was “bad pool header” or something, QR brings me to generic ass troubleshoot page. Google error and it says “something ain’t right, could be either software or hardware related, restart and you should never see it again” it did this every 4 days or so until we just re-imaged the damn thing.
I think part of the other problem is that Linux, in general, is meant for servers. So if anything functions, keep on chugging and rely on external monitoring to alert the user. Only acknowledge failure when all hope is lost.
Which is great for a server, not so much for a desktop where you might realize kde plasma fell over at 15:22 because you can’t click to a new window and the clock hasn’t updated since then (in fairness, I expected that when I finally installed openrazer). Nothing a reboot can’t fix but it still makes debugging a lot more tedious.
Not sure of a good way to handle it, but it would be wonderful to get stuff like that sooner rather than needing to wait for a full on kernel panic catastrophe.
And, in fairness, when I have to deal with a windows PC I have definitely noticed that MS changes the frequency of blue screens a lot from year to year. Back in the day, even a slightly wonky mouse driver would instantly trigger one. Then sometimes it feels like every single service could fail and it still wouldn’t pop one.
I think they stopped showing hex codes by default and just had error messages instead (I think hex codes can be turned on with the registry), it was still quite helpful, I’ve never been given a useless error but I think they were recently trialing removing error codes as well. I don’t know if they actually did that, I stopped using windows a few months ago.
Back in the day, the blue screen hex codes were actually insanely useful for debugging. Honestly? I think most (all?) linux distros could learn a lot from ACTUALLY showing the error when the machine dies a horrible death rather than assuming people will dig through the right log files before they rotate.
But I know a few years back (… so probably over a decade) they started having QR codes instead. Which sounded awesome until you try to use one and realize they all go to the same generic help page. Not sure if the hex codes were still there.
*scans QR code*
I DON’T FUCKING KNOW, YOU’RE THE ONE WHO BROUGHT ME HERE!
Has all the hallmarks of a good idea that never got properly implemented.
Google-fication comes for us all.
My favorite experience was coming to work and seeing a dedicated scientific instrument computer with the blue screen, error message was “bad pool header” or something, QR brings me to generic ass troubleshoot page. Google error and it says “something ain’t right, could be either software or hardware related, restart and you should never see it again” it did this every 4 days or so until we just re-imaged the damn thing.
Linux recently DRM (Direct Rendering Manager, not copyright protection garbage) subsystem recently gained the ability to do something even better.
It can render a QR code containing debug information and kernel logs.
I actually had to use this not too long ago. I wasnt expecting a helpful html file generated by the qr code.
I love it when real information is given to users, instead of esoteric error codes that lead you to bullshit slop sites when googled.
https://github.com/markusfisch/BinaryEye
Default apps tend to suck at recognizing big codes.
Huh this is few kernel versions back. I haven’t had a kernel panic in years, will have to simulate one and see if fedora supports this.
Yeah. Heard about that, never seen it in action.
I think part of the other problem is that Linux, in general, is meant for servers. So if anything functions, keep on chugging and rely on external monitoring to alert the user. Only acknowledge failure when all hope is lost.
Which is great for a server, not so much for a desktop where you might realize kde plasma fell over at 15:22 because you can’t click to a new window and the clock hasn’t updated since then (in fairness, I expected that when I finally installed openrazer). Nothing a reboot can’t fix but it still makes debugging a lot more tedious.
Not sure of a good way to handle it, but it would be wonderful to get stuff like that sooner rather than needing to wait for a full on kernel panic catastrophe.
And, in fairness, when I have to deal with a windows PC I have definitely noticed that MS changes the frequency of blue screens a lot from year to year. Back in the day, even a slightly wonky mouse driver would instantly trigger one. Then sometimes it feels like every single service could fail and it still wouldn’t pop one.
I think they stopped showing hex codes by default and just had error messages instead (I think hex codes can be turned on with the registry), it was still quite helpful, I’ve never been given a useless error but I think they were recently trialing removing error codes as well. I don’t know if they actually did that, I stopped using windows a few months ago.
Insanely useful is quite a stretch here
As long as someone knew what that code meant, it was kinda useful, yeah