Quick! Rush out a new fix without testing it!
...and repeat, ad infinitum...
Microsoft on Tuesday posted KB4464330 (Windows 10 1809 Build 17763.55) in an effort to halt the damage done by last week's Windows 10 version 1809 update, but it hasn't quite worked. Reports of update inflicted file deletion and loss of internet connectivity for Microsoft Edge and Microsoft Store apps have died down. But users …
They've been doing that pretty much since Worries for Windows. Even though it became very much obvious during Windows Vista, they achieved enough lock in to survive that.
I'm very unimpressed with their software, but you must admit that their machinations to keep their claws into a market the quality of their software doesn't deserve for some 3 decades is impressive. If they spent 20% of that investment and effort on the actual products they would not have to troll forums..
This week’s Win10 update ‘only’ deleted my Audio drivers on a modern HP lappy, matter of minutes to d/l 300mb & install successfully from hp.com, as all the built in ‘repair’ tools in the “Service called Windows Ten”, formerly an OS, refused to acknowledge there was any problem with audio.
Likewise....everyone loves waking up to their laptop, starting a teams meeting and then realising there are no audio drivers because Microsoft have decided it should be deleted.
Yet literally everyone accepts this level of shoddiness "well it is kind of free", so is MacOS and not once have i updated my 8 year old Macbook and had it forget how to play sound!
"so is MacOS and not once have i updated my 8 year old Macbook and had it forget how to play sound!"
Funny you should mention that but certain models of iMac had exactly that problem after updating to a particular MacOS, I forget whether it was Sierra or the one before that but essentially all the iMacs we had of a particular model stopped being able to output sound via the internal speakers following the update.
To be fair it was a trivial fix, plugging something into the headphone jack and unplugging it fixed it, but the principal is the same, it should never have happened.
In some ways you could blame Microsoft less for this, they can't realistically test Windows updates on every conceivable bit of hardware that it may be running on. Apple on the other hand have a much smaller set of hardware to test, and they make all of it!
Did the same thing to my ASUS. Exact same scenario. Wife was looking for a new laptop so I scanned some recent reviews and found a log of Lenovo users complaining of what looked like the same thing.
Why is windows trashing, apparently across multiple brands, sound drivers? Have they no QA? (Yes, I know, rhetorical question.)
From the article: "The update debate has divided security experts about whether automatic updates are more trouble than they're worth"
Obviously more 'cons' than 'pros'. See icon.
But, will Micro-shaft do what the CUSTOMERS want? (let alone security experts)
(snarky, hysterical, uncontrollable laughter follows)
You do realise that's like buying a car and the brakes fail then blaming the owner because another model had issues with the same software. I'm not a fan of either to be honest from personal experience (Windows 10 that is and HP to some extent) but you can't generalise like that.
Every OS and hardware manufacturer has had problems over the years.
I'm not sure it was ever best of breed (really ancient Laserjet printers were good I suppose) but if it was, it was a very, very long time ago indeed.
Every manufacturer has their problems but very few have been so consistently poor in my experience than HP, particularly with laptops. The servers are OK I suppose, if you forget about that time when they said you'd need a support contract in order to get continued firmware updates...
It's outrageous IMNHO that Microsoft can cause untold damage to millions of people and businesses worldwide and escape all accountability for their negligence in not properly validating OS and software updates. These preventable losses in the many millions are a result of pure negligence by Microsoft. Why is it that consumer protection agencies have allowed Microsoft and other companies to wreck havoc on PCs via blatant neglect when these updates can be validated prior to distribution? If there is no accountability for damages then there is little incentive to deliver proper updates.
Businesses would be stupid to allow automated patching like this.
at a minimum TEST -> PILOT -> PROD
While MS are cuplable for a bad patch, the business is responsible for ensuring it does not cause problems for their staff. Very easy to do if you are using WSUS, BigFix, LANDesk
If you follow the minmum of a 3 step cycle and expose more machines each time, your TEST and PILOT phases - if correct - should cover most of your client and server configurations (of course, how you get these phases populated is up for discussion)
The above is all well and good if you're running a business
But if you are Mr Average HP owner running win10 home edition and suddenly your laptop is hosed.... what are you going to do about it?
