If you skipped to the last paragraph, I'm an idiot and meant to direct you to the last two paragraphs. Hopefully that's not too much reading for you...
You can't hear below 20Hz, but you can feel it. The same applies over 20KHz. A Nyquist rate of 22.05KHz or 24KHz leaves out a lot of frequencies you can feel.
Of course, so do shitty speakers and literally every pair of headphones, so this may or may not matter to you.
If you don't buy into the actual science behind all of this, you can go ahead and skip right to the last paragraph of this post and read why none of it really matters.
For portable playback, you're right, there's no difference, as you don't feel below 20Hz or above 20KHz with your ears, where your headphones or earbuds are mounted. For playback through shitty speakers, you're also right; you might have a subwoofer that will give you everything down to about 8Hz or so, but you're probably missing everything over 16KHz, so it really doesn't matter if your storage format can preserve those frequencies.
If you happen to have speakers capable of outputting well above 20KHz, suddenly it begins to matter whether your source and equipment can preserve those frequencies. In fact, a lot of perceived sound isn't actually heard but, rather, is the result of a beat frequency generated when two or more higher-than-audible frequencies interact; if that beat frequency is within the audible range, the listener will "hear" it, but only if those inaudible frequencies are reproduced faithfully and not simply discarded by a too-low sample rate or a high-pass filter intended to remove harmonic distortion introduced by a shitty or overdriven DAC.
I'm may not be an audio engineer professionally, because I have other talents that pay better, but I am a very capable audio engineer. My office, a 10' by 11' space which doubles as a hobby recording studio, is acoustically dead in its default configuration, though I can move two strategically-placed items and liven it right up. It's great for conference calls, as well as any type of music you might be in to. My very inexpensive speaker setup has no trouble outputting frequencies as high as 47KHz (my cats hate when I test this) and yes, when I play 47KHz through one speaker and 46KHz through the other, you can swear there's a 1KHz tone playing in the room; step out of the room, or far enough out of the soundfield of either speaker, and the tone disappears.
If you've ever experienced an "audible" sound that you couldn't locate the source of, that's what it was: a beat pattern generated by two inaudible tones. Remember, those tones are lost with headphones and shitty speakers, so they're not really missed if that's what you're listening on.
Mind you, I'm no speaker snob; I listen primarily on headphones and earbuds, and enjoy my digital (even streamed) music just fine through them. But, I do happen to have equipment that can produce much better sound and, well, I enjoy that when I use it, as well, provided the source can support it.
Coincidentally, you can get pretty close to clean vinyl recording 24 bits per sample at 192KHz, until you try to compress it. That's a hair under 1.1MB/sec (that's megabytes), or 1.3GB per LP side (20 minutes); and those are binary megabytes and gigabytes. A 128GB (that's decimal GB because that's how we measure storage for some stupid reason) SD card is only 116GB before being formatted, roughly 114GB after, and will hold about 44 LP records. Mind you, it will do so in slightly lower quality than the original clean vinyl, but it's sure great to have as the records age.
Awesome for portability, right? Well, actually, no. Remember, you don't benefit from feeding inaudible frequencies to your ears, which is what you're doing if you happen to have such a source and a pair of headphones that can produce those frequencies. What it is good for is suddenly disappearing in a pile of junk mail on your desk, because it's just so much smaller than the LP records it's being compared to. For portability, you're better off putting 16-bit 48KHz audio on that SD card, allowin
You say it was reading and writing your hard drive, but you don't at any point mention network activity. Considering that the only way, out of the box, to monitor disk access is Task Manager and you specifically mention that this was the first 5 minutes, that's what you must have been using. Did you see network activity in that time? And which process are you referring to as "diagtrack"? If you have a process that's actually called "diagtrack", that's not Windows and you should contact your OEM about it.
Perhaps they should... What they're doing now seems to be working just as well, if not better, given that Win 10 is more stable than previous versions; but I imagine it would be even more so if they still had a QA team.
Someone has to physically retrieve it, which involves your active consent.
You gave that consent to Microsoft by installing a non-Enterprise version of Windows 10 and accepting the license agreement, or by installing an Enterprise version of Windows 10 and not disabling automatic error reporting.
If, in the event of a crash, Windows asked if it could send the crash report to Microsoft (like it used to!), there'd be no issue.
If Microsoft wasn't, then, forced to deal with idiots who insist they fix their crashing programs, yet refuse to provide crash reports when asked, there'd be no issue.
You're arguing in favor of telemetry, but I don't see anyone arguing against it. What people are arguing against is that it is mandatory.
Except that it's not. Either you work in an industry where Windows is mandatory, in which case you can afford the 5-license minimum for Enterprise and disable the telemetry, or you don't and you can use something else.
If we accept that Microsoft is being forthright and truthful about this, then the telemetry you can't turn off includes "basic device information, quality-related information, app compatibility, and Microsoft Store. When the level is set to Basic, it also includes the Security level information."
Basic device information (such as CPU type, RAM and storage sizes and utilization, and what hardware and drivers are installed) would seem to be somewhat required as part of a crash report. In fact, quality-related information would seem to be more user-friendly name for "why did it crash", coupled with "app compatibility" as a way of saying "what crashed". It really seems as though they've broken "crash report" into its component elements; likely in an attempt to be somewhat more transparent about what's in them. Looks like that backfired.
