Windows Subsystem for Linux

[Kazinsal]

I blanketly discard anything with the word "Microsoft" in it.

Alright then, what has Microsoft done that you don't consider evil?

Apparently nothing.

[FallenAvatar]

Honestly, at this point onlyonemac is nothing but a toxic member inciting BS arguments thru nothing more than his childish ignorance. Why hasn't he been banned?

- Monk

P.S. Or atleast why has this side convo not been moved to the auto-delete forum? (It has absolutely zero relevance in OSDev, or even general programming as its all pure BS rhetoric.)

[zdz]

Honestly, at this point onlyonemac is nothing but a toxic member inciting BS arguments thru nothing more than his childish ignorance. Why hasn't he been banned?

- Monk

P.S. Or atleast why has this side convo not been moved to the auto-delete forum? (It has absolutely zero relevance in OSDev, or even general programming as its all pure BS rhetoric.)

The topic about Ubuntu subsystem for Windows is relevant, but what we've done here (because I'm also guilty) is a total mess. I would really like to get back on the topic, but that seems impossible.

[iansjack]

Honestly, at this point onlyonemac is nothing but a toxic member inciting BS arguments thru nothing more than his childish ignorance. Why hasn't he been banned?

I am totally against anyone being banned just because they hold views that I think are silly. We're all a bit silly when we are young and I'd rather have the chance to refute untruths than just suppressing them.

[onlyonemac]

You have to understand that most BSODs are caused by faulty drivers, not by the OS. Have you had the curiosity to look at the memory dump to pinpoint the actual problem? I've never experienced BSODs caused directly by a Windows driver. Also, long boot times, weird behaviour, unexpected BSODs, there might have been some malware around. I had an old laptop (i5, 4G RAM, can't remember anything else) which booted Windows 8.1 in 15-20 seconds. This tells nothing.
We ran a lot of different tests on our drivers and there are crashes that reproduce in such weird condition that you can only ask yourself if we can catch them all. Sometimes it's not even the fault of one driver, but some weird interaction between two or more. You can't summarize it to "I had a friend who had a BSOD".

It still tells you quite a bit about the design of Windows for it to be capable of crashing so easily. I've only had two Linux kernel panics in about five years, and one was because my root filesystem was on an NFS server with a firewall blocking the connection from the client and the other was because I ran out of memory trying to transcode video on a machine with only 256 MB of RAM (with swapping disabled).

Also, I can guarantee that my friend's laptop had no malware on it. You don't know him, but I do and he is the most security-conscious user that I have ever encountered. The crash was clearly related to how long it had been since he had last booted the system.

[iansjack]

The crash was clearly related to how long it had been since he had last booted the system.

It only takes a little thought to realize that there is very little chance that merely not using a computer for a while can somehow cause an OS installation to be magically corrupted. (Although a hardware fault - say faulty RAM - could have this effect.)

If this had happened the other way round - a couple of months using Windows on a machine that also had a Ubuntu install had resulted in the Ubuntu installation being corrupted - I'm 100% sure that you would conclude that Windows had corrupted the Ubuntu partition. I'd certainly conclude that. And when it happens the other way round I'm pretty sure that one of the OSs caused the problem and it wasn't the one that hadn't been used.

Of course, what is obvious very much depends upon what you want to believe in the first place. It's always better to look at a problem dispassionately rather than with preconceived notions.

So a nice bit of anecdotal evidence, but it really tells us nothing (except, possible, that Ubuntu managed to corrupt a Windows installation on the same computer).

[Kazinsal]

(except, possible, that Ubuntu managed to corrupt a Windows installation on the same computer)

Which it will happily do because it will ignore an NTFS partition saying "please don't mount me unless you plan on restoring saved system state first" and then go ahead and start screwing around with the filesystem. The saved state gets corrupted, and the next time the Windows install tries to boot, the whole thing goes down because some system it didn't know about decided to ruin its known good system state.

[iansjack]

I've only had two Linux kernel panics in about five years

I haven't had a BSOD on any of my several Windows installations for well over 5 years.

But, again, anecdotal evidence is really pretty meaningless. I haven't had a kernel panic on any of my Linux or FreeBSD installations in that time either. I really can't remember the last time that any operating system crashed on me recently. (Applications are another thing - Firefox keeps crashing on my Gentoo installation.)

[zdz]

You have to understand that most BSODs are caused by faulty drivers, not by the OS. Have you had the curiosity to look at the memory dump to pinpoint the actual problem? I've never experienced BSODs caused directly by a Windows driver. Also, long boot times, weird behaviour, unexpected BSODs, there might have been some malware around. I had an old laptop (i5, 4G RAM, can't remember anything else) which booted Windows 8.1 in 15-20 seconds. This tells nothing.
We ran a lot of different tests on our drivers and there are crashes that reproduce in such weird condition that you can only ask yourself if we can catch them all. Sometimes it's not even the fault of one driver, but some weird interaction between two or more. You can't summarize it to "I had a friend who had a BSOD".