Apart from have to pay someone to fix it.
Maybe thats why m$ dont do as much QC work to their software.. because they get a kickback from every mom and pop computer store for fixing 1000s of pcs when auto update hoses them...
"Not all businesses have on staff IT people"
And that is the problem, if you are going to operate a complex machine you need to hire experts who know how to properly maintain it.
Despite what microsoft claim, windows is simply not suitable for non technical users - keeping it running reliably and securely is extremely difficult and requires highly skilled (ie expensive) people to do so.
hogwash. Windows 7 ran happily with updates set to automatic. I cannot think of an updatr that caused as much damage as 1809 w10 update.
A small garage running an MOT PC, an accounts PC, a diagnostic PC and a general office PC would not have extwnsive IT staff. They would buy preset up dells, use a business internet line with managed firewall and away you go.
w10 is a nightmare. it auto updates major releases and has little rollback capability. The inbuilt w10 backup is shite, you need to use the legacy backup (or veeam). provisioned apps are amongst the most stupid things available, especially baked in shite in lower than enterprise versions.
W10 is a train wreck and hopefully a few large businesses will take on a linux rollout to wake MS up.
I’m not kicking back squat to micro$oft! But I sure catch a lot of grief from customers when I have to charge them to fix their computer after a Windoze 10 update borks their computer. Having to wipe and reload is NOT an acceptable answer to a bad update.
> They've got a bunch of keys to launch the browser, printer, calculator, etc...
The last W10 update hosed the key I used, I found it very useful having a key for controlling the screen brightness rather than having to going through all the settings menus. But since it was useful MS decided it would no longer be allowed and stopping it working.
The last W10 update hosed the key I used,
As per that link my setup was in 'C:\SWSETUP\APP\Applications\HP\HPHotkeyS_SS9NB2\220.127.116.11\src'. Ran that, rebooted and keys were working again. If can't change brightness even from display settings, manually update the graphics driver.
These keys (sometimes called media keys) have keycodes attached to them and the OS itself is supposed to recognize these (such does Android). It's pretty much Windows's fault.
But writing a whole DEVICE DRIVER for that is, as you said, truly nonsense.
recording keystrokes? If that's the case, what's different about Microshaft slurping your activity and sending the data back to Redmond? You know, so "the store" can offer you "smart suggestions"...
From Joe Belfiore's infamous keynote speech at a dev conference shortly before win-10-nic released...
"As I'm using my PC, on the client we know which apps you're launching, and which apps you're installing, and so we're able to communicate with the store and bring down suggestions that are personalized for ~you~, to help users learn about great new apps that are available for them to try out on their PC"
In any case, I am starting to think that maybe HP has had WAY too much of Micro-shaft's coolaid...
[they would do well to ship with Linux as their primary OS, instead]
"This is why auto-installed updates are so dangerous," observed Steve Bellovin, a professor in the computer science department at Columbia University, via Twitter.
But Washington Post reporter Barton Gellman defended auto-updates, arguing they do more good than harm.
So lessee here...on the one hand, you have a CompSci professor (who fully understands the Second Law of Computer Science: "You can't change anything without fucking something else up") who thinks cramming updates down one's throat is a bad idea. On the other hand, a newspaper hack (who likely can't tell the difference between Windows Explorer and Internet Explorer) saying its great.
Hmmm...I wonder who I should believe...
Barton and Alex miss two major points:
1) Earlier Windows versions received mainly small security updates and the odd bug-fix. Windows 10 receives huge "feature" updates every few months. There is no security benefit to new features, just the chance of disruption.
2) Modern Windows 10 updates do not have the same stringent testing and QA that updates for earlier Windows enjoyed. Not saying MS never released buggy updates previously, but you have to admit that problems are more common now.
Nobody is arguing much about security updates. The issues being faced by the majority of people however are coming about when big "Feature" updates are forced upon them. These don't provide security, but do leave a trail of borkage in their wake.
Barton Gellman is a little bit more than an average 'newspaper hack'. He led the Washington Post's coverage of the Snowden leaks, interviewing Snowden personally, and quite probably has had to deal with Information Security in an up close and personal way as part of that process - most of which he either can't (or won't) talk about. It's worth reading his Wikipedia profile.
