Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×
Microsoft IT

Global IT Outage Linked To CrowdStrike Update Disrupts Businesses (techcrunch.com) 274

A widespread IT outage, caused by a defective software update from cybersecurity firm CrowdStrike, is affecting businesses worldwide, causing significant disruptions across various sectors. The issue has primarily impacted computers running Windows, resulting in system crashes and "blue screen of death" errors. The travel industry appears to be among the hardest hit, with airlines and airports in multiple countries reporting problems with check-in and ticketing systems, leading to flight delays. Other affected sectors include banking, retail, and healthcare.

CrowdStrike CEO George Kurtz confirmed the outage was due to a "defect" in a content update for Windows hosts, ruling out a cyberattack. The company is working on a fix. CrowdStrike said the crash reports were "related to the Falcon Sensor" -- its cloud-based security service that it describes as "real-time threat detection, simplified management, and proactive threat hunting."

A Microsoft spokesperson told TechCrunch that the previous Microsoft 365 service disruption overnight July 18-19 was unrelated to the widespread outage triggered by the CrowdStrike update.

Editor's note: The story has been updated throughout the day and moved higher on the front page.
This discussion has been archived. No new comments can be posted.

Global IT Outage Linked To CrowdStrike Update Disrupts Businesses

Comments Filter:
  • by BlacKSacrificE ( 1089327 ) on Friday July 19, 2024 @05:53AM (#64636983)

    From CrowdStrike's support portal:

    1. Boot Windows into Safe Mode or the Windows Recovery Environment
    2. Navigate to the C:\Windows\System32\drivers\CrowdStrike directory
    3. Locate the file matching “C-00000291*.sys” and delete it.
    4. Boot the host normally.

    Nothing to do with Microsoft (for once)

    • by BadgerStork ( 7656678 ) on Friday July 19, 2024 @06:01AM (#64636995)
      1a. Enter admin password that you don't have?
      • by thegarbz ( 1787294 ) on Friday July 19, 2024 @08:26AM (#64637277)

        1a. Enter admin password that you don't have?

        Two scenarios here:

        1. You have your computer, which means your Microsoft account is your admin account automatically and you can log in locally with your email address.
        2. Your computer is managed by IT and it's 100% their problem, go back to playing with your phone until they sort it out.

        The problem here for number 2 is that you can't push out a remote update if your computer wont boot so there's a shitton of manual intervention underway.

        • by e3m4n ( 947977 ) on Friday July 19, 2024 @09:06AM (#64637393)

          The hospital my wife works at has called in everyone who works in IT. This includes the training departments who arent computer technical on anything other than the software they train and the tools for training. Even the managers are going in to do fixes. My guess is they will have flash drives and batch files to make it as easy as possible.

          • Re: (Score:2, Insightful)

            by chipperdog ( 169552 )
            Hospital computers should not be able to boot from a thumb drive, nor should a thumb drive be mountable on a running system,
            • by e3m4n ( 947977 )

              Who said anything about BOOTING from a thumb drive? Do you always jump to conclusions? You will never be command material if you keep that up. You have a thumb drive with a batch script so people dont typo the delete command. That expands the number of people rolling out the fix.

          • Today a thumb drive is useless for boot to recover due to bitlocker unless the thumb drive has the key.

            At my workplace it's only the servers that had Crowdstrike so we whad it relatively easy. I feel sorry for all organizations with Crowdstrike on the clients.

            What worries me is a situation where bitlocker goes fubar at an update.

            • by e3m4n ( 947977 )

              You need to reread my post. I didn’t say shit about booting. the fix involves booting into windows recovery mode. When a hospital has over 10,000 workstations and they’re calling in everybody that works in IT, even the ones that are not Technical to do the fixes, you need a batch script so they don’t fuck shit up. So you boot into recovery mode, you plug in your thumb drive, and you execute the batch script so there’s no typos and dumb shit like that. It also speeds up the repair pro

            • by MachineShedFred ( 621896 ) on Friday July 19, 2024 @11:43AM (#64637939) Journal

              The best one is Crowdstrike endpoint security + remote workers.

