Is it worth it to do a full shutdown of Windows 8/10 from time to time

[Plusck]

9 hours ago, Camacha said:

Does it do anything useful? That is an awful lot of continuous power.

I do the same - but it isn't continuous power. When the computer goes to sleep it goes into a very deep sleep - there is still a tiny bit of power consumed (or at least it is needed: turn it off at the wall and it'll reboot) but the fans are all off and there's no noticeable heat given off. This has been the default behaviour for years.

[Pawelk198604]

On 13.06.2016 at 6:47 AM, RainDreamer said:

I didn't know about the shift shut down thing, neat!

So now you know

I learned about it from computerworld 

On 14.06.2016 at 4:18 AM, adsii1970 said:

I restart my desktop every 14 days. I've disabled the EnergyStar crap on it and when it is not in use by me or the family, I am running BOINC software for the Einstein project and SETI@Home.

The laptop gets restarted once a week, but when the screen is down, it is in standby mode.

The tablet is set to completely shut down when not in use after it has been in standby mode for two hours, or if I manually shut it down when I am finished with it.

 

 

On 14.06.2016 at 4:44 AM, CliftonM said:

Have you tried out Prime95?  That's the one where you use your computer to look for prime numbers.

 

On 14.06.2016 at 4:45 AM, adsii1970 said:

No, I have not. Is that part of the overall BOINC project?

I once participated in SETI@home, but it was 5 years ago. 

I think project is dead now

But does all this BIONIC projects serve any purpose?   

[adsii1970]

11 hours ago, Pawelk198604 said:

I once participated in SETI@home, but it was 5 years ago. 

I think project is dead now

But does all this BIONIC projects serve any purpose?   

Seti@Home is far from dead. It just no longer receives federal (U.S. government) funding. There are hundreds of projects that operate through the BOINC software/torrent.

To answer your second question, yes, they do. It allows the research to be generated over a large number of computers once you have set your preferences up. I set mine to run at night (from 11:00 p.m. to about 4:30 a.m.) and when it has been screen saver mode for over an hour. It is as simple as visiting the BOINC website and choosing the projects you want to participate in.

[Camacha]

22 hours ago, Plusck said:

I do the same - but it isn't continuous power. When the computer goes to sleep it goes into a very deep sleep - there is still a tiny bit of power consumed (or at least it is needed: turn it off at the wall and it'll reboot) but the fans are all off and there's no noticeable heat given off. This has been the default behaviour for years.

If it sleeps it has not been running constantly

[razark]

My desktop rarely gets shut down, but I reboot it at least once a week..  It depends on what I have running at the time.

My (alleged) laptop, however, has been continuously running for about 892 days without a reboot.

[Plusck]

4 hours ago, Camacha said:

If it sleeps it has not been running constantly

I interpreted that as "constantly powered on". Deep sleep counts as "powered on" IMO, at least that's how I'd interpret the behaviour under all of the various varieties of OSX I've used. Windows 10 seems to take this further, saving the current state of the machine to disk (which is what OSX does if your laptop battery runs out) which is not the same as the standard "sleep" option.

But yah, there was a little cross-reading between your posts and MajorJim's.

[Camacha]

2 hours ago, Plusck said:

I interpreted that as "constantly powered on". Deep sleep counts as "powered on" IMO, at least that's how I'd interpret the behaviour under all of the various varieties of OSX I've used. Windows 10 seems to take this further, saving the current state of the machine to disk (which is what OSX does if your laptop battery runs out) which is not the same as the standard "sleep" option.

But yah, there was a little cross-reading between your posts and MajorJim's.

I would classify neither sleep and hibernation as powered on, as both literally turn off all or most of the system. The power is actually gone. Sleep still needs power for the memory, hibernation is basically a shutdown without the actual shutdown procedure.

[Plusck]

37 minutes ago, Camacha said:

I would classify neither sleep and hibernation as powered on, as both literally turn off all or most of the system. The power is actually gone. Sleep still needs power for the memory, hibernation is basically a shutdown without the actual shutdown procedure.

Occam's razor: do you need power? Yes / no

If yes, the computer is "on", if no, it is "off".

[Camacha]

Just now, Plusck said:

Occam's razor: do you need power? Yes / no

If yes, the computer is "on", if no, it is "off".

False. A computer is comprised of a large amount of subsystems, many of which have the ability to locally turn on and off. Also, it would mean laptops do not power down or off. It could run without any power at all!

[StrandedonEarth]

17 minutes ago, Camacha said:

It could run without any power at all!

Not if you also remove the battery (or the battery is non-functional)

[Plusck]

10 hours ago, Camacha said:

False. A computer is comprised of a large amount of subsystems, many of which have the ability to locally turn on and off. Also, it would mean laptops do not power down or off. It could run without any power at all!

The ability to turn on or off locally still requires power.