This is not to say his real-word experience trumps Steve Bellovin's knowledge*, but I would say Gellman's background shows signs of at least having relevant knowledge. Purely in journalism terms, he has fairly good credentials:
Gellman has contributed to three Pulitzer Prizes for The Washington Post, winning as an individual, team member and team leader.
I wouldn't characterise him as a hack, and I'd lay good odds on him knowing the difference between Internet Explorer and Windows Explorer.
The actual issue is an interesting one; I suspect a nuanced answer might be appropriate: the average user is best served by applying updates as soon as practicable after they have been applied and found to be good by experts. This allows experts to use their expertise in recovering from unexpected glitches, and the population of average users benefit from across-the-board improvements in security. I suspect Microsoft, in part, moved to forced updates because so many people did not update, exacerbating security vulnerabilities. We all know of devices without firmware updates, even if they are available, because most people don't regard applying updates as important. Microsoft, in theory, should be ensuring that updates go without a hitch (i.e. they are the 'experts' people wait for): and for the most part they do, but failures are, quite rightly, high profile.
I think a good argument can be made for automated updates improving the general level of security, but I would also say that experts should be given the tools to opt out of updates where it is, in their opinion, necessary.
*Steve Bellovin's blog (SMBlog — Steve Bellovin's Blog) is always an interesting read for me. I have learned a great deal from it.
In general keeping the OS and applications updated is a 'good idea/. But the devil is in the details as to the precise installation timing. Some have to vet hardware or software to be sure something mission critical will work afterwards. Others need to wait to when they have some free time so the update does not interrupt work. Also, if there is no active exploit updates or serious security risk updates can be delayed to more convenient time. The idea that one must always update toot suit just because they are available is unrealistic and possibility catastrophic for the user. But must advice is not tempered with any clue about the real world but is only applicable to some academic dream world.
Minor nit. The first thing that generally happens when updates are released (okay, inflicted) is the rush to reverse the update to find what vulnerabilities are patched for by the malware communities. The tools are readily available, I've got most of them here, it's more a matter of how much effort is required for a particular level of return.
Therefore, if those vulnerabilities weren't being actively exploited, you can be assured in a few days/hours it will be actively exploited.
"The update debate has divided security experts about whether automatic updates are more trouble than they're worth because bad code can cause data loss."
There is no worse thing that software can do than cause data loss. That said, I think this is the wrong debate. The real debate should be around forcing people to update.
If someone wants to have updates automatically applied and are willing to accept the risk, I have no problem with that. But it should be optional. I think this is an ethical issue -- it is not ethically supportable to force updates on people, full stop.
Even when you force updates on people they find ways not to reboot their machines.
Auto updating is generally a good thing, Windows suffers because there is a massive range of hardware from a endless stream of manufactures all of which may not quite do things properly so there will always be issues. As for testing, how do test on an infinite range of device combinations?
They have a million odd people running in the insider program and there still issue (perhaps because a big chunk are running VMs and not primary devices).
Botnets have moved to the current big unpatched groups of device, the IoT, becuse Windows is not as easy as it was.
Personally autopatching does not bother me, there is enough ways to secure your data out there it wreally isn't difficult, however I realise some are so IT illiterate they can't do it so some data loss will be enevitable. However it far better than seeing your data encrypted and paying a ransome only for the key never to turn up.
Its your choice, the world is a crap place to live made worse by humans. Live with it.
1. MS don't pick up on the bugs reported in the insider program.
2. Their testing approach is wrong, MS should have common OEM equipment to test on and OEMs themselves should be able to test a pre-release version on their machines and be able to report errors back. An error would mean either Windows needs fixing by MS or the device driver needs fixing by the OEM, but the idea is not to send out an update which hoses thousands of computers.
3. Windows still reboots and applies updates when it feels like it, people in real life with work to do need to be able to keep their work.
4. If MS is part of making it the world worse, they're doing it wrong.
Two objections, in addition to what @Dan 55 said:
Botnets have moved to the current big unpatched groups of device, the IoT, becuse Windows is not as easy as it was.