              Have fun explaining to project managers and marketing folks how to boot to a recovery console, type in that monstrous string of bitlocker recovery key, and how to navigate a command prompt in recovery mode.

              UPS and FedEx are going to have a revenue spike as literally millions of laptops get shipped to remote employees behind this bullshit. By the way: this is why you canary updates if you know what you're doing.

        • 2. Your computer is managed by IT and it's 100% their problem, go back to playing with your phone until they sort it out.

          The problem here for number 2 is that you can't push out a remote update if your computer wont boot so there's a shitton of manual intervention underway.

          You mean the IT guys are going to have to leave their enclaves and go computer-to-computer to fix the problem? Oh noes!

          In some companies I know, getting any support person to come out of their seemingly perpetual meetings for an actual laying-on of hands requires prayers, incantations, and possibly a goat to sacrifice.

          • You mean the IT guys are going to have to leave their enclaves and go computer-to-computer to fix the problem? Oh noes!

            You do know people work from home, right? For a company that could be mean thousands of homes that a single IT guy would have to visit. Do you think that is practical? Or a more obvious solution is everyone has to go into the office to get their computers fixed with maybe IT guys visiting the few that cannot.

      • by chill ( 34294 ) on Friday July 19, 2024 @09:36AM (#64637495) Journal

        No, that's possibly Step 2.
        Step 1a is going to be enter the BitLocker recovery key you probably don't have. Remote laptop users are fucked.

        • No, that's possibly Step 2.
          Step 1a is going to be enter the BitLocker recovery key you probably don't have. Remote laptop users are fucked.

          the TMP should have the key

    • by AleRunner ( 4556245 ) on Friday July 19, 2024 @06:22AM (#64637027)

      Could we please post references? People should not be deleting random files from their computer based on a comment on Slashdot. Despite this having caused multiple major companies to stop working, Microsoft has nothing on their main page!

      Quick seach finds a Forbes article which confirms this fix https://www.forbes.com/sites/k... [forbes.com]

      Nothing to do with Microsoft (for once)

      Wrong. This kind of problem is inherent in the user hostile design of Microsoft systems which force you to use third parties in order to get an approximately working operating system from a Windows install.

    • by arglebargle_xiv ( 2212710 ) on Friday July 19, 2024 @06:41AM (#64637063)

      You forgot a bit:

      0. Enter the Bitlocker key that you need to boot into safe mode. Bill the IT guy probably still has a copy of that somewhere. Oh, but he left the company three years ago to join an ashram in India.

      • Re: (Score:2, Insightful)

        by Anonymous Coward

        That even assumes it was set by a human and not randomly generated and associated with the domain/azure ad/elantra/mdm or whatever the duck cloud account that has been enforced because cloud is good.

        • you can get the key from the domain for each system as long as save backup keys to domain was done as part of the setup.

    • Or, as related in this article [marketwatch.com]:

      “CrowdStrike is actively working with customers impacted by a defect found in a single content update for Windows hosts,” CEO George Kurtz said in a statement.

      “Mac and Linux hosts are not impacted. This is not a security incident or cyberattack. The issue has been identified, isolated and a fix has been deployed.”

    • by gweihir ( 88907 ) on Friday July 19, 2024 @06:48AM (#64637081)

      Nothing to do with Microsoft (for once)

      Not quite. Their pathetic design of MS Windows is what makes tools like Crowdstrike necessary and what requires them to be placed deeply in the systems internals.

      • by Junta ( 36770 ) on Friday July 19, 2024 @08:09AM (#64637227)

        Crowdstrike supports Linux and Mac too.

        Maybe 20 years ago Windows design and track record produced an environment of companies demanding tiger repelling rocks today. However now all platforms are subject to "needing" this stuff, and I would say that modern Windows design is largely competitive with macOS and Linux for security enforcement facilities.