Also, I would certainly hope Microsoft, and not some other party, is getting information about how I use the Microsoft Store. How do you suspect the Windows Store works? Do you think every Windows install comes with a full copy of everything that has ever resided in, or will ever reside in, the Windows store (including the app I am currently writing), and just calculates the current state of the store based on the current date and time? Or do you think, more reasonably, that the current state of the store resides on Microsoft's servers and you have to send data back to those servers so they know what to serve you?
I don't think Microsoft has devised a way to see into the future and determine every single piece of software that will even be submitted to the Microsoft Store, nor have they invented a compression algorithm efficient enough to fit all of that onto a single DVD, so I'm leaning toward the server solution.
This is quite a bit more than only crash reports. Also, crash reports are not exactly innocuous. They can contain very sensitive information themselves.
There is actually a setting (set to disallow by default) to allow or disallow automatic sending of potentially sensitive contents (e.g. contents of RAM or files) along with crash reports. I don't recall where I saw it, but I do know it's there and defaults to asking the user prior to sending such data.
Update to what? Not from where. Windows XP? Because that's what I was talking about... you know, when I said "disabling automatic updates in earlier versions" and "before Microsoft played these games".
If people hadn't done that, then not held up their responsibility (to the rest of the users, not to themselves or Microsoft) to install security patches to ensure their machines didn't become shit-spewing bot nests, perhaps Microsoft wouldn't have taken away the ability.
You seem to only be able to mentally go as far back as the release of Windows 10, but we're discussing things that happened long, long before then, which lead to many of Microsoft's (admittedly ill-thought) decisions regarding Windows 10. Logical fallacy: attributing decisions made prior to an event to occurrences which followed. Correct that, then we'll talk.
WSUS servers weren't even available at one time.:)
When were WSUS servers not available for Windows 10? It's been a standard offering wince Win2k3.
As for 5 licenses, no, most won't have a running WSUS.
Won't and can't are two different things. I was stating the 5 license minimum for Enterprise versions of Windows.
Maybe at 50 or 100, when they decide they actually need an IT service or person and that person goes "Hey, for a mere 5K (license and hardware) you can manage all this with minimal fuss" and then still have issues as stuff gets pushed.
No competent IT person would quote $5k, as you only the Windows Server 2012 or newer system that is already running your domain controller and AD; tick the box to enable WSUS and add it to your policies; done. As an added bonus, a competent IT person would thoroughly test each update before adding it to WSUS and avoid the "issues" you allude to. It shouldn't take more than a day for a mid-level ($75-100/hr) tech to get working; in fact, it should only take an hour or two. That's $75-200 for a competent tech, up to $800 if he's really really slow.
As for the ongoing cost of having someone review updates, if your team is big enough that bad or poorly-timed updates are actually incurring a measurable cost in lost productivity, paying someone to properly test updates and only apply those which don't break anything will surely be cheaper than the lost productivity. The peace of mind that comes with knowing what's running on your systems, on the other hand, is priceless.
This isn't even what I do for a living, but I could set it up with one hand tied behind my back.
But, as for Win10, that doesn't make any difference either because everything will get pushed within 9 or 12 months, whatever the latest arbitrary deadline is.
Not so with WSUS. No version of Windows that is configured to use a WSUS server looks anywhere other than the configured WSUS server for updates.
Home users had just as many reasons to turn off WU as businesses did.
And anyone who turns it off also has a responsibility to periodically install that shit themselves. Guess who didn't live up to that responsibility and lost the ability to decide for themselves!
In fact, they more likely would be affected by an update screwing up their system and would be less likely to be able to fix it.
Then they should have learned what they were doing before they did it.
That they did not know enough to intelligently apply security fixes over time isn't really their fault.
Will you say the same of Mac users who disable automatic updates because Apple has released a few bad video drivers (more than just that, but it's what I recall off the top of my head) for older Macs? What of Linux and BSD users?
The fault still lies with MS for not breaking up "updates" into mandatory "security patches" and optional everything else and then not abusing that system with crap like the "Upgrade to Windows 10" program.
So it's Microsoft's fault people disabled Windows Update during the time before Windows 10, back when Microsoft did allow you to install updates by category? Going back at least as far as XP SP2, I know you could opt to have just "Critical Updates" installed, and those were just security patches.
It wasn't until Windows 10 that they began abusing that, so you can't really cite that abuse as the reason Windows XP users disabled Windows Update on day one and never installed a single update.
MS is still the root of the problem, and always will be.
Interesting opinion; I prefer to believe that ignorant users are the problem, as they are on any system.
If Microsoft wants telemetry data to resolve issues with system crashes, they can earn it. Start by actually reading through the forum posts with thousands of people reporting the same issue, and work to address that issue
How many of those thousands of people do you think can actually accurately describe the actual problem they're facing, let along provide the technical details that come from crash telemetry? It's honestly like Ford asking someone who was involved in a car accident due to a bug in their car's anti-lock braking system to help them fix it, rather than asking the car itself what went wrong; cars store post-crash and post-fault telemetry for a reason, and Windows does for the very same reason. Only the system knows why the system failed.
This is the norm in the Microsoft support forums. Microsoft cannot simultaneously argue that they need telemetry in order to address crashes, performance issues, and system instabilities, while also ignoring the green pastures of such information volunteered to them that goes unaddressed and unresolved unless another end user provides a workaround.
A feature or function not behaving as expected and a program crash are two different things. One (the program crash) will provide telemetry and the other will not. Microsoft does not need telemetry to learn that sometimes the Start menu does not open when you click it; and telemetry will not tell them that, either. Those types of issues do belong in forums, as they're not crashes but, rather, UI and UX bugs that telemetry can't possibly nail down; they're not failures of the system, they're failures of the design of the system.