Non sequitur. Is it necessarily the cause?
Botnets have moved on to IoT because they are an easier target with lots of them, not necessarily because Windows is "harder" now.
the world is a crap place to live made worse by humans.
The world is a crap place to live because of humans.
Planet Earth itself is a GREAT place to live. Try a walk in nature, in a green forest, or some mountains .. it does wonders.
It didn't give me a BSOD but did screw up the audio on my couple of weeks old HP laptop. First start after the update it decided there was no sound devices at all, after I reinstalled the initial driver the sound was weedy and had no bass. I eventually manually updated the drivers after several other attempts at updating to a more recent one failed. Only good thing in all this is it uses an SSD so reboots, which there were numerous, take almost no time.
Well done for getting it to work.
However given the complaints about updates removing manually loaded drivers esp Display Drivers I'd be prepared to repeat the operation each and every time that MS issues even the smallest security patch.
Such is the way with the redomnd control freaks these days. 'Do it our way and when we say or not at all'.
I'd be quite happy for a default auto-update and you can change that if you know what you're doing, approach.
A better idea would be the system update process to have an option to take a backup before applying system updates. BUT, this should also have the ability to defer the update until somebody is there to attach the backup medium or to use a permanently attached backup device, which could be anything from an external USB drive to a NAS box or the cloud.
Its all perfectly feasable: this would just automate what I've been doing for years with Linux:
(1) disable the auto update system
(2) manually make a backup immediately before triggering an update.
I do this on a weekly basis. Its a three step manual process (1) Backup, (2) System update, (3) reboot.
This has remained manual for three reasons: first because its never been enough of a nuisance to try automating it, second because the backup disks are stored offline and thridly I use encrypted partitions so the encryption password has to be entered on the local keyboard at boot time.
My house server, which runs 24x7, also keeps seven generations of a compressed nightly backup of user space on a permanently attached disk, but this is for fat finger and disk crash protection rather than surviving a system update. For the latter you need a full backup rather than something that just secures locally created files and data.
From the Article: "I cannot boot to safe mode and it does not allow me to do a system restore before the updates have gone through. …"
If the users statement is true, that's the real WTF here. Sure, changes can go wrong and you can't test for every configuration out there. But it's up to Microsoft to ensure there's a robust recovery mechanism before starting an update.
The best recovery mechanism for Windows is none other than the BOFH's rubber mallet.
Safe Mode is a really terrible design actually. It is supposed to only allow a subset of programs and drivers. But what if, like in this case, a keyboard driver fails? This happens in Safe Mode because, er, how else are you going to type?
Safe Mode is a really terrible design actually. It is supposed to only allow a subset of programs and drivers. But what if, like in this case, a keyboard driver fails? This happens in Safe Mode because, er, how else are you going to type?
Windows will load the generic keyboard driver in that case. It may not allow use of the special buttons, but it will be enough to let you fix whatever is wrong with the PC.
Yep, that'd be genius-- having an option within the Settings -> Update & Security that you can set once 'Force a system-image backup before any update' with perhaps a secondary 'Save space by only backing up system files' option. I honestly can't see a downside to that, except for the expectation that it increases the worry of 'Huh, updates could be bad' vibe, which, well-- seems just at the moment.
The HP/Update error would have killed 3 machines in four remote locations (so 12 in total), but thankfully we discovered it when two BSOD'd (deleting the hpq*.sys driver fixed those), and even luckier it was the failover machines it hurt, one lasting casualty remains that we have to fly back to HQ to diagnose. Maddening to say the least.
"A better idea would be the system update process to have an option to take a backup before applying system updates."
If we have to choose one or the other, the better idea is to let people turn off auto-updating. But, really, the best thing would be to allow people to turn off autoupdating and provide some automatic backup before updating.
Automatic backups are a difficult thing to accomplish, though, because you have to put that backup somewhere, which means the user is going to have to configure something.