When a computer is "off", there is only one thing still running and that is the system clock running off a small battery. Remove that and you lose the current date and time but that's all. In this state, throwing the whole thing into a bath then drying it thoroughly will make zero change to its operating state or any stored data. It is "off".

When a computer is "on", you can't do that. Removing batteries - even for a fraction of a second - will cause substantial data loss (whether or not that data is important is irrelevant - and quitting applications and shutting down services will obviously leave virtually nothing of any importance to be 'lost').

With modern laptops, tablets and phones, there is an apparent blurring of boundaries, but it is only "apparent". Chuck an incriminating phone or laptop into a river, fish it out and repair the shorted power circuitry, and it'll boot just fine. Data loss will be zero (with the exception of the current date) if you turned it "off" first, and non-zero if it was "on". As far as consumer-oriented electronic devices are concerned, there is only a real blurring of boundaries for things like cars: turning a modern car "off" leaves a lot of subsystems still running which will lose data (sometimes significantly) if you then roll the car into a river. At the same time, merely turning the engine off (for some cars - I know for a fact that this can happen for BMWs) will cause data loss from some of the error-logging buffers. There is therefore a blurring of boundaries since the car has several different levels of being "on", and is only ever "off" when you need to change the battery, with no user control over the processes.

But computers are not cars, we still have control over these functions, and there is a very definite difference between being "on" (and requiring power to be available at all times) and being "off". Windows 10 has started to blur this boundary by making a stable, unpowered state of hibernation the default when you shut the computer down, but it's the exception.

So I stand by what I said before. My computer thinks it has been running for a week. I would say it has been "on" for a week. For about half of that time it has been using only a tiny amount of power, but has needed that power to be available without even a pause of a fraction of a second.

I mentioned Occam's razor simply because the question, "do I need power", is the one that requires the fewest assumptions about the state of the multiple subsystems. If I lose data by removing power, or if removing power makes a radical change to how the device behaves when power is restored, then the device was "on". There will surely come a day when, like many cars today, our computers are never fully "off" unless they're taken apart, but we haven't reached that stage yet.

Edited June 16, 2016 by Plusck

[Camacha]

1 hour ago, Plusck said:

So I stand by what I said before. My computer thinks it has been running for a week. I would say it has been "on" for a week. For about half of that time it has been using only a tiny amount of power, but has needed that power to be available without even a pause of a fraction of a second.

I mentioned Occam's razor simply because the question, "do I need power", is the one that requires the fewest assumptions about the state of the multiple subsystems. If I lose data by removing power, or if removing power makes a radical change to how the device behaves when power is restored, then the device was "on". There will surely come a day when, like many cars today, our computers are never fully "off" unless they're taken apart, but we haven't reached that stage yet.

You can only stand by what you said before because you conveniently choose to interpret the conditions in your favour. Hibernate has been a significant part of Windows for a long time now. As you indicate correctly, it is a massive exception to your fairly makeshift throw-it-in-the-river-rule. For all intents and purposes, a computer is off. The same goes for the data loss, which really needs to be bent to your definition to work out - the time and date is data too. Therefore, according to your definitions, any computer with a clock battery counts as powered on, which basically means all of them.

Also, the premise that on and off are mutually exclusive and that any power in the system means it is powered on is false. It would not only mean that the battery would count all computers as always powered on, but it also totally ignores the fact that computers have many parts that can partially or completely turn off independently from the rest of the computer. It also ignores other edge cases, such as WoL, where the computer is turned off, but still has the capability to detect a signal to turn on. According to your story, you would consider such a computer to be turned on.

Even though it is a neat conversation trick to pretend you apply Occam's razor, as it appears to give weight to your point of view, the story completely unravels as soon as you look into it. It is not the simplest explanation, it is a contrived construct to support your point of view. The only properly unpowered computer is one with all the battery removed and the plug unplugged, but that is unnecessarily strict and not the convention in every day conversation. A computer than has been turned off can be fully shut down (but plugged in and with a battery) just fine, just as it can be put in hibernation. Even sleep makes a reasonable candidate for the moniker powered down, though I accept it could lead to some discussion. Calling it properly turned on, however, obviously does not fly.

If you want to be selectively strict with definitions and reasoning, I do not mind, but we should take it for the artificial construct that is.

[Laie]

On 6/14/2016 at 7:26 PM, Plusck said:

 a very deep sleep - there is still a tiny bit of power consumed (or at least it is needed: turn it off at the wall and it'll reboot)

That's "susped-to-RAM". Very similar conceptually to suspend-to-disk, but faster; actual power demand in that state is typically on the order of 5W, at least half of which is losses in the power supply. With a crummy PS it can be a lot more, but then it would probably be warm to the touch.

[Plusck]

1 hour ago, Camacha said:

You can only stand by what you said before because you conveniently choose to interpret the conditions in your favour. Hibernate has been a significant part of Windows for a long time now. As you indicate correctly, it is a massive exception to your fairly makeshift throw-it-in-the-river-rule. For all intents and purposes, a computer is off. The same goes for the data loss, which really needs to be bent to your definition to work out - the time and date is data too. Therefore, according to your definitions, any computer with a clock battery counts as powered on, which basically means all of them.

