For those who don't support Windows Update handling UEFI/BIOS Updates . . .

[britechguy]

. . . the machine I have in my hands now is exactly why I do.

An old HP EliteBook 8560P with an i5 2nd gen in it. Started out life with Windows 7, but was upgraded to Windows 10.

The BIOS/UEFI that was still on board was Version F22 dating from 2011. The latest version from HP for this machine, which I've manually downloaded and installed, is F67 from 2018.

If HP supplies such updates for Windows Update to apply, it would be better if it did. I know that end users are not going to do it.

 

[NETWizz]

. . . the machine I have in my hands now is exactly why I do.

An old HP EliteBook 8560P with an i5 2nd gen in it. Started out life with Windows 7, but was upgraded to Windows 10.

The BIOS/UEFI that was still on board was Version F22 dating from 2011. The latest version from HP for this machine, which I've manually downloaded and installed, is F67 from 2018.

If HP supplies such updates for Windows Update to apply, it would be better if it did. I know that end users are not going to do it.

Unless it is fixing an issue like Meltodwn or Spectre, I do not see a need to just push firmware automatically... at least unless it is exceptionally well tested.

Sure, I like the latest firmware, but think of the disaster if we have 12,000 laptops update and even 1% fail to boot.

 

[Markverhyden]

Yep. I'm far more comfortable managing my own stuff. That being said it is a valid point with consumers. Even today, with virtually all BIOS updates running from within the OS, customers still won't keep up with them. I think a happy middle ground would be to set a flag during OOBE.

 

[britechguy]

Even today, with virtually all BIOS updates running from within the OS, customers still won't keep up with them.

Yup.

And since I haven't had a BIOS/UEFI update "go south" in more than 10 years, I have zero fear of them now. And the few that I did have go south were not after a successful flash, but because something went tragically wrong during the flashing process (and who knows what that was).

If the device manufacturers cannot be trusted to only issue well-tested BIOS/UEFI, then no one can.

For the vast majority of computer users, automatic updating is the very best option. But I have no objections to there being a way for the tech savvy to disable this. Very few people would.

 

[fincoder]

Automatic UEFI updates have been rock-solid in my experience. With today's complex UEFI firmware and potential security exploits, I think firmware updates are almost as important as OS updates.

I want UEFI updates installed on a new computer, to fix anything found after manufacturing which could be 6 months earlier. If it's recommended or automatic, then the vendor warranty will cover it if bricked. Better to do that before setting up for the end user.

 

[NETWizz]

Automatic UEFI updates have been rock-solid in my experience. With today's complex UEFI firmware and potential security exploits, I think firmware updates are almost as important as OS updates.

I want UEFI updates installed on a new computer, to fix anything found after manufacturing which could be 6 months earlier. If it's recommended or automatic, then the vendor warranty will cover it if bricked. Better to do that before setting up for the end user.

They probably technically wouldn't fix it if you specifically told them it died during a firmware update; however, they would probably never ask and ultimately it would probably be covered... or if you politely rejected it and raised a concern holding firm, you would likely still get it covered. Also, if you get told "no," you can call back and get someone different. Usually that is the trick with big companies.

Depending upon what I read, many sources say BIOS updates aren't covered, yet other sources say they are. I think in writing they are NOT, but in practice they actually are.

 

[HCHTech]

My concern stems from the number of calls we've had where a customer repeatedly "got tired of waiting" for an update to install and forced the computer off in the middle - probably while staring at the screen that says "Do not turn off your computer" - ha. This confession usually isn't offered at the outset, either...you have to drag it out of them.

Edit: I see TN helpfully tells me that the :rollseyes: emoticon is my most-frequently used....directly correlated with the frequency I deal with end-users, I suspect!

 

[britechguy]

My concern stems from the number of calls we've had where a customer repeatedly "got tired of waiting" for an update to install and forced the computer off in the middle - probably while staring at the screen that says "Do not turn off your computer" - ha.

The problem(s) being:

1. This sort of behavior is stupid, and you can't fix stupid.

2. It can apply to any sort of major update.

3. UEFI/BIOS updates tend to be some of the fastest, and the "most active on the screen," of all of them. And most of that "turn off" kind of behavior is due to "nothing happening."

