Road To Damascus

“Dread started to creep up my spine as I sat there, eyes focused on the Blue Screen of Death. At the time, there were a few spare PC parts around the house but nothing to build a workable desktop from. Luckily, I had a USB stick floating around with a Linux on it, and this USB stick stored a special “Live” version of that operating system. That meant I could plug the USB stick in and run the operating system without needing to install anything on computer..”
 
See My Open Source Story: Adam Sims
My conversion was a lot less dramatic than Adam Sims. Rather than a single crisis, my revelation was a period of time with frequent failures to deliver a useful desktop. That Other OS was freezing hourly in my classroom. GNU/Linux ran trouble-free for months. I instantly saw the light, however. Since then I’ve grown in knowledge and capability with GNU/Linux just like Sims. The route is not as important as the destination.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in technology and tagged , , , , , , , . Bookmark the permalink.

30 Responses to Road To Damascus

  1. oiaohm says:

    –kurkosdr X11 crash does happen under Windows–

    Opps wording.

    The problem with using TM bullshit is you completely miss the fact that. Under Windows users can be automatically logged out by the system same way a X11 server crash does. Session management failure of Windows will log you out.

    Worst part is someone at Microsoft thought that this was a design feature.

    http://answers.microsoft.com/en-us/windows/forum/windows_7-desktop/how-to-disable-frequent-auto-logoff-in-windows-7/0d690c5e-54be-49bb-98ad-caf1153ce876?auth=1

    Yes added to Windows 7 a feature to automatically log you out completely when system is left unattended. Good fun go to loo come back and find Windows has vaporised your work. This is not like X11 where you have a crash log to show to boss as evidence that something when wrong. Its recorded as if you logged yourself out.

  2. oiaohm says:

    kurkosdr X11 crash does happen under Windows. You do have cases where users magically find themselves logged out under windows without them doing anything.

    Usermode setting blocking BSOD message happened in Windows 3.1x and Windows 95 with particular video card drivers. Change to NT kernels designed kernels removed this possibility.

    Good to know that Microsoft has, if nothing else, addressed X11 userspace mode blocking kernel panic … As always, Microsoft represent the Good Guys here.
    Dr Moron you don’t even know how to rewrite this correctly.

  3. DrLoser says:

    M$ updates are known reasons for BSOD as well, but eh…let’s not go there.

    Now, to be fair here, Dog-Brain, I’m generously going to assume that this recent period of “lack of economic activity,” as it were, is nothing at all to do with the obvious fact that you have nailed your colors to the Linux mast.

    I have no particular reason to assume that a man without an HSE cannot, absent that rather trivial qualification, manage something that is recognisable as a sabbatical.

    What larks! Let’s start here then, Dog-Brain:

    Explain the difference between a “Blue Screen of Death” and a “Red Screen of Death.”

    For added value here, do your best to explain the equivalents in the Linux kernel. Or, alternatively, and this would be a very useful explanation — tell us why and how the Linux kernel has no need of the equivalents of BSoDs or RSoDs.

    Over to you, Dog-Brain.

  4. Deaf Spy says:

    M$ updates are known reasons for BSOD as well, but eh…let’s not go there.

    Come on, dougie, go there 🙂

  5. dougman says:

    Re: Ancient history. Bad/glitching hardware and the occasional bad driver (usually unsigned) are the only times one sees BSOD’s these days.

    You don’t say? S0 occasionally but repeatedly, is that like often seldom, hardly softly or always never? LOL..

    http://www.tenforums.com/bsod-crashes-debugging/

    M$ updates are known reasons for BSOD as well, but eh…let’s not go there.

  6. Wizard Emeritus wrote, “In most case a BSOD is due to a hardware problem, repair/replace the hardware and the BSOD is gone.”

    That was not my experience. In the first school where I installed GNU/Linux, it ran flawlessly on the same hardware that Lose ’95 failed daily. XP was failing several times each year on TLW’s PC. GNU/Linux has not failed since except for power failures after our UPS died on the same hardware. Beast has been running Debian GNU/Linux since its inception as an AMD64 machine back in 2005 or so and I just moved the hard drives over to get the quad-core version going in 2008. Similarly, when the motherboard failed, I just replaced it and rebooted without any changes to the software. TOOS would have wanted many reinstallations in that period. Beast’s drives have 28 to 84K hours of operation. Only one has failed in all those years. The OS has just kept going and I’ve worked it hard. Lately, it runs 200 processes.

  7. DrLoser says:

    Linux biggest failure is X11 userspace mode blocking kernel panic from being displayed but with this disappearing Microsoft really need to start addressing some of the other problems.

    Babble, Fifi. Nothing but babble.