Recently, we have had some problems with our latest Windows update, and we've had to pull it twice. We'd like first to apologize to all the users that this problem has affected. We didn't do the testing we should have, and we're sorry that our failures have caused trouble for you. We will be compensating the users that this has affected, and we will be changing our policies and practices to ensure that any updates we release go through much more stringent quality assurance and testing before we release them. We'd also like to apologize to those in our Windows Insider Program, who have generously tested our products without receiving the attention they deserve from Microsoft. For now, the Windows April 10, 2018 update (version 1803) will remain our latest release. We will continue to work on new features, but our major focus will be in fixing bugs and ensuring something like this doesn't happen again. For this reason, we won't be releasing a feature update to Windows this fall. We do plan to release one in the spring of 2019, which will be thoroughly tested before release. We hope you will continue to work with us and forgive us for this failure.
1 Microsoft Way
Both funny and sad how as soon as a read what looked like a reasonable and professional response from Microsoft I knew it was a joke by the forth sentence.
Still I think the main issues are 1) M$ QC and testing has gone to crap and 2) the nanny state of Win10 blocks you from setting manageable controls. They could get away with one or the other, but the combination of both is truly toxic, and WSUS is no escape hatch as it is buggy, miserable to use if you turn of auto approval, and a space hog unless carefully configured.
Zero tools provided to help their users implement the best practice acceptance setting, and no real hepl for alternative patching systems. Get out of the way of my MDM, or give me something at least as good as Munki.
As much as I hate Microsoft (and I do), this seems to be a symptom of some larger problem. Almost everything these days seems to be loaded with high-tech features that are of use to no one, are poorly or not-at-all tested and have bloated, broken and idiotic UIs.
There was a time when vacuum cleaners were made like tanks. There are probably perfectly functioning vacuum cleaners under 50 years of garbage accumulation out in the landfills of western civilization. The companies that made these vacuum cleaners are pretty much all long gone, once people had their tank-vacuums then they didn't ever need any new vacuums and the companies went out of business.
Over time durable goods became less and less durable as the time between one sale and the next depends on lack of durability. Now it is almost impossible to purchase anything that isn't garbage.
Of course software is not a durable good, but it seems to to be following the same pattern. Slap something together, make it pretty (for very strange values of "pretty"), load it up with bullshit and shove it out the door. Fitness for purpose isn't an afterthought, it's never thought of at all.
Something has gone very wrong.
In addition to an interest in computers, I spend a lot of time dealing with structural steel fabrication. It's astonishing how different the attitudes in these two worlds are. Computer programmers tend to take the approach of "it compiles - ship it". Security, efficiency, quality in general seem to be an afterthought at best. Structural engineering is governed by various industry codes, which make frequent use of the word 'shall', and often have the force of law. If a structure fails, and the designer or fabricator did not follow the relevant code(s), they can be held liable for that failure. I wonder if something similar is needed to tame the Wild West of programming: an RFC or ISO standard that establishes requirements for the design and testing of quality software. I wouldn't want it to be too restrictive, to allow development of new languages and methods of development, but at least set a minimum standard for proper engineering of software, extent of testing, what sorts of bugs identified during testing must be fixed before shipping.
It's probably impossible to prevent all bugs and security flaws, but we ought to be able to do something about this chronic parade of embarrassingly bad mistakes from companies that have the resources to do better.
"Computer programmers tend to take the approach of "it compiles - ship it". Security, efficiency, quality in general seem to be an afterthought at best. Structural engineering is governed by various industry codes, which make frequent use of the word 'shall', and often have the force of law. If a structure fails, and the designer or fabricator did not follow the relevant code(s), they can be held liable for that failure."
Only in your part of the world. Take a look in some second world countries. Codes and liable don't stop incompetent.
The client wants this thing,
the designer do that thing,
the contractor do another thing,
and you hope nothing will fail.
It's just that some places there are more competent people take responsibility. And you hope their responsibility reach far enough to stop failure.
AC due to the experience.
You've described the problem very well. The cause is actually quite simple but is likely to be unpalatable.
Software development is the only branch of engineering where the practitioner is entitled (indeed generally expected) to be entirely self taught and not validated to a ratified standard for professional competence. Such training as is available largely concentrates on a language, an 'environment' (i.e. a bunch of proprietary libraries) or a dev system, but the basic principles of sound engineering practice are not likely to be on the syllabus.