In the end, and regardless of the type, I have become firmly in favor of update automation to the maximum extent. Over the course of my career I have seen (and had to clean up) more smoldering heaps that could be traced back to refusal to update "because Tom's friend down the street said I shouldn't" than I ever have from the much maligned, but seldom occurring in practice "bad update."

I'd rather take my chances with keeping things up to date than the opposite. And when it comes to end users, they will never, ever be on top of all of this. When it comes right down to it, given the pace of updating these days in any number of areas, I couldn't be even if I wanted to be as far as getting them all in if I had to research them. I'm happy to take Windows Update, Intel Driver & Support Assistant, HP Support Assistant, and AMD Radeon Software updates. If the companies that produce the hardware and software don't know what they're doing in 99.99% of the situations, I have no hope of even coming close to that average.

 

[SAFCasper]

3. UEFI/BIOS updates tend to be some of the fastest, and the "most active on the screen," of all of them. And most of that "turn off" kind of behavior is due to "nothing happening."

Agreed I can understand a client powering off during a feature update when for example it's stuck at the same place for 30+ minutes and looks like nothing is happening. Not the same as a UEFI update which take on average less than 5 minutes with a constantly moving progress bar.

Comparison of sorts - what about firmware updates to iOS and Android? They come through automatically. Powering off during a firmware update can easily brick the device. Do people here also have concerns about this? If not then why not? Maybe because they have a battery and can't be "unplugged" ?

 

[Porthos]

Not the same as a UEFI update which take on average less than 5 minutes with a constantly moving progress bar.

I have them just seem to be off during the restarting stage. It can be concerning for the average user who is not patient.
also, with laptops, Users are in a hurry and shut it and go somewhere.

 

[Sky-Knight]

EFI based systems have an OS on the BIOS now. It has security issues, it must be updated quarterly, and I'm rapidly approaching monthly. The idea that firmware should be updated only when needed to fix some issue is outmoded. Brains must be updated.

Users are going to have to let update processes complete, or they're going to have to get used to paying us to fix bricks. These are their choices. The more automation we can bring to this table the better. And again, brains must be updated.

Firmware updates for Android won't apply unless forced when the device isn't plugged in. At least... that's how GOOGLE does it, Samsung and the rest muck with this process.

Failed BIOS updates are replaced main boards under warranty... typically. There's no way to tell the difference between a bad BIOS and defective power circuit when troubleshooting.

 

[britechguy]

I have them just seem to be off during the restarting stage. It can be concerning for the average user who is not patient.
also, with laptops, Users are in a hurry and shut it and go somewhere.

The following isn't aimed at you, but I'm sorry, BIOS/UEFI updates are about as clear as you can get that there will be a restart after flashing. I actually expect users, all users, to read what is presented to them (other than EULAs) and the dialogs for a BIOS/UEFI update are very, very brief indeed. The whole process is very, very brief indeed.

If it requires, as @Sky-Knight said, learning the hard way by paying for bricked devices, then that's what's going to have to happen. It's quite like getting many users to accept that doing routine, cyclic full system image backups is not optional, but essential insurance. Many only finally accept and acknowledge this after their first total loss of data with no ability to recover anything.

I also refuse to accept, at this point in time, that anyone who's been using a computer for any extended period of time is not familiar with update processes in general. There have been plenty of Feature Updates during the Windows 10 era and of differing complexities so that pretty much any variant has really already occurred, and more than once.

I can't make people learn, or pay attention, but I have every reason to believe that they have enough experience that if they ever cared to learn by paying attention that this has already occurred.

The PC world in 2021 is very, very far from new and most of my clients, of any age, now have years of ownership under their respective belts. If they insist on making foolish choices, I certainly cannot be held responsible, nor can companies trying to keep them safe by auto-updating things that desperately need it.

 

[Sky-Knight]

Exactly, they make dumb decisions, and then pay people like us to fix it.

The only part I worry about is the encryption, because that's not fixable...

