Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Microsoft suggests command line fiddling to get Windows 10 update installed

theregister.com

Post reply on HN
Print view View original

201–210 of 223 posts

Re: Microsoft suggests command line fiddling to get Windows 10 update installed

#201

Post by iforgotpassword »

brettermeier wrote: post #74
squarefoot wrote: post #46

FTA: "Showing an unhelpful error message and then requiring a user to delve into the world of the command line to fix things. What is this? Linux?" What did the author want, a button to push for each solution to every possible exotic error, even critical ones that would require to resize a system partition from a running OS? Seriously!? Of course one solve this through the command line. Or, if that wasn't what the au…

The beauty of Windows is that you have one-button-push-solutions that just work. If Linux would have that too, maybe even tech-unsavvy-users could handle it, but since Linux is made from freaks for freaks that won't happen. I mean, Linux people can't even see the problems of a not-one-click-solution, but for every normal user it's crystal clear. Command line is for tech users and really for nobody else. Nobody who is…

Sure, and I wouldn't want it any other way, or I'd just use Windows in the first place. That said, on Linux I still get error messages, not a big number – an actual message that tells me what didn't work so if googling that error message doesn't yield a solution, I can go ahead and try do fix the issue myself.

I still remember from the Windows days when you were unlucky enough to hit a more unusual error code, where googling it didn't immediately result in a dozen forum threads with people posting the same issue and then some dude replying with a fix where you don't even know how they possibly figured that out. If you didn't find a solution googling the error code it basically meant you can try random shit until you accidentally fix it, or reinstall. On Linux I could just ultimately look at the source code of whatever component is broken and track back from where the error is printed.

Re: Microsoft suggests command line fiddling to get Windows 10 update installed

#202

Post by concerned_user »

currysausage wrote: post #181
Zenst wrote: post #14

I've followed the instructions and understand that they may seem overwhelming for some users. Instead of using command line methods, I opted for Windows Disk Management to shrink the drive's partitions and create the necessary space. It's surprising that a script hasn't been provided for this process, which suggests the complexity and potential pitfalls involved. This leads me to believe that the absence of a scripte…

It looks like a PowerShell script is available now, but I don't think it's changing the partition size. Has anyone tried this out? https://support.microsoft.com/en-us/topic/kb5034957-updating...

I can only see it using dism to apply the patch but there is no commands to manipulate partition size so if the cause is lack of free space it will fail same as update did.

Re: Microsoft suggests command line fiddling to get Windows 10 update installed

#203

Post by crimbles »

GTP wrote: post #199
crimbles wrote: post #87

Earlier quoted context omitted.

Depends really. If it's some kind of runtime or VM you pretty much have zero idea what the exit code was. Python, CLR, JVM tend to leave you in this mire.

Yes, but remember that usually nowadays a piece of software also logs something to stderr instead of relying just on exit codes to signal error.

Not necessarily. Most of the 139's I get these days are SIGSEGV related. Obviously if you annoy the MMU at that level it's not going to get given an opportunity to write anything useful to stderr.

Re: Microsoft suggests command line fiddling to get Windows 10 update installed

#204

Post by kotaKat »

sixothree wrote: post #200
kotaKat wrote: post #105

Earlier quoted context omitted.

You have to wait for all the back-end updates to finish. Install all the Windows Updates, then open Microsoft Store and go to the Library and wait for it to update even more shit, then you finally will have Terminal and WinGet. Yeah, I found that one out the annoying way too.

Maybe I was experiencing something else on top of that because I was getting exception codes at first. But then once I started getting "source errors" I decided to check the store and such and that's when things started working more normally. What's surprising though is that it listed the msstore source as operating normally when I did winget source update. Regardless, kind of crappy this doesn't work out of the box…

My experience (and path that lead to this) was realizing that a fresh install of Windows 11 gave me PowerShell and not Terminal in the Win+X menu.

WinGet would... launch, but provide no output. And of course, it was my first time trying to use WinGet in general...

A few reboots later suddenly Win+X had Terminal and WinGet started working, and on a subsequent reinstall in a VM I spotted that, yeah, apparently all the extra WinGet, Terminal, etc stuff all has to be pushed from the MS Store instead of Windows Update.

Total pain in the ass.

Re: Microsoft suggests command line fiddling to get Windows 10 update installed

#205

Post by mook »

reactordev wrote: post #86
maximus-decimus wrote: post #62

Earlier quoted context omitted.

This is how I imagine the error code looks. How else can you even return the wrong error? ```c# public class FooError { public Message(){ //TODO: I copy pasted this from somewhere else. I still need to update to a meaningful message return "this is a BarError" } } ```

In C/C++ land, errors are often mapped to const size_t int values. Whether it's an enum or just a list of "codes". In this case, I suspect (like OpenGL) it's just a giant list of codes defined as "extern const uint32_t MY_SUPER_ERROR = 0x0001" and some other part of the code has defined the same value to another error type. Then it's trying to value match and give a human readable value (0x001) but based on the match…

0x80070643 is ERROR_INSTALL_FAILURE, so yeah basically somebody clobbered the error code in some installation routine. It's in winerror.h, according to https://learn.microsoft.com/en-us/windows/win32/debug/system...