When I was practising just over 20 years back as a systems engineer with sole responsibility for automation and data acquisition on projects with multi-million budgets, it was automatically part of the culture to test and verify everything before releasing it into the production environment, just the same as was the case with the mechanical and electrical components of the systems. But the 'webbification' of things has led to the rise of more than one echelon of back bedroom 'developers' who uncritically mash up code from fragments found online, with no consideration of security, robustness, or, in many cases, even ergonomics.
One only has to examine the questions and answers on online 'developer' communities to recognise the shallowness of the general level of understanding, typified by instant launch into example code without any explanation or discussion of objectives or mechanisms. So Dunning and Kruger rule. Ignorance can not identify itself, and all the more when the general standard of ignorance makes it a societal norm.
It's just like aviation. We only learn lessons when people die, and we do learn because of either conscience or fear of consequences.
Somebody has to die because of bad programming habits. Only then is the world going to learn how to write better software.
Until then, beancounters rule and politicians are plain uninterested (because money is coming either way)
"Almost everything these days seems to be loaded with high-tech features that are of use to no one, are poorly or not-at-all tested and have bloated, broken and idiotic UIs."
"Something has gone very wrong."
There's a giant cloud of SMUG hanging over Redmond, these days. I blame THAT.
the 'next generation' of developers took over when 'Ape' released. It was "their turn now". And they're arrogant, smug, and don't care about the customer, because THEY know best! I could ALSO say something about them being raised by helicopter parents and indoctrinated by left-leaning teachers and even MORE indoctrinated by outright communist professors in college, which for them is a VERY recent memory, but I won't. Wait, I just did...
"There's a giant cloud of SMUG hanging over Redmond, these days. I blame THAT."
But this is a wider problem than just Microsoft. It's practically epidemic across Silicon Valley as well. I'm not sure what to root cause is, but I started noticing many years ago that the younger programmers entering the field, by and large, knew how to produce programs in specific languages and frameworks, but tended not to know anything but the very basics of software engineering.
I suspect that the real problem is that many people are entering the field solely because they want to make a fat paycheck, not because they want to be engineers, so they learn pretty much the bare minimum.
Exactly - it would be interesting to know what borked the driver - a bad kernel change, or a change that triggered some bad code issue in HP code which went unnoticed before?
Anyway, why MS thinkers with kernel code so often is a bit worryingly - it got issue with the Intel audio driver, now with the HP keyboard one.
Don't trust them any more..
Microsoft already "helped" with that update. I always liked Windows, but this upgrade... I have no backup, so I've been needing in recovery software. Btw, if someone is interested in - DiskInternals (Uneraser and Partition recovery) have a discount for software all October.
Or TestDisk or some other FOSS tool. They have 100% discount for eternity, and do the job better than commercial counterparts.
Why do you pay when somebody has an alternative that they are giving away (with code) for free, and all you have to do is test it and possibly contribute back or donate, or in the least, say thanks?
My HP laptop (ProBook 470 G5) seems okay but I did have a bit of hassle. After the reboot my screen was at full brightness and the hot keys to adjust it didn't work. There were no options in the Windows display settings page either. When I checked Device Manager there was an exclamation mark against the Intel display driver. I manually updated it and it found a newer version. After that the display settings page allowed me to adjust the brightness but hot keys still didn't work. Ran the HP hot keys installer and they came back.
Just had a Windows auto update and after login faced with blackscreen and message device not found, yet again, Outlook hanging. Same procedure, switch off router, reboot then power off, switch router on, then reboot again. Simple. So our thanks to MS for making our lives so exciting, stressful and saving us from having to do real work.
Consumer auto-updates is a good thing, is things bork for an extremely small percentage, well thats what backup to the cloud is for right, who wnats millions of zombie machines pushing of spam !
Business, well should never happen, as tehy should be usuing best practice patch testing and central deployment? After all, they have shareholders and profit margins to worry about, don't want big fines for exaple in the financial industry for upacthed holes?
Hey, it's almost like you should wait until the user is at a point that they feel confident they have backups of the system and everything they need before you push updates to them.
Almost like... you know... you should ask the user. "Hey, I'd like to update now... are you ready?" rather than just forcing it through anyway.