 

[britechguy]

The only part I worry about is the encryption, because that's not fixable...

Hence the reason I hate it being overly broadly applied, and these days it absolutely is.

It gives a very false sense of security and very few understand what potential disaster lurks, particularly if they don't take backups. One small "burp" and your stuff is gone, gone, gone with no chance of recovery.

 

[Sky-Knight]

@britechguy and Microsoft has no interest in "fixing" that problem, because it sells more Onedrive.

I want to be annoyed about that... but I just cannot be. Endpoints must be disposable, that's for everyone's sanity!

 

[britechguy]

Endpoints must be disposable,

And we shall have to hold diametrically opposed opinions on that.

The computer that sits in front of me should be "the bastion" and first and most reliable source of data. That doesn't mean that occasional breakdowns may not occur nor that I should not have backups, but cloud storage should be entirely optional.

 

[NJW]

The following isn't aimed at you, but I'm sorry, BIOS/UEFI updates are about as clear as you can get that there will be a restart after flashing. I actually expect users, all users, to read what is presented to them (other than EULAs) and the dialogs for a BIOS/UEFI update are very, very brief indeed. The whole process is very, very brief indeed.

I agree that the visible part of a firmware upgrade is sufficiently chatty to keep users from interrupting it and it doesn't take long, but the reboot can be very silent and quite long.

I updated the firmware on a Lenovo (consumer) laptop yesterday, by remote connection, but with the user in attendance. Applying the update was straightforward, warning before starting that a restart would be required, with progress reported throughout. However, after briefly flashing on the screen that it was finished, the restart was immediately initiated (no further prompt or request to proceed) and the machine sat with a black screen for several minutes, but with the lights on. I was mentally moving later appointments so I could go out and rescue it, when it came back up.

I'm sure that if I hadn't been in telephone contact with the user, they would have done something to hurry it up. If it had been here, on the bench, I would have been saved by taking time to Google the symptoms before intervening.

 

[britechguy]

I'm sure that if I hadn't been in telephone contact with the user, they would have done something to hurry it up.

It's interesting how there do seem to be two rather distinct "camps" when things like this occur: those that get impatient after less than 5 minutes time and want to hurry things up (and start all sorts of stupidity in hopes of doing so) and those that freeze like they think the computer has, refuse to touch it, and call me.

I don't know why, but the preponderance of folks who've somehow ended up as my clients fall into the second camp. That could also be the result of things I've told them over time, though, but a great deal seem to start in that camp. I routinely tell my clients several things:

1. If you're asked whether you want to install/allow/change something and you have no idea how you should answer, the correct initial answer is always No or whatever refusal the dialog gives you.

2. Lengthy installs often have one or more "stall points" where it looks like absolutely nothing is going on for a very long period of time. Just wait (and wait and wait) and in almost all cases you'll end up going from wherever you were in the process to way further along in an instant when things "let loose."

3. Random button pushing, particularly when frustrated, is the way to madness and destruction.

 

[GTP]

Random button pushing, particularly when frustrated, is the way to madness and destruction.

The root cause of ID10T errors.

 

[britechguy]

It's funny, but today I applied the latest BIOS/UEFI update to the Acer TC-885 I recently acquired. This was, by far and away, the most "primitive" of all the updaters I've used in recent years, but it was still entirely straightforward and gave the user the warnings they would have needed.

The only thing it didn't do, and I think it should have, is mention that the reboot that comes after this part completes:



is particularly long, slow, and unusual. The machine not only rebooted, but clearly was doing something akin to restarting itself several times without anything ever showing on the screen, at all, it remained black. I watched the power light turn off and on several times over what felt like forever, even for me, and it should be no issue to add the information about the odd nature of the reboot to the message.

All of the other recent BIOS/UEFI updates I've done occurred almost entirely within Windows itself, and then on reboot was nothing more than the standard "white blocks status indicator" on the black screen prior to Windows booting and away you go. The Acer process was slower and longer and a long period of time was spent on the screen shown above.

But if you can read warnings like that one, no matter who you are, and ignore them there is no one to blame but you.