Re: Microsoft suggests command line fiddling to get Windows 10 update installed

#206

Post by reactordev »

mook wrote: post #205
reactordev wrote: post #86

Earlier quoted context omitted.

In C/C++ land, errors are often mapped to const size_t int values. Whether it's an enum or just a list of "codes". In this case, I suspect (like OpenGL) it's just a giant list of codes defined as "extern const uint32_t MY_SUPER_ERROR = 0x0001" and some other part of the code has defined the same value to another error type. Then it's trying to value match and give a human readable value (0x001) but based on the match…

0x80070643 is ERROR_INSTALL_FAILURE, so yeah basically somebody clobbered the error code in some installation routine. It's in winerror.h, according to https://learn.microsoft.com/en-us/windows/win32/debug/system...

“What’s a null?”

“Well, it’s the absence of a value, or 0, or nullptr, or…”

“Forget it, you’re insane.”

- My son.

Re: Microsoft suggests command line fiddling to get Windows 10 update installed

#207

Post by gverrilla »

swells34 wrote: post #162
jstummbillig wrote: post #128

Earlier quoted context omitted.

In that case I will have to disagree. I run a stock, up to date Ubuntu. Yes, there is a GUI for a lot of things, but I will most certainly run into something that will require me to drop to the CLI before long – which I am fine with, but I am not going to pretend it won't inevitably happen. My experience is: On Windows, for most enduser use cases (and this absolutely excludes anything dev related), I would expect the…

I disagree. On stock Ubuntu, there is no normal situation you'll run into where you must use a terminal. If you decide to do something that requires a terminal, that's a different story. But basically every normal operation you'd do on windows GUI (and more...), you have a GUI option available on Ubuntu; it is the Windows of the Linux world, and it behaves as such.

Cmon my ubuntu stock can't even remember my sound input options, which are reset on each reboot. There's no option for that on the GUI. Also upgrading linux can be a big pain for non-experts. I've been using dualboot windows/linux for a long time, and I simply don't upgrade ubuntu: I'll do a fresh install if I want/need the new version.

Re: Microsoft suggests command line fiddling to get Windows 10 update installed

#208

Post by swells34 »

JoeAltmaier wrote: post #151
swells34 wrote: post #149

Earlier quoted context omitted.

Yes, and some people just want to take their car to the mechanic, while other people want to fix theirs. It's a classic struggle in any area. The big difference though, is if I use my car in a professional capacity , I want to know how to change my own oil, because I'm not an idiot. I want to be able to fix simple issues, because that's just being professionally responsible. Windows discourages this kind of professio…

Might be a bit of hyperbole. Microsoft had extensive documentation of nearly every API they support, and some they don't. Comprehensive documentation with notes and caveats. Automatically available from their tools. Compare, for instance, with Apples almost total non-documentation of anything. Mostly just scraped comments with arguments barely described e.g 'a string'. No notes, no architecture, nothing.

Yes, my comment contained more hyperbole than it should have; I was frustrated at the idea of "hiding information about a problem" as a good design goal, and let that frustration influence my writing. Apologies to jstummbillig for that. Thank you for being the voice of reason in the room, Joe.

Interesting, I've never found Microsoft documentation for Windows thorough enough to reason about a problem, especially when compared to Arch or Gentoo's docs. I always feel like I'm reading the abstract of a paper, with no access to the rest of it. Is there some kind of special access or login requirement for deeper documentation, or am I just spoiled by the great Linux docs?

Apple products are the king of the "you don't own your device" mantra. I just refuse to develop anything for them, because the time expense of trying to guess how to get the OS to comply hasn't been worth the reward. That being said, I've found the documentation for Swift to be extremely good, better than most other languages I work with.

Re: Microsoft suggests command line fiddling to get Windows 10 update installed

#209

Post by swells34 »

lupusreal wrote: post #179
maleldil wrote: post #95

Earlier quoted context omitted.

What hardware? Your experience with Linux can be very different depending on that. For example, my laptop's WiFi refuses to work properly on Linux, so I'm forced to use Windows. Edit: oh, and not to mention graphics (Nvidia), multiple monitors and HiDPI, especially with fractional scaling and different multiples per display. HiDPI on Linux doesn't work nearly as well as Windows and macOS.

If you know you're going to use Linux before buying a machine, you can generally save yourself all the trouble by simply avoiding broadcom and nvidia. Those two have been the source of the most trouble for years. Intel for graphics and wifi is generally the safest choice. I pick laptops with those and have great success.

A decade ago I stopped having problems with both BCM and NVidia devices in Linux. Last year I purchased a very recent wifi dongle and had to manually install broadcom drivers, but other than that it's been smooth sailing with no issues.

Re: Microsoft suggests command line fiddling to get Windows 10 update installed

#210

Post by ParetoOptimal »

trelane wrote: post #167
ParetoOptimal wrote: post #156

Earlier quoted context omitted.

Without Wayland I'm pretty sure this is still the case.

I use Wayland. Perhaps that's your issue?

I moved to Wayland mostly, but gaming performance is worse. Mostly latency and 1% fps lows.
Post reply on HN