Hell, maybe even a 30-day "Have you backed up yet?" nag might not be enough... maybe they are desperately trying to get the system backed up to a sufficient state, and are struggling to do so as a home user on an OS with PRECISELY ZERO GUIDANCE on how to do just that, or restore it if it goes wrong. So maybe someone should a) ask them to backup, including asking them if they'd like to see the backup tools available, b) not proceed with an update until a backup has been taken and verified, c) not proceed with an update until the user has said "Yes, I have adequate backups" no matter how long that takes.
And maybe, just maybe, put an System Restore feature into the SO that actually fecking works?
Gosh, I wonder what the easiest solution would be to this dilemma if you don't want to hand-hold your users through this, but also don't want to trash their systems by updating before they've managed to take a backup.
I don't know... allow them to indefinitely postpone updates maybe?
whenever I update FreeBSD's kernel and install it, the old kernel is preserved, and if something goes wrong, I can boot the old kernel and fix things.
Micro-shaft: take a lesson from open source. Give us THAT option on bootup after your "cram-it-up-our-asses" 'up'dates FAIL...
either that or let us EASILY AND COMPLETELY DISABLE THE UPDATES [make them 'manual only'] like we USED to do, instead of making it a choice between "now" or "later" (and then sneak it in when we're not looking).
Firstly, he doesn't offer an argument, you quote a tweet, which is an assertion. It's his point of view,
Secondly, if there are any statistics on this, let's see 'em.
There may be users who for whatever reason don't mind a bit of downtime, but people who own their computers, and want to do, you know, work, might want to choose when or whether to update
With over 30 years working with PC's in the enterprise and for personal use, I can't remember a single time not having an OS update resulted in data loss, BSOD, security breaches or other glitches. On the other hand, the number of time an auto-update has resulted in data loss, BSOD, security breaches or other glitches in just the last few years will soon take more than the fingers of two hands to count.
Microsoft have repeatedly shown that they are incompetent, bordering on malicious, when it comes to delivering software and updates that actually work. We've seen time and time again that Windows has dragged its bloated carcass over the line: it will never be reliable again. Microsoft has lost control.
Yet many of you continue to self-flagellate by installing this muck, then getting all surprised and outraged when once again, a Microsoft product doesn't work. Every article like this gets a plethora of Windows installation and update based woes from people who surely must've read some of the millions of other articles and comments stating how unreliable and shoddy Windows is, and yet carry using it and taking the pain regardless! Why haven't you Windows users learned? I truly don't get it.
The culture of today is to defend shit. If somebody kicks you in the face, it is your duty to jump to their defence and make up a story to make it right. M$ destroys you computer and behaves recklessly? It is the user's fault. Forcing something on an unwilling victim is good. Treating people with respect is bad. M$ aren't alone. Tesla sell electric cars then hobble the battery life to cut the range down. Then they sell the extra capacity as an 'upgrade'. Fine. But in times of distress like a hurricane they reach in and OTA update your property to increase the range. What happened to buying something then owning it. Why do companies feel they have the right to alter things after the sale is made? Years ago these practices would be illegal and the companies brought to justice quickly. Now we get to hear screwed customers defending it. Go figure.
Auto updates for SECURITY patches are a good idea, Forced feature updates every 6 months are NOT a good idea.
All versions of Windows 10 should have the option to receive only security updates and turn off feature updates. Currently I have to set my Windows 10 PCs network adapter to a metered connection to stop it download feature updates but there should be a proper way of switching it off in the OS settings.
Sure the odd security update breaks stuff but not as frequently as these feature updates do.
As for the new features, why not let US decide what ones we want to download and install?
You mean MS should let you beta test whichever version of Windows you wish? That doesn't make any sense. They know which version they need tested, which is why they assigned you to the task of testing that specific version.
Microsoft Updates put 1809 on both my laptop and main PC and just this Tuesday, had 3 more updates.
I noticed a problem with my hp Pavilion M6 laptop when the switchable graphics weren't working as it should. For example: when playing a game, I set the Catylist program to use the AMD RADEON 7470M Graphic Card instead of the built-in Intel 4000 Chipset. With the new updates, Switchable Graphics weren't working as I was getting between 2 and 0 FPS.