A program crash, on the other hand, is much easier to track down and fix when you have the actual system that experienced it provide details about it that the end user who was sitting at that system can't possibly even be aware of. Sure, you can have a thousand people report the crash, each giving a slightly different account of the issue, and you can assume that all of those similar-sounding crashes follow the same root cause, spend countless hours attempting to reproduce an intermittent problem, finally get it to happen once so you can now confirm that a problem does exist, then spend countless more hours trying to reproduce it again and again with every proposed fix because, well, it's an intermittent problem, it doesn't happen every time you do the thing that triggers it...
Or, you can have the failing system tell you how and why it failed, immediately know what needs to be fixed and how to verify that it ha been fixed, and possibly learn that there are a handful of "whys" for a given "how". That's something a thousand forum posts can't give you.
Imagine a thousand people posting about Word crashing when they open files saved by a certain older version of Word. You read all thousand reports, they all say Word 2016 sometimes crashes when opening files saved by Word 2003. Do you know, from a thousand descriptions of the crash scenario, what caused those crashes? Do you know that there was just one cause? Might there be multiple causes? I mean, come on, we're talking about Microsoft, right? Even you should agree that a single issue in their software is likely to have multiple causes.
So, what, they see the forum posts, reproduce the issue on their end--they found a working test case, they're not gonna keep looking for more of them--and fix the issue they reproduced. Well... They fix one cause of that issue. Then they report back that it has been fixed.
And it has, for about 10% of the people who reported it.
Telemetry lets them see the actual problem, and not just the result of the problem, so they can fix it right the first time.
You can't honestly be sitting there with a straight face, comparing pre-XP Windows to post-XP Windows, and telling me it doesn't work. Every version of Windows released since XP has been more stab
Businesses with have IT staff who manage updates and ensure that security updates are installed regularly are not the problem. Though, with more than say 5 (the minimum enterprise license) workstations should have had their own WSUS servers in the first place.
Home user's were not doing all of that. They would just (stupidly) turn off WU and never install a single update. Those idiots are why we now cannot turn off automatic updates in Win 10.
End users have no basis for bitching about Microsoft removing the ability to turn off Windows Updates when they were the reason it was removed. Ever growing botnets, re-infection with virus after virus after the user just cleaned them up? Everything or nothing disabling of updates where your system remains vulnerable.
Script kiddies using exploits that got patched. You want to ban automatic updates for consumer computers because we're supposed to trust the user to install them to close the backdoor used by the bad guys?
No, end users made this mess and are hoping to blame Microsoft.
Because getting some kind of virus is a hypothetical
Until it happens to you or you see it happen to someone else. According to you, seeing it happen is enough; after all, you did say
seeing several people's presentations ruined [...] are tangible experiences.
All of the "ruined presentations" I've seen have been reported in the news media. The very same news media who reports on these viruses people are getting, mind you.
And yes, I've had updates interrupt my work before. Twice, on two different systems. I treat those incidents as bluescreens and, well, even with those, Windows 10 is still more stable and reliable than any previous version.
Because they do care about what crashes on your computer and why, so they can fix those issues. That's more to do with what other people (software developers) do on your computer than what you do on it.
By disabling automatic updates in earlier versions, before Microsoft played these games, the end users put themselves in a position of trust, in control of the security, stability, and performance of not only their own computers, but every computer connected to the internet, regardless of OS or version. Those same users also put themselves in a position of trust regarding the perception of the security of Microsoft's OS.
Those users failed miserably to live up to the position they chose for themselves.
As a result, Microsoft have, and perhaps rightly so, removed the option that previously allowed those users to put themselves in that position.
It does have the side effect of screwing those of us who both disabled automatic updates and manually installed updates within a reasonable timeframe (or took sufficient security measures to mitigate the risk of not having installed updates). Now, we no longer have that choice and yes, that does suck.
It sucks a bit less, though, when you take a moment to realize that, over time, the mess that is older versions of Windows with Windows Update disabled and manual updates literally never applied will clean itself up as those systems naturally remove themselves from the environment, either through obsolescence or hardware failure. They'll be replaced with new systems on which the user can't disable updates irresponsibly (that is, turn it off and "forget" to ever manually install updates that are actually important) and we'll all be better off for it.
Yes, even those of us who are suffering with forced updates now.
Indeed they did. That, coupled with the fact that they would then never go manually apply updates, and the ensuing malware shitstorm, is why we have forced updates in Win 10.
Therein lies the rub. John Deere can't stop a 3rd party from making an aftermarket ECU, control unit, or diagnostic unit, but copyright law dictates that anyone wishing to do so must not use John Deere's copyrighted software in order to do so unless they have John Deere's permission to use it. No law says they have to licence their copyrights, only that they have to not sue to prevent the sale and manufacture of parts. There is a healthy aftermarket for parts for John Deere tractors which can be replaced without touching the computer systems; that market would be a lot larger if we were allowed to reverse engineer those systems, but that is a separate issue.
And even then, existing laws only cover vehicle's intended to be licensed and driven on public roads, which does not cover most of John Deere's line, making them more like phones than cars. John Deere doesn't want to have to deal with those laws for all of their tractor business, so they allow what 3rd party parts are possible without the use or reverse engineering of their software (because they can still get away with preventing that) and, perhaps more importantly, they provide service well past the expected life of the product.