        Now the largest sincere driver of these solutions is the fact that:
        -The desktop paradigm is still largely all applications are treated equal, so a random text editor and your web browser have equal access to each other's content and networking.
        -Users are really lazy and will make remote access poorly secured

        • by Ol Olsoc ( 1175323 ) on Friday July 19, 2024 @09:01AM (#64637381)

          Crowdstrike supports Linux and Mac too.

          Yeah - but Linux and Mac didn't have this issue.

          Maybe 20 years ago Windows design and track record produced an environment of companies demanding tiger repelling rocks today. However now all platforms are subject to "needing" this stuff, and I would say that modern Windows design is largely competitive with macOS and Linux for security enforcement facilities.

          Gotta respectfully disagree. One file, deeply buried within an OS, provided by a third party, that can bring systems using windows computers to BSOD or not boot at all - worldwide- simply has no defense, yet people are trying to defend it. It also shows an attack vector for someone that wants to go scorched earth worldwide I'm pretty concerned.

          • by Junta ( 36770 ) on Friday July 19, 2024 @09:16AM (#64637433)

            Without knowing the nature of the bug, hard to say for sure, but vaguely the concept of a third party add-on making kernel panics or unbootable system is *easily* imagined in Linux (and presumably MacOS).

            Looking it up, Crowdstrike in Linux requires a kernel module of their design to be loaded. In fact, they don't even provide source code seemingly:
            https://github.com/CrowdStrike... [github.com]

            So yes, Crowdstrike could easily make a screw up that renders Linux systems panicking and "unbootable" (note that Windows can still boot safe mode, and analagously, you should have some rd.break or similar scheme to avoid loading a buggy 'falcon' kernel module they would have provided. They happened to screw up Windows today, but next time it could be Mac or it could be Linux. The issue is that a mistake was made by a third party that demands kernel level access to whatever it is they are doing.

            • which is why many distros allow you to boot into single user mode where such kernel modules is not loaded
              • by Junta ( 36770 )

                Yes, like "safe mode", and why I mentioned 'rd.break' as an example of getting to an analogous state. Even if they for some reason injected gorp into initramfs, rd.break=cmdline, for example, would preempt *all* kernel module loads in dracut style initramfs.

          • by deKernel ( 65640 )

            Well, the OS might not be an issue, but you sure can blame MS people for rolling such an update out in such a fashion that it takes the whole Central region down. Sounds like someone didn't have a proper testing process before rolling out such deep changes as a driver can be.

            • Why are we blaming Microsoft for an entity not named Microsoft rolling out a shitty update to their 3rd party software that isn't maintained, distributed, or otherwise acknowledged as existing by Microsoft?

              Just curious.

              I'm all for shitting on Microsoft for their numerous and constant fuck-ups, but this isn't that.

    • If you need to boot into safe mode for the system to get far enough to apply a fix, it rather begs the question of how exactly either Microsoft or CrowdStrike are going to "fix" this. The fix has been identified and applied on the CrowdStrike server side, but any system that downloaded the defective channel update is dead until someone competent enough to apply this fix gets to it.

      • by e3m4n ( 947977 )

        Safe mode and a batch script to delete the offending module.

        • If the effected systems are critical enough to use this super-security Crowdstrike product it would be negligence for those systems to also be configured to allow for easy remediation (i.e. booting off external media, file access to the appropriate folders, drive encryption, etc.

          Crowdstrike will be bankrupted by this. This update intentionally evaded policies that orgs put in place to prevent automatic updates and they used a secret back door to push it out.

          • by e3m4n ( 947977 )

            At the hospital, my wife works at, all of the desktop machines are affected. They all have a blue screen of death. These are simple workstations.