    Good to know that Microsoft has, if nothing else, addressed X11 userspace mode blocking kernel panic … As always, Microsoft represent the Good Guys here. Time for them to address “some of the other problems.”

    You really are a colossal ignorant tin-foil hat wearing moron, aren’t you, Peter Dolding?

  8. DrLoser says:

    How can you fix TOOS when you don’t have the code and the EULA forbids it?

    Nota Bene, Robert: this question assumes an axiom.

    Quite a specific axiom, really. I’ll attempt to summarise it as follows: “In order to fix BSOD issues, you need access to the code.”

    Personally, I see this as a completely ludicrous axiom. But I’m willing to discuss the consequences …

    … should you wish to accept that axiom, or re-define it for your own purpose, Robert.

    I’m 99.9999% convinced that you, Robert, are an ignorant evasive coward and will not accept this challenge.

    Mind you, there’s always that other 0.0001% …

  9. Wizard Emeritus wrote, “You wouldn’t accept my talking about my ancient Linux experiences”.

    I have no such objections. I’ve used GNU/Linux since 2000 and have had a few problems too but nothing like TOOS. e.g. malware. I still have not encountered any on GNU/Linux. e.g. re-re-reboots – When TOOS was running a few hours between reboots GNU/Linux would run as long as I wanted. Then there’s cost. Why should I pay anyone for permission to use my own hardware? Why should I have to limit the number of machines on my LAN?

  10. kurkosdr says:

    Dear Pog, nobody cares about your “it BSODs for mysterious reasons!!” talk.

    Didn’t you guys used to have fictional security problems with Windows? At least those were fun to argue about. Where did they go? Or we are supposed to not care about security now because Android is running unpatched for most users (who don’t know or want to flash unofficial ROMs) and hence a mobile security risk?

    http://www.infosecurity-magazine.com/news/80-of-android-users-have-outdated/

  11. Wizard Emeritus says:

    “. XP still has such issues…”

    Ancient history. You wouldn’t accept my talking about my ancient Linux experiences, why should I accept your talking about your ancient experiences?

  12. kurkosdr wrote, “What mythical system is that which caused XP to fill all the virtual memory and crash, but still have a powerfull enough CPU to be usable for day to day tasks? A PIII with 16MB ram?” as if there was any relationship between the two. I’ve never used ~few MB systems since my DOS/TOOS3 system died 15 years ago, before XP was around. XP still has such issues

  13. Wizard Emeritus says:

    “GNU/Linux is eminently more fixable than TOOS which usually just gets re-re-rebooted.”

    Nope, Linux has different problems and issues. Oh and BTW a linux kernel panic requires reboots, and some times “re-re-re-re boots” in my experience.

  14. Wizard Emeritus says:

    “How can you fix TOOS when you don’t have the code and the EULA forbids it? ”

    In most case a BSOD is due to a hardware problem, repair/replace the hardware and the BSOD is gone. No fixing of the OS is involved. At worst case, a call to microsoft tech support would produce a resolution. The same was true for Linux kernel panics.

  15. Wizard Emeritus wrote, “We simply fixed the problems.
     
    And we do the same with windows.”

    How can you fix TOOS when you don’t have the code and the EULA forbids it? GNU/Linux is eminently more fixable than TOOS which usually just gets re-re-rebooted.

  16. Wizard Emeritus says:

    “Nonsense. I’ve seen TOOS crash just idling or running out of memory in normal usage. Linux has never done that for me.”

    Ancient history. Bad/glitching hardware and the occasional bad driver (usually unsigned) are the only times one sees BSOD’s these days. ALL of those events are rare, especially from windows 7 and beyond.

    Talking of BSOD as if it is some deal killer for Windows is nonsense. I have watched Linux kernel panics due to the same problems in a production environment. No one was looking to throw linux out for that reason, We simply fixed the problems.

    And we do the same with windows.

  17. kurkosdr says:

    I’ve seen TOOS crash just idling or running out of memory in normal usage.

    What mythical system is that which caused XP to fill all the virtual memory and crash, but still have a powerfull enough CPU to be usable for day to day tasks? A PIII with 16MB ram?

    Also thanks to FiFi for reminding me that apart from kernel panic, our dear X.org can also take the session with it. Often for no valid reason.

    http://www.tmrepository.com/trademarks/xserverrestartsarenotrealcrashes/

  18. oiaohm says:

    kurkosdr there is a extra set of BSOD causes that Windows systems have that Linux systems don’t.

    Digital rights management drivers. Yes some game and application copy protection under Windows is in fact done by inserting kernel drivers. Yes these do like to cat fight badly at times.