John Deere actually provides a good model for phone manufacturers to follow, at least in terms of ensuring security updates are available: Full warranty lasting at least as long as the financing period, software updates free of charge during the warranty period, software updates still available after the warranty period, and nothing technically stopping the end user from replacing the software with their own once the warranty ends.
I see where you're going re: Right to Repair and I agree, we'd be better off in that world. We would also be better off that our current situation if phone manufacturers follwed John Deere's model. In either case, I think you and I both realize it is going to take an act of law ir a very hardcore court ruling to make it happen.
A company can disclaim liability for the end user's actions or modifications, and they may limit how long they will warrant a product to be free of manufacturing defects; however, it is not unheard of and would not be unexpected for the court to expect maintenance services (such as software updates) to be made available for a duration exceeding the average usable life of a product, especially when lack of such maintenance is likely to result in financial or social harm to the user. Moreso when the device manufacturer is the only legitimate (and legal re: the DMCA) source of such maintenance. That is, a manufacturer and seller of a product cannot stipulate that their product should only be fit for the purpose for which it was sold for a certain period of time, save for reasons following spoilage, mechanical breakdown, electrical failure, or some other naturally occurring destructive force. And, indeed, even if that were not true, we have phones being sold as new, covered by a 1 year (sometimes longer) warranty, well after manufacturer updates have ceased; if your implication is that the manufacturer should only be liable for updates during the warranty period, well, there's your violation right there.
Put another way, no court would expect Ford to still provide parts and maintenance for a '69 Mustang (though they happily will) because it is trivial to find parts, service manuals, and indeed service stations, who will perform such maintenance. Even if it were not, so long as they provided said parts and maintenance during the typical usable life of the average car (this may seem a poor example as the industry accepted service term, during which the manufacturer must maintain parts inventory is around 10 years, while most cars last much longer -- more on this later, but it's actually quite a salient example indeed), which they did, they would be in the clear. The law, specifically around automobiles, stipulates that manufacturers must allow aftermarket parts to be manufactured and sold, so there are still parts available and plenty of places to have a vehicle serviced even after the manufacturer shuts down the lines.
Not so for phones.
As for why 10 years is acceptable for cars when cars typically last much longer, and why that makes cars a salient example of what the courts might expect in terms of manufacturer-provided support (especially absent a viable aftermarket for said support) if it can be shown that lack of availability of maintenance for a product might cause the end user financial or social harm, it's actually pretty simple. You see, leading up to the mid 1960's, auto makers got in the habit of stopping production of parts when they felt it was time to start selling more new cars, and they were aggressive about curtailing the manufacture of any 3rd party replacement parts. In the 1960's, a law was enacted requiring them to make parts available for a minimum of 20 years; this was repealed in the 1980's as the law was effectively forcing manufacturers to maintain inventory of parts which would literally never sell; but, before then, another law was introduced requiring them to allow 3rd party replacement parts to be manufactured and sold.
Despite no longer being forced by law to maintain parts inventories for 20 years, the US auto industry has adopted a minimum standard of 10 years of factory support, so as to prevent such a law from ever being deemed necessary again. Why 10 years? Well, back in the 1980's when this policy was adopted, 5 years was a common car loan term; roughly half the new-car-buying population would wither trade in their car before it was fully paid for, or do so shortly thereafter, which meant roughly half the new-car-buying population was maintaining a roughly 5-year cycle of vehicle ownership. Of the remaining half, roughly half (so 25% of the total new-car-buying population) would keep their car for 5 years after paying it off, while the remainder would keep it for longer. The math worked out that the averag
Part of me wants to suggest we say "fuck it" and factor historical lack of updates into our purchase decisions. That is, purposely buy phones based on lack of updates, given that we know they will eventually get hacked.
And when they do, the entire lot of us stick it to the manufacturers, who refused to provide updates, in the form of a series of class action suits, one for each model sold in each market, layered atop one for each manufacturer in each market.
Vendors don't provide updates right now because they believe it is cheaper not to do so. We can't change the equation, but we sure can change the variables and, by effect, the result.
There's a disconnect somewhere in the editing (go figure). In reality, the silicon and device vendors need only be involved in the initial development phase of a given device's ROM. After that, provided Google doesn't change any APIs (and they generally don't as doing so would break apps as well), the underlying Android core can be updater, even wholesale replaced, by Google.
New APIs and features can be added by Google as they are developed, and the silicon and device vendors can update their apps and drivers to take advantage of those new APIs and features, or not; the user still gets to benefit from them in the form of being able to run the newest apps on the newest version of Android. Also, security patches for the Android core can come direct from Google. Security patches for vendor-supplied apps and drivers still need to come from the vendor, but there's really nothing Google can do to change that.
In this case, Microsoft built the bomb. You can't blame the guy who saw Microsoft's bomb and reported its existence when someone else takes Microsoft's bomb and kills your wife with it. You can blame Microsoft and whoever took and used the bomb, but not the guy who reported on the bomb's existence in the first place; the bomb still existed before he reported it and would still have been used had he not reported it.
Online, in this context, means a connected disk. An offline disk, one which is not connected to the system, cannot be affected in any way. That's why any sane backup system involves multiple volumes which are only connected to the system (e.g. brought online) as needed.
Slashdot Mirror
If you skipped to the last paragraph, I'm an idiot and meant to direct you to the last two paragraphs. Hopefully that's not too much reading for you...