It still tells you quite a bit about the design of Windows for it to be capable of crashing so easily. I've only had two Linux kernel panics in about five years, and one was because my root filesystem was on an NFS server with a firewall blocking the connection from the client and the other was because I ran out of memory trying to transcode video on a machine with only 256 MB of RAM (with swapping disabled).

Also, I can guarantee that my friend's laptop had no malware on it. You don't know him, but I do and he is the most security-conscious user that I have ever encountered. The crash was clearly related to how long it had been since he had last booted the system.

I suppose you understand how the kernel works, right? There is this big chunk of resources shared by all the kernel mode code. And if one such kernel mode piece of code happens to be buggy and, let's count just a few things: access memory it had freed, or stay too long at a high IRQL level, or is corrupting other driver's code / data or... (I can go on and on) how is this the OS design fault? How is Linux protecting your system from a faulty driver / rootkit / hardware problem?
I only had 0 BSODs in five years on my three laptops. And bare in mind that I do a lot of stupid things with them.
If we count the virtual machines I had one BSOD due to a faulty BIOS (who reported that all memory was usable and then, when Windows discovered that the ACPI driver tried to map OS memory it put a stop to it - I presume this is also Windows' fault).
You write OS code, right? Explain to me how can a crash happen because you haven't booted your OS in a long time? I'm curious what's the logic behind this.
You can guarantee it had no malware? Take a look at this https://apt.securelist.com/#secondPage and tell me more about it, I'll wait. But hey, I guess you are now a security expert now.
I'll ask again, do you remember (did you even check) the BSOD message? A BSOD is a sign that something is wrong with your system (hardware related or software related) and it's Windows' method of protecting your data and itself. I'd rather have a BSOD if Patch Guard detects some kernel IAT hooks than let system run. But hey, I can actually read a memory dump and figure out what happen instead of blaming Microsoft for everything.
And those two panics you had with Linux, those would have never happened on Windows.

iansjack wrote:
(except, possible, that Ubuntu managed to corrupt a Windows installation on the same computer)

Which it will happily do because it will ignore an NTFS partition saying "please don't mount me unless you plan on restoring saved system state first" and then go ahead and start screwing around with the filesystem. The saved state gets corrupted, and the next time the Windows install tries to boot, the whole thing goes down because some system it didn't know about decided to ruin its known good system state.

Whoch is weird, because Lubuntu will refuse to mount a NTFS partition if the Windows crashed, or went through a forced shut down or is in hybernate. You have to go the extra mile and force it to do this.

[alexfru]

Explain to me how can a crash happen because you haven't booted your OS in a long time? I'm curious what's the logic behind this.

Well, people have been fncking up and with date/time and calendars for thousands of years now.
You can't rule out some stupid sh!t like this or this or this or this. Screwy updates (or update processes) are not unheard of. You may get "lucky".

[Brendan]

Hi,

You have to understand that most BSODs are caused by faulty drivers, not by the OS. Have you had the curiosity to look at the memory dump to pinpoint the actual problem? I've never experienced BSODs caused directly by a Windows driver. Also, long boot times, weird behaviour, unexpected BSODs, there might have been some malware around. I had an old laptop (i5, 4G RAM, can't remember anything else) which booted Windows 8.1 in 15-20 seconds. This tells nothing.
We ran a lot of different tests on our drivers and there are crashes that reproduce in such weird condition that you can only ask yourself if we can catch them all. Sometimes it's not even the fault of one driver, but some weird interaction between two or more. You can't summarize it to "I had a friend who had a BSOD".

It still tells you quite a bit about the design of Windows for it to be capable of crashing so easily. I've only had two Linux kernel panics in about five years, and one was because my root filesystem was on an NFS server with a firewall blocking the connection from the client and the other was because I ran out of memory trying to transcode video on a machine with only 256 MB of RAM (with swapping disabled).

Also, I can guarantee that my friend's laptop had no malware on it. You don't know him, but I do and he is the most security-conscious user that I have ever encountered. The crash was clearly related to how long it had been since he had last booted the system.

Does your guarantee that the friend's laptop had no malware on it include a guarantee that the laptop didn't have GRUB2?

The problem with GRUB is that they consistently fail to separate "boot manager" (choosing an OS to boot) and "boot loader" (booting the chosen OS); which results in a complicated config file that contains settings for all OSs. GRUB2 makes this worse by over-complicating that single config file that all OSs depend on so badly that Ubuntu (and most other Linux distros) have implemented "auto-magic scripting" to generate the GRUB2 config. This leads to "auto-magically generated over-complex config file that all OSs depend on" with a huge risk of Ubuntu screwing up other OSs when Ubuntu updates various things (e.g. when it updates GRUB2).