    • by EvilSS ( 557649 )
      I feel bad for the sysadmins at large companies this morning. Imagine having to a manual process like that on 20,000 PCs.
      • For on-site machines, if PXE boot is enabled it can go pretty fast, boot to a script that does the fix.
    • by evanh ( 627108 )

      I don't understand how come all these large scale commercial systems can be apparently just blowing in the wind. Isn't there IT staff testing before deploying the updates?

      • by ThosLives ( 686517 ) on Friday July 19, 2024 @09:24AM (#64637451) Journal

        I was going to say this... there are at least two points of massive process failure here:

        How did this get released in the first place?

        Why does any IT department have auto-updates enabled at all? Any update can break a system... they really really need to be vetted.

        It's like the software industry is trying to get itself legislated to require something like Professional Engineer licensing. After all, IT infrastructure is at least as important now as civil infrastructure, which already requires PE.

        When we have daily system outages and software breaches... this really needs to be addressed. I had hoped that the industry would have been able to keep itself under control without regulation, but we software engineers have not demonstrated the ability to maintain due diligence. It's a shame that for those that do execute well, because now they're soon likely to have regulatory burden.

        • by khchung ( 462899 ) on Friday July 19, 2024 @09:46AM (#64637527) Journal

          Why does any IT department have auto-updates enabled at all? Any update can break a system... they really really need to be vetted.

          Because these updates included patches for zero-day exploits that, if they waited for the vetting process, could result in compromised machines?

          Unfortunately for admins, they are between a rock and a hard place, there is no right way to support a lot of Windows machines without hitting problems down the line.

        • does CrowdStrike have WSUS like control of role outs of updates? Like with test groups, stage groups, prod groups of systems?

        • It's like the software industry is trying to get itself legislated to require something like Professional Engineer licensing.

          That wouldn't change anything. The people behind this outage all would have gotten such a license easily.

    • by 2TecTom ( 311314 )

      classism leads to corruption which breeds incompetence and irresponsibility

    • by CEC-P ( 10248912 )
      Have fun, everyone with Bitlocker. Actually, does RE bypass that? Somehow that never came up. Oh well, have fun everyone doing it with PE's or using Veracrypt instead.
    • delete C:\Windows\System32\
    • First deal with the update windows Explorer prompts, then the patch prompts, etc....

  • by ls671 ( 1122017 ) on Friday July 19, 2024 @05:54AM (#64636987) Homepage

    Microsoft said in a statement Friday that it was aware of the global outage that was affecting Windows devices,

    That's reassuring, thanks!

    • by leonbev ( 111395 )

      It is kind of weird that there zero mention of this issue and how to fix it if you go to support.microsoft.com. The first thing that you get when you go here is a ad for Copilot. While being totally on-brand for Microsoft to try to sell you more service add-ons in 2024, it's kind of annoying.

      • Support.microsoft.com is a generic landing page for all Windows users, not a breaking news site, not a site for managed windows installs, not a site for people who are running Crowdstrike. Why would it cover it? The overwhelming majority of Windows users do not run Crowdstrike.

        If you want news about why the cloud was down the actual site is https://status.cloud.microsoft... [status.cloud.microsoft] which at the very top gives you information about Crowdstrike. That is 100% of what Microsoft is responsible for.
        If you want to know ab

        • If you want news about why the cloud was down the actual site is https://status.cloud.microsoft... [status.cloud.microsoft] which at the very top gives you information about Crowdstrike. That is 100% of what Microsoft is responsible for.

          Microsoft is 100% responsible for the fragility of its OS, especially since much of it is related to deliberate decisions they made in the design of the kernel for NT4, which are still affecting us to this day.

          • by DarkOx ( 621550 )

            Not really - Microsoft has more or less fixed the issues since NT6.x kernels. The problem is government.

            Microsoft would have (and rightly so) killed off the entire third-party AV/EDR industry. The could not and did not though because dumb dumbs like you and rlsivergun would be here screaming about anti-trust, and some Pocahontas..err uh Liz Warren rather would be dumb enough to listen to you.