You can't hear below 20Hz, but you can feel it. The same applies over 20KHz. A Nyquist rate of 22.05KHz or 24KHz leaves out a lot of frequencies you can feel.
Of course, so do shitty speakers and literally every pair of headphones, so this may or may not matter to you.
If you don't buy into the actual science behind all of this, you can go ahead and skip right to the last paragraph of this post and read why none of it really matters.
For portable playback, you're right, there's no difference, as you don't feel below 20Hz or above 20KHz with your ears, where your headphones or earbuds are mounted. For playback through shitty speakers, you're also right; you might have a subwoofer that will give you everything down to about 8Hz or so, but you're probably missing everything over 16KHz, so it really doesn't matter if your storage format can preserve those frequencies.
If you happen to have speakers capable of outputting well above 20KHz, suddenly it begins to matter whether your source and equipment can preserve those frequencies. In fact, a lot of perceived sound isn't actually heard but, rather, is the result of a beat frequency generated when two or more higher-than-audible frequencies interact; if that beat frequency is within the audible range, the listener will "hear" it, but only if those inaudible frequencies are reproduced faithfully and not simply discarded by a too-low sample rate or a high-pass filter intended to remove harmonic distortion introduced by a shitty or overdriven DAC.
I'm may not be an audio engineer professionally, because I have other talents that pay better, but I am a very capable audio engineer. My office, a 10' by 11' space which doubles as a hobby recording studio, is acoustically dead in its default configuration, though I can move two strategically-placed items and liven it right up. It's great for conference calls, as well as any type of music you might be in to. My very inexpensive speaker setup has no trouble outputting frequencies as high as 47KHz (my cats hate when I test this) and yes, when I play 47KHz through one speaker and 46KHz through the other, you can swear there's a 1KHz tone playing in the room; step out of the room, or far enough out of the soundfield of either speaker, and the tone disappears.
If you've ever experienced an "audible" sound that you couldn't locate the source of, that's what it was: a beat pattern generated by two inaudible tones. Remember, those tones are lost with headphones and shitty speakers, so they're not really missed if that's what you're listening on.
Mind you, I'm no speaker snob; I listen primarily on headphones and earbuds, and enjoy my digital (even streamed) music just fine through them. But, I do happen to have equipment that can produce much better sound and, well, I enjoy that when I use it, as well, provided the source can support it.
Coincidentally, you can get pretty close to clean vinyl recording 24 bits per sample at 192KHz, until you try to compress it. That's a hair under 1.1MB/sec (that's megabytes), or 1.3GB per LP side (20 minutes); and those are binary megabytes and gigabytes. A 128GB (that's decimal GB because that's how we measure storage for some stupid reason) SD card is only 116GB before being formatted, roughly 114GB after, and will hold about 44 LP records. Mind you, it will do so in slightly lower quality than the original clean vinyl, but it's sure great to have as the records age.
Awesome for portability, right? Well, actually, no. Remember, you don't benefit from feeding inaudible frequencies to your ears, which is what you're doing if you happen to have such a source and a pair of headphones that can produce those frequencies. What it is good for is suddenly disappearing in a pile of junk mail on your desk, because it's just so much smaller than the LP records it's being compared to. For portability, you're better off putting 16-bit 48KHz audio on that SD card, allowin
You say it was reading and writing your hard drive, but you don't at any point mention network activity. Considering that the only way, out of the box, to monitor disk access is Task Manager and you specifically mention that this was the first 5 minutes, that's what you must have been using. Did you see network activity in that time? And which process are you referring to as "diagtrack"? If you have a process that's actually called "diagtrack", that's not Windows and you should contact your OEM about it.
Perhaps they should... What they're doing now seems to be working just as well, if not better, given that Win 10 is more stable than previous versions; but I imagine it would be even more so if they still had a QA team.
Someone has to physically retrieve it, which involves your active consent.
You gave that consent to Microsoft by installing a non-Enterprise version of Windows 10 and accepting the license agreement, or by installing an Enterprise version of Windows 10 and not disabling automatic error reporting.
If, in the event of a crash, Windows asked if it could send the crash report to Microsoft (like it used to!), there'd be no issue.
If Microsoft wasn't, then, forced to deal with idiots who insist they fix their crashing programs, yet refuse to provide crash reports when asked, there'd be no issue.
You're arguing in favor of telemetry, but I don't see anyone arguing against it. What people are arguing against is that it is mandatory.
Except that it's not. Either you work in an industry where Windows is mandatory, in which case you can afford the 5-license minimum for Enterprise and disable the telemetry, or you don't and you can use something else.
If we accept that Microsoft is being forthright and truthful about this, then the telemetry you can't turn off includes "basic device information, quality-related information, app compatibility, and Microsoft Store. When the level is set to Basic, it also includes the Security level information."
Basic device information (such as CPU type, RAM and storage sizes and utilization, and what hardware and drivers are installed) would seem to be somewhat required as part of a crash report. In fact, quality-related information would seem to be more user-friendly name for "why did it crash", coupled with "app compatibility" as a way of saying "what crashed". It really seems as though they've broken "crash report" into its component elements; likely in an attempt to be somewhat more transparent about what's in them. Looks like that backfired.
Also, I would certainly hope Microsoft, and not some other party, is getting information about how I use the Microsoft Store. How do you suspect the Windows Store works? Do you think every Windows install comes with a full copy of everything that has ever resided in, or will ever reside in, the Windows store (including the app I am currently writing), and just calculates the current state of the store based on the current date and time? Or do you think, more reasonably, that the current state of the store resides on Microsoft's servers and you have to send data back to those servers so they know what to serve you?