Worse, Windows uses various technologies to (try to) ensure that its boot code hasn't been tampered with (and hasn't been infected with a rootkit or something); so simply updating GRUB2 (in Ubuntu) can be enough to break Windows.

Note that it is not Microsofts fault that GRUB2 is a mis-designed pile of puss that screws up security. In a similar way it's not Ubuntu's fault that GRUB2 is so ugly that it needs "auto-magic scripting" just to shield end users from the sheer horror of GRUB's configuration mess.


Cheers,

Brendan

[iansjack]

Explain to me how can a crash happen because you haven't booted your OS in a long time? I'm curious what's the logic behind this.

Well, people have been fncking up and with date/time and calendars for thousands of years now.
You can't rule out some stupid sh!t like this or this or this or this. Screwy updates (or update processes) are not unheard of. You may get "lucky".

If there was such a fault in Windows we would have heard many more reports of similar happenings. It's a question of probabilities; without access to the computer and/or details of the BSOD it is impossible to say what happened. But it is an order of magnitude more likely that the cause was an external influence (Ubuntu screwing with the Linux partition, probably with help from the user) rather than some one-in-a-billion date-related bug in Windows.

[alexfru]

Well, people have been fncking up and with date/time and calendars for thousands of years now.
You can't rule out some stupid sh!t like this or this or this or this. Screwy updates (or update processes) are not unheard of. You may get "lucky".

If there was such a fault in Windows we would have heard many more reports of similar happenings. It's a question of probabilities; without access to the computer and/or details of the BSOD it is impossible to say what happened. But it is an order of magnitude more likely that the cause was an external influence (Ubuntu screwing with the Linux partition, probably with help from the user) rather than some one-in-a-billion date-related bug in Windows.

The thing is, we're in the business of genetically engineering bugs. The toughest survive and lurk around in the darkness. You can ship a product with race conditions that nobody ever sees in the field but they are there in the source code (I've witnessed this happen in different products at different companies) or you just see some rare and inexplicable crash dumps and have no idea how that could ever happen. I mean, the probability can be very low as well. A couple of years back I was installing updates on Windows 7 on my work PC. It was taking time and it was Friday night, so I just left it enjoying itself over the weekend. Monday morning it was still installing the updates. About to finish, showing 100% progress. I waited a little more (like an hour) and it still didn't finish. Since it must've clearly been broken by that time I figured it wouldn't make things any worse if I poked it a little. So I pressed ctrl+alt+del (as I'd do in the olden win9x days) and guess what? I got to log in. The thing had finished with the updates as normal, but due to a bug the screen was wrong. That's an interesting bug. I've been using Windows 7 (and Vista as well) since before RTM. And I never saw updates behave that way. Nor had I ever heard about it before. Updates hogging the CPU and stuff like that, broken updates, all that is pretty widespread. Hisenupdates like this are rare.

[zdz]

Explain to me how can a crash happen because you haven't booted your OS in a long time? I'm curious what's the logic behind this.

Well, people have been fncking up and with date/time and calendars for thousands of years now.
You can't rule out some stupid sh!t like this or this or this or this. Screwy updates (or update processes) are not unheard of. You may get "lucky".

A Windows Update Bugcheck will usually end up with an option to roll back that update at the next boot - starting with Windows 8(.1?) you won't even get to the loading screen and a message telling you about this will appear. I can't guarantee it will always do this, but I've never seen it not do this.

Note that it is not Microsofts fault that GRUB2 is a mis-designed pile of puss that screws up security. In a similar way it's not Ubuntu's fault that GRUB2 is so ugly that it needs "auto-magic scripting" just to shield end users from the sheer horror of GRUB's configuration mess.

So what you want to say is that an open source project isn't perfect? Oh no Windows should stop doing whatever it does in order to achieve boot time security (I can't say I'm 100% ok with some things it does - but that is another topic) and just use GRUB2 as a bootloader. I'm mostly joking, ignore this paragraph

I can say that I've seen Windows corrupting it's MBR. This was ~2 years ago. I had an older version installed after a newer one and it refused to boot anymore. I fixed this in 5 minutes by doing a MBR fix from a USB Stick with a random Windows .ISO on it. There were no other external forces involved in this. It never happened again and I simply ignored it so I can't tell you more about what exactly happened.

[iansjack]

The thing is, we're in the business of genetically engineering bugs. The toughest survive and lurk around in the darkness.

I'm not saying that rare bugs don't occur, just that - with no other information - this is not the most likely explanation. For every anecdote that you relate of rare Windows bugs I could give you links to a dozen posts on Tom's Hardware (for example) where people have screwed up a Windows installation by installing Linux alongside it.

So either is possible, but it is not reasonable to jump to the conclusion that the only explanation of this behaviour is a bug in Windows.

As another anecdote, my main computer has three Windows installations and two Linux ones. I've not - in living memory - had a BSOD or kernel panic in any of them. (But I have had to do a little manual work to put things right from time to time when I've managed to trash a partition table, boot order, or similar.)