            • Microsoft would have (and rightly so) killed off the entire third-party AV/EDR industry.

              You mean with their product that is neither faster nor better at detection than most of the competition?

              screaming about anti-trust

              Antitrust is real, it really affects consumers, and Microsoft was found by the USDOJ to have done basically everything which is considered antitrust.

        • considering just how large an outage this is, I for one would argue that support.microsoft.com, which is a site I guess many would first go to, should have some info on it regarding this (and the fix).
  • by gavron ( 1300111 ) on Friday July 19, 2024 @06:02AM (#64636997)

    When you sleep with someone you're not just sleeping with them.
    You're sleeping with everyone they've slept with.

    When you run a business, whether a multi-billion dollar airline or the health network of the UK or all the airports in Spain or the Berlin airport or not, have a technical plan with backups. That way if you outsource your tech and your "outsource partner" is a failure, you have a backup.

    Don't outsource to Microsoft. They've spent every year since the 1990s (that's 30+ years!!!) ensuring their sofware is backward compatible so you can run DOS games, but in the process introducing more holes, bugs, and security holes than any other software firm on this planet PERIOD. Don't outsource to a "partner" that can't be trusted to be online and safe.

    Microsoft knows they suck. So they outsource "security" to Crowdsource, and Crowdsource just knows their farts don't smell so they go around crowing about how awesome they are and how their stock is going up. I'm sure investors are excited today to find out that Crowdsource fucked it up royally. Microsoft is excited that their "security partner" sucks. Trillions of dollar companies including most major airlines, airports, health networks, and even food chains and stores who cound on their "cloud partner" are 100% offline.

    "Degraded mode" means you've failed as an IT person. Print your resume... IF your printer is local and doesn't require the "Cloud" to auth, and use the manual door to exit the building.

    Don't let the door hit you in the ass. Hopefully you have your own car... rideshare requires the cloud.

    Like dominos in a pile, all we have left is a pile.

    • Don't outsource to Microsoft. They've spent every year since the 1990s (that's 30+ years!!!) ensuring their sofware is backward compatible so you can run DOS games, but in the process introducing more holes, bugs, and security holes than any other software firm on this planet PERIOD.

      Is that still true? I thought they had dumped that particular principle with Windows 10, I know there is a lot of older software around which either had reduced functionality or stopped working altogether at that point. There

      • Is that still true? I thought they had dumped that particular principle with Windows 10

        In fact they dumped it with Vista, A lot of Windows software from before then won't run on it. They also dumped it with 64 bit, where the virtual machine used to provide backwards compatibility will no longer run 16 bit software even though amd64 has features which enable that (and there is even a free 3rd party replacement which does it.) It did get a lot worse again in 10, though.

        This brings up the question, why run Windows on servers at all? Windows makes a reasonable amount of sense for desktops if you

    • by gweihir ( 88907 )

      Indeed. A halfway decent OS does not need things like Crowdsource.

  • That might be a first for Microsoft.

  • by Megane ( 129182 )

    Just use the old three R's:

    1. Reboot
    2. Reinstall
    3. Red Hat

    (not that I'd use deadrat these days, I told you it was the "old" three R's)

    • That sounds like it would take a really long time. Seems easier just to fix the actual problem by deleting a single file on your PC.

    • by mjwx ( 966435 )

      Just use the old three R's:

      1. Reboot
      2. Reinstall
      3. Red Hat

      (not that I'd use deadrat these days, I told you it was the "old" three R's)

      I'll propose an update.

      1. Reboot
      2. Reinstall
      3. Reimage with Linux Mint

      This will still work with your distro of choice.

  • Quality Control (Score:4, Insightful)

    by jd ( 1658 ) <imipak&yahoo,com> on Friday July 19, 2024 @07:05AM (#64637127) Homepage Journal

    A failure on this scale should simply not be possible.