    Driver duplication. Linux has less drivers to cover the same volume of hardware. So Linux has less driver duplication than Windows. This is a factor for stability. Every time a vendor duplicates a driver for a bit of hardware instead of using shared development increases failure rate.

    Also, the reasons that make Windows BSOD are the same that make Linux kernel panic. Bad hardware, voltage fluctuations or bad drivers.
    True but you missed factors why Windows has so many bad drivers compared to the number Linux has. Classes of drivers that end up in kernel space is a factor.

    Linux could start moving more drivers that don’t need ring 0 performance to user-space as this would increase stability. Windows has moved particular drivers to user-space. So number of drivers not in kernel space in one way Windows is ahead here but undone by volume of driver duplication and copy protection drivers Windows suffers from.

    So the reason for Windows BSOD numbers being higher than Linux kernel panics is perfectly explainable. I don’t really see how Microsoft can afford to address the problem.

    Yes the reason for Linux not having a in kernel space stable ABI makes device developers who support Linux look at each others code and share resources maintaining it so getting more stable drivers. This is not something magical.

    Other issue is Microsoft BSOD messages are not clear on what in fact caused the the failure due to the error catch setups so hal or kbd drivers for example show up in BSOD message when in reality it some sub driver that was loaded by them todo something had a error that it did not catch. Linux and OSX don’t have in driver design cascading exception handling so same case in a driver generate a panic naming the poorly designed driver. Yes a NT design selection that makes debugging harder.

    Linux biggest failure is X11 userspace mode blocking kernel panic from being displayed but with this disappearing Microsoft really need to start addressing some of the other problems.

  19. kurkosdr wrote, “I said there are no “mysterious” BSODs that aren’t caused by the same reasons as kernel panics”.

    Nonsense. I’ve seen TOOS crash just idling or running out of memory in normal usage. Linux has never done that for me.

  20. kurkosdr says:

    I can do Google search me too:

    https://www.google.ca/search?q=windows+xp+bsod#q=linux+kernel+panic

    I never said BSODs never happen Pog, I said there are no “mysterious” BSODs that aren’t caused by the same reasons as kernel panics, unless a computer is under your presence, in which case apparently it does fire mysterious BSODs. And only then.

    Meet the ManitobanMan, the man who can make Windows computers mysteriously BSOD (measuring 98 published Marvell comic-book issues, slapped-together popcorn flick currently in the works by Warner Bros)

  21. kurkosdr wrote, “how it’s possible for Windows to work BSOD-less for everyone”.

    It’s not.

    I was using Lose ’95 when I discovered GNU/Linux worked for us, not against us. TLW was definitely used official XP. Came right off a proper installation CD, over and over again.

    Oh, yes, Bill, one of the richest guys on the planet with all kinds of highly skilled employees working for him had endless BSODs during a presentation at CES2005, about when TLW switched to GNU/Linux.

  22. kurkosdr says:

    @Pog

    XP has always being rock-solid for me… no crashes, no BSODs except for once, when an IDSN modem which was made by a Greek company (Intracom) and hence was poorly engineered, had a bad USB driver and needed a driver update.

    Dear Pog, how it’s possible for Windows to work BSOD-less for everyone, tire rotation shops, point-of-sale systems, ATMs, video editing houses, stay-at-home moms editing their facebook profile, and only fire mysterious BSODs (not attributed to bad drivers, bad hardware or voltage fluctuations) around your presence?

    https://www.youtube.com/watch?v=Sm3_qEMTdc4&t=1m50s

    Don’t get me wrong… I can find many faults of Windows… not being able to install all updates in one go if there are too many of them (Microsoft please fix this… I don’t care how many restarts it will take, as long as I don’t have to see mysterious “update failed” messages), windows update being ass-slow in general, system restore rendered inoperable by third-party AVs, Windows 10 being all around sucky and randomly changing file assosciations… But BSODs? Are you sure someone didn’t prank you and put a Windows XP theme (or Vista, or 7, or 8 or 10) logo in a Windows 95 installaton?

    Or you could be just blowing BS out of your mouth…

  23. matchrocket wrote, “I was using a Windows computer. I knew I couldn’t do the same trick with it so I shut the machine down, then plugged in the new mouse and turned the computer back on. It went through its lengthy boot up as usual but then it began installing the mouse. It had to install the mouse?”

    Yes. I had a lab like that. Three different kinds of machines and several kinds of mice. The kids would move mice around and shut down the lab. I replaced the OS with GNU/Linux and that lab purred.

  24. DrLoser says:

    Not entirely sure what happened at the end there.

    However, for completeness.

    I have a reasonable grasp on how to design an OS that handles BSODs and RSODs and for that matter Linux Driver panics and … do we still live in an era of double-panics? … those too.

