XPost: alt.comp.os.windows-11   
   From: nospam@needed.invalid   
      
   On Sat, 1/24/2026 11:26 PM, Jeff Barnett wrote:   
   > On 1/24/2026 7:23 PM, Paul wrote:   
   >> On Sat, 1/24/2026 8:29 PM, Jeff Barnett wrote:   
   >>> On 1/24/2026 5:37 PM, ...w¡ñ§±¤ñ wrote:   
   >>>> Jeff Barnett wrote on 1/24/2026 12:56 PM:   
   >>>>> On 1/24/2026 10:07 AM, ...w¡ñ§±¤ñ wrote:   
   >>>>>> Jeff Barnett wrote on 1/24/2026 9:07 AM:   
   >>>>>>> On 1/23/2026 12:44 AM, ...w¡ñ§±¤ñ wrote:   
   >>>>>>>> micky wrote on 1/22/2026 4:26 PM:   
   >>>>>>>>>   
   >>>>>>>>> I suppose most of the readers of these newsgroups know that for a   
   full   
   >>>>>>>>> start from scratch it's required to run Restart, instead of Shutdown   
   >>>>>>>>> followed by starting later.  Even though I only learned this a year   
   ago   
   >>>>>>>>> and even though it's a bit counter-intuitive, you all know that,   
   right,   
   >>>>>>>>> (because if you're running with Fast Startup*** shutdown saves the   
   >>>>>>>>> kernel and reuses it, but Restart never uses Fast Startup and it   
   >>>>>>>>> rebuilds the kernel, with new values if the values have changed).   
   >>>>>>>>>   
   >>>>>>>>> But maybe you did not know that you can get the benefit of Restart   
   and   
   >>>>>>>>> still turn your machine off if you press Shift while clicking on   
   >>>>>>>>> Shutdown.   How about that!!   
   >>>>>>>>>   
   >>>>>>>>   
   >>>>>>>> If Hibernation (and by extension, Fast Startup) is disabled (e.g.,   
   via powercfg -h off), the "Shutdown" button functions as a full, traditional   
   shutdown - clears the system state(shuts down all processes) and clears temp   
   memory and powers Off.    
   Subsequent Powering on, re-initializes the o/s   
   >>>>>>>>    - which, incidentally is the same results as Shift-Shutdown.   
   >>>>>>>>    i.e. you don't need to the keystroke combo if Hibernation is   
   disabled.   
   >>>>>>>>   
   >>>>>>>> Restart(when Hiberation is disabled) does the exact same clearing but   
   instead of shutting down, powers the device back on and re- initializes the   
   o/s.   
   >>>>>>> I have Hibernation set to never in my power profile. However, I have   
   an APC UPS that will hibernate the machine in case of an extended power   
   outage; in other words, the APC software is able to work around that setting.   
   Does this setup change    
   anything said in this thread so far?   
   >>>>>>   
   >>>>>> Afaik, the UPS Hibernation feature requires enabling Windows   
   hibernation.   
   >>>>>>    i.e. When a power outage occurs, the UPS has to communicate with   
   Windows and validate Windows hibernation state(enabled/disabled), if enabled   
   instruct Windows to hibernate, once done the UPS provides backup battery power   
   for the duration of    
   its capability.   
   >>>>>>   
   >>>>>> Lacking Windows hibernation, the UPS only provides battery power for   
   the device(and Windows) for the duration of its backup battery capability.   
   >>>>>>     - no Windows hibernation   
   >>>>>   
   >>>>> I've not enabled hibernation (as said above) but the APC software seems   
   to be able to do it anyway. Perhaps it changes the power profile when   
   necessary. I remember one of the disk tester/diagnostic packages that turned   
   sleep off when it was called    
   and never bothered to restore it on its way out. So there is certainly some   
   way to diddle the settings.   
   >>>>   
   >>>> See Paul's response, that echo's my input.   
   >>>>    => Hibernation needs to be enabled in Windows.   
   >>>   
   >>> I believe (from my own experience) that Paul is incorrect in this specific   
   case. The APC software (OS level stuff) can when it decides it's time to   
   shutdown because of battery drain, etc, can arrange that hibernation works   
   whether the user has    
   enabled it or not. I've noted it's happening a few times.   
   >>   
   >> I think there has to be a hiberfile (C:\hiberfil.sys) and   
   >> it is typically equal in size to 50% of memory or a larger   
   >> percentage. Via compression, the RAM content written out to   
   >> the hiberfile is smaller than the physical size. Most sessions   
   >> on Windows are quite small, so only a fraction of the reserved   
   >> hiberfile space is used.   
   >>   
   >>     powercfg /a     # check supported ACPI states.   
   >>                     # Hibernation is S4. Mine is   
   disabled right now.   
   >>                     # A   shutdown /h   would   
   generate an error of some sort.   
   >>   
   >>     powercfg /h on  # Add Hibernation, then verify with the first   
   command   
   >>                     # Check for hidden file on the root   
   of C: with dir /ah in command prompt   
   >>   
   >> There is also some procedure at startup, which may be writing   
   >> something to the header of the hiberfile. The hiberfile is "invalidated"   
   >> as part of startup, such that it cannot be reused immediately to   
   >> restore the same session ("like it was GroundHog day").   
   >>   
   >> Otherwise, without using the hibernation feature, little in the   
   >> way of wear, happens because of the existence of the hiberfile.   
   >>   
   >> If you were installing the APC software, it is possible there is   
   >> sufficient elevation at install time, for the APC software to be   
   >> using powercfg and switching the hiberfile on and reserving the   
   >> necessary space.   
   >>   
   >> Once the hiberfile exists, the normal set of system commands   
   >> should work, to generate a hiberfile. The command from the UPS   
   >> should be terse -- it isn't always a super-intelligent processing   
   >> thing. Any application loaded into the host OS, has to do the   
   >> more intelligent part. The UPS should not be sending this   
   >> command, as written. That would be an exploit waiting to happen.   
   >>   
   >>     shutdown /h   
   > Paul, the power plan/profile that I've selected specifies "never" for   
   hibernate. However, there is a ~27GB hiberfil.sys for my 64GB memory on the C   
   disk. I presume that the APC software ensures that such a file exists and   
   knows how to flip bits in the    
   header to inform the BIOS that this is a hibernate. There was no hiberfil.sys   
   with the "same" software and power plan/profile on my prior Windows 7 machines   
   so I assume the necessary structures were consed when needed. I believe that   
   APC takes    
   responsibility to ensure this all works properly - I believe the majority of   
   their users are office works using company machines. It would be lunacy on   
   their part to expect them to know how to dick around with power plans, etc. In   
   fact, their IT    
   department has locked them out of anything at all interesting.   
      
   I don't think there is a way, to tell who did that :-)   
      
   You could check the Event Viewer, for the evidence, but I don't   
   know if it identifies the executable that is doing it.   
      
   In any case, the weird hiberfil.sys size, suggests Windows   
   didn't do that. I thought 50% was a "popular" Windows choice.   
   While there are ways to select custom percentages, there is   
   likely to be some lower limit on size choice.   
      
      
   [continued in next message]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)