    1. It IS Microsoft's fault, in that Windows is demonstrably fragile. It should be able to cope with buggy files, gracefully, and it clearly can't. Windows is used on critical systems, including in intensive care wards. You simply can't afford this level of fragility in something that is used in such a manner.

    2. The buggy software should never have been released. This sort of error should have been picked up in QA, given that it seems to always happen.

    OSS software that is own-risk is one thing, but commercial software needs to be released to far higher standards and the "no warranty" thing for commercial software needs to go. The manufacturer should damn well be held liable.

    • Re: (Score:3, Insightful)

      by Sloth77 ( 6645172 )
      1. No, sorry but protecting against a bug in a kernel driver running in ring 0 with no address space protection is not possible. You would get exactly the same type of crash with a bug in a Linux kernel module.

      2. Agree completely
      • You would get exactly the same type of crash with a bug in a Linux kernel module.

        I wouldn't because a) I would be getting that kernel module from my OS vendor who would do actual QA on the whole system before releasing it and b) because my particular OS vendor provides a way to step back to a couple of previous configurations of the kernel if they do break something.

        I actually had an equivalent problem with a new WiFi driver at one point which broke with my particular, rare, WiFi card (thus the QA failure) and simply stepping back to an older kernel release allowed me to work around it

    • It is not a random "buggy" file. It is a file that tells crowdstrike how to recognise dangerous code. The contents of the file is incorrect but neither the crowdstrike code nor Microsoft can possibly now that. And because the contents of that file is incorrect, crowdstrike wrongly believes that some required code is dangerous and stops windows from running. Exactly what it should do if that required code had indeed be hacked.

      What you are saying is like âoewindows is broken because it doesnâ(TM)
      • by jd ( 1658 )

        But since it always does this, any actual test of the code would have revealed that there was a showstopper bug. It's kinda difficult to miss.

    • 1. It IS Microsoft's fault, in that Windows is demonstrably fragile.

      Sorry but that is horseshit. Literally any OS on the planet can be rendered unusable by 3rd party software that has administrator privileges to modify system files. At the level that CrowdStrike runs on your machine, it can just as easily wipe your /boot partition. Or load a kernel module on boot that causes a panic. Or cause a recurring MDM reset on your Android / iPhone (I have had this happen).

      No OS is safe from software that is implicitly trusted to manage your OS.

      • by sinij ( 911942 )

        Literally any OS on the planet can be rendered unusable by 3rd party software that has administrator privileges to modify system files. .

        Please don't ruin /. favorite pastime of hating on MS with your reason and logic.

    • 1. Bullshit.
      I've experienced plenty Linux-based machines not booting up after a 3rd party driver/software update.
      The last time something similar happened? A couple weeks ago, with a Home Assistant update from 2024.6.4 to 2024.7.2, which broke critical services. I admit it DID boot-up, but it was largely unusable.
      Even after reverting to 2024.6.4 from backups, beside losing some data, that particular system is still riddled with issues (some integrations no longer work, historical data randomly stops recordin

    • by e3m4n ( 947977 )

      Security works against itself. Think about GRUB booting in interactive mode. At some point it was determined that allowing a local user the use bootable media or bypass modules was a security risk. So they started rolling boot security into the BIOS and UEFI. This prevents a simple fix of booting and telling the kernel to skip certain sick modules that prevent you from booting. So a design to prevent this sort of clusterfuck inherently introduces a risk to security that also presents a problem. Bitlocker is

    • 2) is impossible to prevent, therefore you have to have a method to recover from it.

      The only recovery I can think of is before doing the stupid fucking update, once you got the user's explicit permission to update, automatically make a copy of the OS image, so that in case of failure to boot it copies the image onto the current, now buggy, OS, and reboots.

      That also means no stupid fucking update every fucking day. Once a month at most and let everyone know, and let them opt out of it.