I don't think Microsoft has devised a way to see into the future and determine every single piece of software that will even be submitted to the Microsoft Store, nor have they invented a compression algorithm efficient enough to fit all of that onto a single DVD, so I'm leaning toward the server solution.
This is quite a bit more than only crash reports. Also, crash reports are not exactly innocuous. They can contain very sensitive information themselves.
There is actually a setting (set to disallow by default) to allow or disallow automatic sending of potentially sensitive contents (e.g. contents of RAM or files) along with crash reports. I don't recall where I saw it, but I do know it's there and defaults to asking the user prior to sending such data.
Update to what? Not from where. Windows XP? Because that's what I was talking about... you know, when I said "disabling automatic updates in earlier versions" and "before Microsoft played these games".
If people hadn't done that, then not held up their responsibility (to the rest of the users, not to themselves or Microsoft) to install security patches to ensure their machines didn't become shit-spewing bot nests, perhaps Microsoft wouldn't have taken away the ability.
You seem to only be able to mentally go as far back as the release of Windows 10, but we're discussing things that happened long, long before then, which lead to many of Microsoft's (admittedly ill-thought) decisions regarding Windows 10. Logical fallacy: attributing decisions made prior to an event to occurrences which followed. Correct that, then we'll talk.
WSUS servers weren't even available at one time. :)
When were WSUS servers not available for Windows 10? It's been a standard offering wince Win2k3.
As for 5 licenses, no, most won't have a running WSUS.
Won't and can't are two different things. I was stating the 5 license minimum for Enterprise versions of Windows.
Maybe at 50 or 100, when they decide they actually need an IT service or person and that person goes "Hey, for a mere 5K (license and hardware) you can manage all this with minimal fuss" and then still have issues as stuff gets pushed.
No competent IT person would quote $5k, as you only the Windows Server 2012 or newer system that is already running your domain controller and AD; tick the box to enable WSUS and add it to your policies; done. As an added bonus, a competent IT person would thoroughly test each update before adding it to WSUS and avoid the "issues" you allude to. It shouldn't take more than a day for a mid-level ($75-100/hr) tech to get working; in fact, it should only take an hour or two. That's $75-200 for a competent tech, up to $800 if he's really really slow.
As for the ongoing cost of having someone review updates, if your team is big enough that bad or poorly-timed updates are actually incurring a measurable cost in lost productivity, paying someone to properly test updates and only apply those which don't break anything will surely be cheaper than the lost productivity. The peace of mind that comes with knowing what's running on your systems, on the other hand, is priceless.
This isn't even what I do for a living, but I could set it up with one hand tied behind my back.
But, as for Win10, that doesn't make any difference either because everything will get pushed within 9 or 12 months, whatever the latest arbitrary deadline is.
Not so with WSUS. No version of Windows that is configured to use a WSUS server looks anywhere other than the configured WSUS server for updates.
Home users had just as many reasons to turn off WU as businesses did.
And anyone who turns it off also has a responsibility to periodically install that shit themselves. Guess who didn't live up to that responsibility and lost the ability to decide for themselves!
In fact, they more likely would be affected by an update screwing up their system and would be less likely to be able to fix it.
Then they should have learned what they were doing before they did it.
That they did not know enough to intelligently apply security fixes over time isn't really their fault.
Will you say the same of Mac users who disable automatic updates because Apple has released a few bad video drivers (more than just that, but it's what I recall off the top of my head) for older Macs? What of Linux and BSD users?
The fault still lies with MS for not breaking up "updates" into mandatory "security patches" and optional everything else and then not abusing that system with crap like the "Upgrade to Windows 10" program.
So it's Microsoft's fault people disabled Windows Update during the time before Windows 10, back when Microsoft did allow you to install updates by category? Going back at least as far as XP SP2, I know you could opt to have just "Critical Updates" installed, and those were just security patches.
It wasn't until Windows 10 that they began abusing that, so you can't really cite that abuse as the reason Windows XP users disabled Windows Update on day one and never installed a single update.
MS is still the root of the problem, and always will be.
Interesting opinion; I prefer to believe that ignorant users are the problem, as they are on any system.
Facts are fac
If Microsoft wants telemetry data to resolve issues with system crashes, they can earn it. Start by actually reading through the forum posts with thousands of people reporting the same issue, and work to address that issue
How many of those thousands of people do you think can actually accurately describe the actual problem they're facing, let along provide the technical details that come from crash telemetry? It's honestly like Ford asking someone who was involved in a car accident due to a bug in their car's anti-lock braking system to help them fix it, rather than asking the car itself what went wrong; cars store post-crash and post-fault telemetry for a reason, and Windows does for the very same reason. Only the system knows why the system failed.
This is the norm in the Microsoft support forums. Microsoft cannot simultaneously argue that they need telemetry in order to address crashes, performance issues, and system instabilities, while also ignoring the green pastures of such information volunteered to them that goes unaddressed and unresolved unless another end user provides a workaround.
A feature or function not behaving as expected and a program crash are two different things. One (the program crash) will provide telemetry and the other will not. Microsoft does not need telemetry to learn that sometimes the Start menu does not open when you click it; and telemetry will not tell them that, either. Those types of issues do belong in forums, as they're not crashes but, rather, UI and UX bugs that telemetry can't possibly nail down; they're not failures of the system, they're failures of the design of the system.