Thought it was the game, it was working fine in Version 1803.
So yesterday, I decided to rollback the update...but that process encountered a problem and had to reboot. When it did, the OS wouldn't start and I kept getting an ERROR 0xc0000225 message. Kept getting it over and over again after I tried my best to run the Repair Tool and System Diagnostics from the new installation Bootable USB I created.
After 3 hours of frustration, I couldn't do anything and am now looking at having my laptop serviced by my company's IT Guys today...which will cost $$$. It's either that or a full system wipe, which I can do myself.
Moral of the story here is that Micro$oft shouldn't have rushed into this, shouldn't have released it to be automatically downloaded by a select grouping...which, BTW, I'm NOT part of the Microsoft Insiders people...and they might have to pay for what their Frankenstein Monster has done.
HP Devices seem to be HP Computers, where a HP Keyboard can be attached. Did anyone experienced in Microsoft Logic, try to boot their HP Device without a HP Keyboard attached, and check if that circumvents the blue screen error WDF_VIOLATION ? If so, you might get a clue whether you have a Genuine HP Keyboard or Chinese import version with keylogger inside.
I have driver updates turned off. This was in response to never obtaining quality driver updates from Intel or Dell for the 7260 AC wireless adapter. New adapters previously downloaded, either by the Win 10 updates or by me trying to find a fix for intermittent outages, left me with a buggy wireless adapter and the decision to leave the stock MS drivers as installed and quit farking with it, which actually solved the problem. It seems this might have also prevented some of the recent issues reported due to the Win 10 update. I'm thinking I should probably set the delayed updates to the 35 day limit. I've never had an issue with Win 10 updates but think I should probably take some preventative measures as I'm due for borkage.
Shutting down last night I realised too late that the update to Win 10 1803 had slipped in to my HP laptop and I feared the worst today when I switched in on. However, so far, so good.
I already had a 4 day delay on "quality updates" ROFL...
...but now upped it to 7 days.
"Feature updates" (ROFL hyterically...) are set on a maximum 365 days delay.
That Linux Mint install is looking more tempting by the day.
"auto-updates has been one of the few bright spots"
Why can no one write software that is free of bugs, and does not require updates?
Self driving cars don't have a hope in hell.
The world needs one common patent free software frame work with no variations that we can focus on removing all of the bugs from.
My Android devices update the applications automatically on a regular basis, so in this context auto updates are probably a very good thing.
However Android usually gives you the choice when to update the OS.
Whilst you are still in scope for an OS update, of course.
Support for consumer versions of Windows is a real pain in the neck for Microsoft. A one off payment back in the mists of time and the expectation of continuous free support until the heat death of the PSU and beyond. In itself this is not a sustainable business model.
The proposed solution - force migrate everyone to the same version of the operating system and force everyone to stay up to date - probably looked good in theory. Only one version to support and the prospect of a subscription model in the misty future to offset the enormous cost of ongoing support. As far as I can see that isn't working out yet.
Eventually something has got to give; the constant expectation of free software and support amongst consumers is going to be up against the eventual realisation that ad-supported software isn't giving advertisers value for money. Who then pays for the "free" software?
IoT is the prime candidate for automatic security updates. However the security implications of giving write access to devices on your home network to some potentially fly by night cowboys are not good. This is another case where pay once and expect free support forever will not cost in.
TL;DR we're screwed.
"My Android devices update the applications automatically on a regular basis, so in this context auto updates are probably a very good thing."
Mine don't, because updates problems do happen, so it's important for me to be able to choose if/when I'm ready to take that risk.
This is for folk that have managed to install Windows 1809 (from the v1 release of Win10 1809).
This week Microsoft released a cumulative update for 1809 KB4464330, this will always fail on first install (which messes up the Update History aesthetically) due to a missing service stack update KB4465477.
It will normally install on the second or third time.
To get round this, after installing Windows 1809 from ISO, download KB4465477 manually from WinCatalog, and install. **before checking for updates**.
This will update the 1809 Service Stack.
Updates should install correctly after this manually applying update.
Windows Update...more Clunk, than a rusty bag of nails.
Biting the hand that feeds IT © 1998–2021