On Fri, 6/5/2026 12:16 PM, Ed Cryer wrote:
I keep getting fooled by these happening unknown to me. When they're in progress, some of Windows' functions just stop being available, but they don't let me know; they just hang.
Here's a typical example that's just occurred.
I ran DISM in CMD, preparatory to an SFC scan. It finished, I started the SFC, but it just hung there. Then I set about looking for the cause. It took some time to realise that a system update must be in progress.
Is there a way to have the updates inform me that one's in progress? I can't find one in Settings.
Ed
It's complicated, and by the looks of it, the activity
is spewed all over the place on purpose.
There can be a notification in the task bar, when a reboot
needs to be done (but typically the reboot request appears
before it is really time to be doing the reboot). A reboot
notification can be too late, to factor into your usage
of SFC.
There is Hours of Activity. It defines when you're working,
for the "purposes of reboots". I have noticed, that major updates,
they are "time sensitive" but this is for the fakey "saving
the planet on CO2 by delaying your update" logic. Which does
not save on CO2 and is a failed policy.
Windows Update has two kinds of updates, it has Patch Tuesday (in-band)
and it has out-of-band or emergency updates. A typical emergency update
might be to repair a bug in a Patch Tuesday.
Windows Update has a "Pause" function of up to five weeks. That's
not really much of a control. And, it does not Pause Everything,
it's not that clever.
There are Driver Updates. They can be turned off, in cases where
your choice of 580, conflicts with the 550 that Microsoft keeps
trying to serve. If you are a 3D gamer and are using the latest
NVidia, the Microsoft part will keep trying to back out your
bleeding edge driver.
Security definition updates come in several times per day.
Windows Metro.Apps update async to Windows (OS) Update. For example,
yesterday, I had around 23 of them, but they might not
affect SFC.
The Windows Update for Patch Tuesday, the installation will stop services,
and you would hope, start services again. This can have spill-on effects
for the services the user maintenance activity is relying upon (the
user notices that a number of things on the OS aren't working
properly at the moment and oh, wouldn't a reboot be nice right now).
*******
Things that change the disposition of the OS.
1) A reboot tends to flush annoying activity. Who knew.
2) Pulling the network cable, removes a root of trust, if
the OS is "checking" some materials it is using for maintenance.
By pulling the network cable, you can have some peace and quiet
while you work.
You can couple these two things together to make a PC you yourself
can use, even if you cannot use the network because of it. An extended Microsoft ICE Panorama Run, that's what I had to do, to get it to
run to completion. Pull the network cable, let the good times roll.
This is all, well, "so so convenient". Like driving a bus, by
tying a rope to the bumper, and putting the rope over your
shoulder and hauling the bus around with you. You will need
broad shoulders and a sense of humour, to be a "bus driver".
So it is with Windows.
Fortunately, Linux *also* has this habit. Just not on all distros.
Any companies that offer service contracts, they tend to have
update policies rivaling Microsoft. For example, Ubuntu can
start updating, or thrust an update window into your screen
when you don't want it. Linux Mint can do this too, but
you can dismiss the window and it will come back, and you
can dismiss it again. But officially, you have control.
Whereas Ubuntu, I've even power cycled the machine in the
middle of an update, to show it who is boss, corrupting
the sequence, and it *still* with one leg cut off it
and hobbling along, it still tried to install stuff,
blew errors and so on. It's not like this practice of
Microsoft, has not had an impact on how other product
offerings treat their customers.
Any time I do experiments, I have to set aside one hour
of misery for clearing updates, so the experiment is
run on a stable footing. It's a giant waste of time
for any materials which might be of a temporary nature.
I roll back a lot of VMs, after an experiment, and
only occasionally do I "keep" the updates bubbling in.
*******
There are third party tools for controlling Windows Update,
but this does not cover enough of maintenance behavior
to give the user absolute control over everything.
For example, if I want to do an HDTune disk benchmark
run, on purpose Microsoft will start various activities
in the background, screwing up the benchmark. You can find
a dozen trace files in the OS, "sleep studies", all
sorts of trash. It's not like there is any possibility
of "sane business practices" in this thing.
Apparently, in a SKU you are not using, InTune can
define a "maintenance window", but Microsoft
is still clueless on what customers want as of course
this does not work properly.
I hear that Windows 2000 runs pretty well, and is
sweetness and light regarding disturbing the user
while they are working. I can run an HDTune there
and there would be no glitches in it. My copy cost me
$300 at Circuit City (a company no longer in business),
back in the day, the software being locked in a
glass cabinet in the main aisle. But I did
"use the hell out of the thing", so I got my moneys worth.
Paul
--- PyGate Linux v1.5.15
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)