A program crash, on the other hand, is much easier to track down and fix when you have the actual system that experienced it provide details about it that the end user who was sitting at that system can't possibly even be aware of. Sure, you can have a thousand people report the crash, each giving a slightly different account of the issue, and you can assume that all of those similar-sounding crashes follow the same root cause, spend countless hours attempting to reproduce an intermittent problem, finally get it to happen once so you can now confirm that a problem does exist, then spend countless more hours trying to reproduce it again and again with every proposed fix because, well, it's an intermittent problem, it doesn't happen every time you do the thing that triggers it...
Or, you can have the failing system tell you how and why it failed, immediately know what needs to be fixed and how to verify that it ha been fixed, and possibly learn that there are a handful of "whys" for a given "how". That's something a thousand forum posts can't give you.
Imagine a thousand people posting about Word crashing when they open files saved by a certain older version of Word. You read all thousand reports, they all say Word 2016 sometimes crashes when opening files saved by Word 2003. Do you know, from a thousand descriptions of the crash scenario, what caused those crashes? Do you know that there was just one cause? Might there be multiple causes? I mean, come on, we're talking about Microsoft, right? Even you should agree that a single issue in their software is likely to have multiple causes.
So, what, they see the forum posts, reproduce the issue on their end--they found a working test case, they're not gonna keep looking for more of them--and fix the issue they reproduced. Well... They fix one cause of that issue. Then they report back that it has been fixed.
And it has, for about 10% of the people who reported it.
Telemetry lets them see the actual problem, and not just the result of the problem, so they can fix it right the first time.
You can't honestly be sitting there with a straight face, comparing pre-XP Windows to post-XP Windows, and telling me it doesn't work. Every version of Windows released since XP has been more stab
Crash reporting is literally the only telemetry you can't turn off...
Second or third update to...???
Businesses with have IT staff who manage updates and ensure that security updates are installed regularly are not the problem. Though, with more than say 5 (the minimum enterprise license) workstations should have had their own WSUS servers in the first place.
Home user's were not doing all of that. They would just (stupidly) turn off WU and never install a single update. Those idiots are why we now cannot turn off automatic updates in Win 10.
End users have no basis for bitching about Microsoft removing the ability to turn off Windows Updates when they were the reason it was removed. Ever growing botnets, re-infection with virus after virus after the user just cleaned them up? Everything or nothing disabling of updates where your system remains vulnerable.
Script kiddies using exploits that got patched. You want to ban automatic updates for consumer computers because we're supposed to trust the user to install them to close the backdoor used by the bad guys?
No, end users made this mess and are hoping to blame Microsoft.
I'm not buying what you're trying to sell.
Because getting some kind of virus is a hypothetical
Until it happens to you or you see it happen to someone else. According to you, seeing it happen is enough; after all, you did say
seeing several people's presentations ruined [...] are tangible experiences.
All of the "ruined presentations" I've seen have been reported in the news media. The very same news media who reports on these viruses people are getting, mind you.
And yes, I've had updates interrupt my work before. Twice, on two different systems. I treat those incidents as bluescreens and, well, even with those, Windows 10 is still more stable and reliable than any previous version.
Because they do care about what crashes on your computer and why, so they can fix those issues. That's more to do with what other people (software developers) do on your computer than what you do on it.
Playing devil's advocate, here...
By disabling automatic updates in earlier versions, before Microsoft played these games, the end users put themselves in a position of trust, in control of the security, stability, and performance of not only their own computers, but every computer connected to the internet, regardless of OS or version. Those same users also put themselves in a position of trust regarding the perception of the security of Microsoft's OS.
Those users failed miserably to live up to the position they chose for themselves.
As a result, Microsoft have, and perhaps rightly so, removed the option that previously allowed those users to put themselves in that position.
It does have the side effect of screwing those of us who both disabled automatic updates and manually installed updates within a reasonable timeframe (or took sufficient security measures to mitigate the risk of not having installed updates). Now, we no longer have that choice and yes, that does suck.
It sucks a bit less, though, when you take a moment to realize that, over time, the mess that is older versions of Windows with Windows Update disabled and manual updates literally never applied will clean itself up as those systems naturally remove themselves from the environment, either through obsolescence or hardware failure. They'll be replaced with new systems on which the user can't disable updates irresponsibly (that is, turn it off and "forget" to ever manually install updates that are actually important) and we'll all be better off for it.
Yes, even those of us who are suffering with forced updates now.
Indeed they did. That, coupled with the fact that they would then never go manually apply updates, and the ensuing malware shitstorm, is why we have forced updates in Win 10.
Thanks, assholes. And I don't mean Microsoft.
Ugh... vehicles, not vehicle's... this is why I don't post from my phone...
Therein lies the rub. John Deere can't stop a 3rd party from making an aftermarket ECU, control unit, or diagnostic unit, but copyright law dictates that anyone wishing to do so must not use John Deere's copyrighted software in order to do so unless they have John Deere's permission to use it. No law says they have to licence their copyrights, only that they have to not sue to prevent the sale and manufacture of parts. There is a healthy aftermarket for parts for John Deere tractors which can be replaced without touching the computer systems; that market would be a lot larger if we were allowed to reverse engineer those systems, but that is a separate issue.
And even then, existing laws only cover vehicle's intended to be licensed and driven on public roads, which does not cover most of John Deere's line, making them more like phones than cars. John Deere doesn't want to have to deal with those laws for all of their tractor business, so they allow what 3rd party parts are possible without the use or reverse engineering of their software (because they can still get away with preventing that) and, perhaps more importantly, they provide service well past the expected life of the product.