    And I am wholly neutral on the issues involved. Not MS. Not Torvalds. Not DeHoMag(2015). Do not care.

    Rather shamefully, I find that listening to Dougie making a complete ass of himself is far more appealing than actually doing any useful work.

    After all … if I were to get around to doing useful work … it’s not as though any of you ignorant free-loaders would pay me or my agent or my company or whoever I sold a contract to, is it?

    In lieu of payment, therefore, I select the alternative option of laughing at ignorant incompetents such as Dougie.

    Hey, it’s free! In all four senses!

  25. DrLoser says:

    Re: Also, the reasons that make Windows BSOD are the same that make Linux kernel panic. Bad hardware, voltage fluctuations or bad drivers.

    BS….

    Good point, Dog-Brain. Your first paragraph was indeed masculine cow-patties. May I suggest that next time you hit the delete button, rather than embarrassing yourself by flinging your faeces at the screen?

    Let’s get our terminology right here. Shouldn’t be too difficult, Dog-Brain … after all, before your “hiatus” or “hibernation” or “intercalculation” or “intermission” or “suspension of studies” or …
    … Oops, the term for your current “suspension of gainful employment,” according to you, is … what’s that other thing?
    Ah yes, in Collegiate terms, it’s a “Sabbatical.” I’m not wholly convinced that, as somebody who is proud to lack a High School Equivalency, that you would really qualify for a “Sabbatical,” Dog-Brain.
    Or, as it happens, for a “suspension of studies.”
    Or for, and let us cut this badinage short, an “intercalculation.”

    Nevertheless, I feel your own hopeless attempts to conceal a regrettable lack of employment prospects are instructive in this case, Dougie. You want to know why? Well, it’s all about categorization.

    See, the thing about BSODs is that they genuinely are “the same that make Linux kernel panic. Bad hardware, voltage fluctuations or bad drivers…” Do you have a better solution for a “generic” OS than to panic in these circumstances?*

    Well, Dog-Brain, clearly you have a lot of spare time and nothing much else to occupy it, except for the contemplation of infinite futility. I’d better help you on that. You seem like a resourceful sort of chipmunk.

    Why not have a go at Red-SODs?

    [* I do. But I’m absolutely certain that nobody here reaches my pay-scale on this advanced, professional stuff. (It’s to do with how one designs the OS kernel, btw.]

  26. matchrocket says:

    I’ve been using Linux for over 10 years now. I’ve gotten use to switching the (USB) mouse on-the-fly. While the computer is running. I’ve done it many times now and Linux doesn’t miss a beat. One time, however, I was using a Windows computer. I knew I couldn’t do the same trick with it so I shut the machine down, then plugged in the new mouse and turned the computer back on. It went through its lengthy boot up as usual but then it began installing the mouse. It had to install the mouse? Sure enough it went looking for a driver, found it, installed it and then? Reboot! Reboot? For a friggin’ mouse? You’ve got to be kidding me. At least it was only one reboot. Oh, the manufacturer of the mouse? Microsoft.

  27. dougman says:

    Re: Also, the reasons that make Windows BSOD are the same that make Linux kernel panic. Bad hardware, voltage fluctuations or bad drivers.

    BS….

    That may be the case in very few occasion as hardware deteriorates, but the majority of the time, the main culprit is Microsoft software.

  28. kurkosdr wrote, “Also, the reasons that make Windows BSOD are the same that make Linux kernel panic. Bad hardware, voltage fluctuations or bad drivers.”

    Well, why could a bunch of “amateurs” back in the day get good uptime on random hardware while M$ could not? BTW. Our hardware was perfect. The machines ran six months without a hiccough after I switched to GNU/Linux. I’ve not seen any such problems on all the conversions I’ve made over the years with dozens of types of machines exposed to wild users. Meanwhile TOOS failed everywhere. The reason TLW switched to GNU/Linux was that I finally said “No More!” to re-installing her OS from M$. In the beginning, XP was failing weekly. Lately it was doing better but it was still failing. GNU/Linux just keeps humming.

  29. kurkosdr says:

    Dear Pog, you know that GNU/Linux is not advanced enough to BSOD, right? All you get is a “kernel panic”.

    Also, the reasons that make Windows BSOD are the same that make Linux kernel panic. Bad hardware, voltage fluctuations or bad drivers.

    IMO Windows 7 and Windows 8 are the best OSes a person can have.

    Windows 10? It’s as if Microsoft saw that flash animation of “Windows RG” and said “hey? I guess we could do this”

  30. matchrocket says:

    What a great story. For every one like that told, there are many more untold.

Leave a Reply