  • by xack ( 5304745 ) on Friday July 19, 2024 @08:36AM (#64637303)
    Many games already cause BSODs with conflicting DRM and Anticheats, it's only a matter of time that it migrates to the corporate world. Combine that with Sony deploying anti-Proton patches to their games and people have no choice but to use Windows, and be subject to constant unreliability. The internet already has disease like symptoms with all the malware bots out there, we will eventually seen an attack where a billion computers get bricked all at once. I don't fully trust Linux either, expect flatpak or snap to get compromised soon.
  • This is not a "Microsoft outage," it's caused by a third party tool that is used by a huge swath of major institutions across the West.

  • From what I read, CrowdStrike's a Falcon Sensor update is causing BSODs, so is this a Microsoft issue or CrowdStrike?
  • 40 years of non-stop markets consolidation is that there is so few players so everybody uses the same software and when there's a large issue with a single software package instead of a few barely noticeable blips the entire world grinds to a halt...
  • by JustAnotherOldGuy ( 4145623 ) on Friday July 19, 2024 @09:42AM (#64637509) Journal

    Crowsdstrike: Keeping You Safe From Accessing Your Data

    Our motto is, "They can't load malware if you can't login!"

  • Do you mean the forced update solely impacts Windows computer?
  • -Obsession with having the absolute latest software update at the expense of testing time and quality control.

    -"Security" bloatware that causes more problems than it protects against.

    -Large organizations that could reasonably provide functions on-premises moving every function and service to the cloud with real-time dependence on third parties. It's one thing to buy software from a third party and run/update it locally. It's another to rely completely on that third party being up and managing their own up

  • In my opinion companies should have some workstations running Windows, some Linux, and some FreeBSD, and some MacOS, about equal shares.

    What is the problem with that? Yes, there would be some disruption with loosing 25% of workstations, but 75% would be still working. People would be standing in lines, but it would be still much better that this.
  • After Microsoft's whoopsie this and now this. I've been called a fossil for insisting on non cloud based security and computer access. Any company can do this but they are unwilling to. They seek some rainbow marshmallow candy land in the cloud. Forgetting the second its in the cloud its not in their control.
  • by BigFire ( 13822 ) on Friday July 19, 2024 @10:46AM (#64637707)

    Advance incompetence is indistinguishable from malice and all.

  • I've watched for decades while smart, dedicated, educated, experienced people have invested enormous amounts of time and money desperately trying to make Windows systems function -- where "function" includes not just performing computing tasks, but performing them securely. All of them have failed. Some of them took longer to fail than others, but all of them failed.

    It's not hard to figure out why: Windows isn't designed to be secure. It's designed to maximize Microsoft's revenue, and everything about it is dedicated to that goal to the exclusion of all others. And when viewed through that lens: it's quite successful. They've made a fortune. And so have all the parasites -- like Crowdstrike -- who've attached themselves to it like remoras on a shark. The secondary market for products and services marketed to do what Microsoft didn't do and will never do is also quite successful. They've made a fortune too.

    I could enumerate all of the specific technical reasons why this happened, but others have already chronicled the litany of problems that goes all the way back to the initial release of Windows and constitutes an unbroken record of failure. So I'm going to skip it in favor of making my point: if you can't compute securely, then you can't compute. And you can't compute securely using Windows. Everyone who's tried has failed, and there's no reason to think that you're going to succeed where everyone else has failed.

    (Of COURSE that doesn't mean that by using something else you're guaranteed success: there are no guarantees. But using something else gives you a fighting chance. I know: I've been doing it for a long, long time. And the last time I had to deal with a security incident was November 2, 1988 -- a date that you should recognize. )
  • by JoeyRox ( 2711699 ) on Friday July 19, 2024 @11:18AM (#64637809)
    I have to think they're going to face billions in lawsuits over hits, no matter what their agreement says about being held liable for software defects.

All theoretical chemistry is really physics; and all theoretical chemists know it. -- Richard P. Feynman

Working...