John Deere actually provides a good model for phone manufacturers to follow, at least in terms of ensuring security updates are available: Full warranty lasting at least as long as the financing period, software updates free of charge during the warranty period, software updates still available after the warranty period, and nothing technically stopping the end user from replacing the software with their own once the warranty ends.
I see where you're going re: Right to Repair and I agree, we'd be better off in that world. We would also be better off that our current situation if phone manufacturers follwed John Deere's model. In either case, I think you and I both realize it is going to take an act of law ir a very hardcore court ruling to make it happen.
Eh? Which Android devices exist that cannot be updated by the OEM? We're talking about devices that aren't updated. There's a pretty big difference.
Also my home country, thanks.
A company can disclaim liability for the end user's actions or modifications, and they may limit how long they will warrant a product to be free of manufacturing defects; however, it is not unheard of and would not be unexpected for the court to expect maintenance services (such as software updates) to be made available for a duration exceeding the average usable life of a product, especially when lack of such maintenance is likely to result in financial or social harm to the user. Moreso when the device manufacturer is the only legitimate (and legal re: the DMCA) source of such maintenance. That is, a manufacturer and seller of a product cannot stipulate that their product should only be fit for the purpose for which it was sold for a certain period of time, save for reasons following spoilage, mechanical breakdown, electrical failure, or some other naturally occurring destructive force. And, indeed, even if that were not true, we have phones being sold as new, covered by a 1 year (sometimes longer) warranty, well after manufacturer updates have ceased; if your implication is that the manufacturer should only be liable for updates during the warranty period, well, there's your violation right there.
Put another way, no court would expect Ford to still provide parts and maintenance for a '69 Mustang (though they happily will) because it is trivial to find parts, service manuals, and indeed service stations, who will perform such maintenance. Even if it were not, so long as they provided said parts and maintenance during the typical usable life of the average car (this may seem a poor example as the industry accepted service term, during which the manufacturer must maintain parts inventory is around 10 years, while most cars last much longer -- more on this later, but it's actually quite a salient example indeed), which they did, they would be in the clear. The law, specifically around automobiles, stipulates that manufacturers must allow aftermarket parts to be manufactured and sold, so there are still parts available and plenty of places to have a vehicle serviced even after the manufacturer shuts down the lines.
Not so for phones.
As for why 10 years is acceptable for cars when cars typically last much longer, and why that makes cars a salient example of what the courts might expect in terms of manufacturer-provided support (especially absent a viable aftermarket for said support) if it can be shown that lack of availability of maintenance for a product might cause the end user financial or social harm, it's actually pretty simple. You see, leading up to the mid 1960's, auto makers got in the habit of stopping production of parts when they felt it was time to start selling more new cars, and they were aggressive about curtailing the manufacture of any 3rd party replacement parts. In the 1960's, a law was enacted requiring them to make parts available for a minimum of 20 years; this was repealed in the 1980's as the law was effectively forcing manufacturers to maintain inventory of parts which would literally never sell; but, before then, another law was introduced requiring them to allow 3rd party replacement parts to be manufactured and sold.
Despite no longer being forced by law to maintain parts inventories for 20 years, the US auto industry has adopted a minimum standard of 10 years of factory support, so as to prevent such a law from ever being deemed necessary again. Why 10 years? Well, back in the 1980's when this policy was adopted, 5 years was a common car loan term; roughly half the new-car-buying population would wither trade in their car before it was fully paid for, or do so shortly thereafter, which meant roughly half the new-car-buying population was maintaining a roughly 5-year cycle of vehicle ownership. Of the remaining half, roughly half (so 25% of the total new-car-buying population) would keep their car for 5 years after paying it off, while the remainder would keep it for longer. The math worked out that the averag
Part of me wants to suggest we say "fuck it" and factor historical lack of updates into our purchase decisions. That is, purposely buy phones based on lack of updates, given that we know they will eventually get hacked.
And when they do, the entire lot of us stick it to the manufacturers, who refused to provide updates, in the form of a series of class action suits, one for each model sold in each market, layered atop one for each manufacturer in each market.
Vendors don't provide updates right now because they believe it is cheaper not to do so. We can't change the equation, but we sure can change the variables and, by effect, the result.
There's a disconnect somewhere in the editing (go figure). In reality, the silicon and device vendors need only be involved in the initial development phase of a given device's ROM. After that, provided Google doesn't change any APIs (and they generally don't as doing so would break apps as well), the underlying Android core can be updater, even wholesale replaced, by Google.
New APIs and features can be added by Google as they are developed, and the silicon and device vendors can update their apps and drivers to take advantage of those new APIs and features, or not; the user still gets to benefit from them in the form of being able to run the newest apps on the newest version of Android. Also, security patches for the Android core can come direct from Google. Security patches for vendor-supplied apps and drivers still need to come from the vendor, but there's really nothing Google can do to change that.
In this case, Microsoft built the bomb. You can't blame the guy who saw Microsoft's bomb and reported its existence when someone else takes Microsoft's bomb and kills your wife with it. You can blame Microsoft and whoever took and used the bomb, but not the guy who reported on the bomb's existence in the first place; the bomb still existed before he reported it and would still have been used had he not reported it.
Online, in this context, means a connected disk. An offline disk, one which is not connected to the system, cannot be affected in any way. That's why any sane backup system involves multiple volumes which are only connected to the system (e.g. brought online) as needed.