Also, the premise that on and off are mutually exclusive and that any power in the system means it is powered on is false. It would not only mean that the battery would count all computers as always powered on, but it also totally ignores the fact that computers have many parts that can partially or completely turn off independently from the rest of the computer. It also ignores other edge cases, such as WoL, where the computer is turned off, but still has the capability to detect a signal to turn on. According to your story, you would consider such a computer to be turned on.

Even though it is a neat conversation trick to pretend you apply Occam's razor, as it appears to give weight to your point of view, the story completely unravels as soon as you look into it. It is not the simplest explanation, it is a contrived construct to support your point of view. The only properly unpowered computer is one with all the battery removed and the plug unplugged, but that is unnecessarily strict and not the convention in every day conversation. A computer than has been turned off can be fully shut down (but plugged in and with a battery) just fine, just as it can be put in hibernation. Even sleep makes a reasonable candidate for the moniker powered down, though I accept it could lead to some discussion. Calling it properly turned on, however, obviously does not fly.

If you want to be selectively strict with definitions and reasoning, I do not mind, but we should take it for the artificial construct that is.

I'm not "conveniently" doing anything.

I consider "on but in deep sleep" to be "on". If I touch the keyboard the fans spin up and my desktop and applications are instantly available - exactly as if I'd simply turned the monitor off for a second. Any interruption in power will be fatal to any unsaved work. So it is "constantly on" as per the original post that all this relates to, and also "on" by that suggested Occam's razor analogy: make the least assumptions about the state of the system, therefore consider removing power (which also is the very basic definition of "on" or "off" in any event) and see whether that changes things significantly; if it does, it's "on".

Given that computers are not intended to be clocks, losing the date and time is hardly essential and it would be obtuse to claim that the computer is "on" just because the clock battery is connected.

In other words, I am trying very hard to avoid all the artificial constructs that you are trying to strawman into the conversation.

You could even avoid the "power" question by trying another test, such as removing vital bits and pieces and replacing them, and seeing whether that affects your system. Start with the keyboard, monitor, hard drive, graphics card, RAM, maybe even the CPU... without even touching the power supply, how much can you remove from your system before you risk damaging it? From experience, I know that I can go all the way down to the RAM (haven't tried CPU) on a computer that is "off" but with the power supply connected, and cause no damage or data loss, but unplugging a keyboard will wake it if it is "on" and in deep sleep.

So yes, I'm trying to give lots of nice clear, binary definitions of "off" and "on". They're also very useful everyday definitions. If I go out to the shops and get caught in a thunderstorm, or meet a friend who wants to fly me to a desert island for a few days, do I worry that I left the computer "on" or not?

And they're also perfectly normal, everyday, valid and workable definitions for telephones, tablets, TVs or whatever. If I turn my phone off it is most definitely off, not just sleeping or silent or in airplane mode or whatever. If you're at the theatre and an announcement is made asking people to turn off their phones, nobody does it and everybody knows that they're obeying the spirit, rather than the letter, of the instructions. Before airplane mode was invented, however, everybody knew that "off" meant "off" when cabin crew made the announcement, so phones would go "off" and laptop owners would hope they could get away with "on but closed and sleeping". Likewise, if a policeman stops you and asks you to turn the engine "off", he does not mean "wait until the onboard computer decides you're at rest and cuts the engine".

And your WoL example has actually become, in time, a further addition to this very simple and clear definition of "off" and "on": Windows 8 and 10, and Mac OS, do not allow WoL for a computer that is "off" (according to my "can you remove power without changing the state" definition), rather than just in deep sleep.

And that WoL question also leads to the ACPI specifications: G0 "working", G1 "sleeping", G2 "soft off" and G3 "mechanical off". Guess what - my binary "on" and "off" neatly splits those four states along the most logically obvious line, with an odd case at the very bottom end of G1 (S4 Hibernation or Suspend to Disk) which is also, neatly, exactly what we agree is getting blurred with Windows 10's default "off" behaviour. OSX is simpler to categorize since it doesn't hibernate to disk, requires power to avoid data loss and is therefore always "on" until powered down. And this is what we were talking about from the very beginning.

So no, I'm not trying to be "selectively strict" in definitions but rather trying to be as broad as possible, considering technical definitions and common speech and normal user expectations.

[Camacha]

1 hour ago, Plusck said:

So no, I'm not trying to be "selectively strict" in definitions but rather trying to be as broad as possible, considering technical definitions and common speech and normal user expectations.

Rehashing the same argument in different words does not change a lot. If you wish to construct this definition for yourself, that is great, but it is in no way, shape or form logical or universal, nor are any shaving utensils involved. As the arguments remain the same, you will find the replies in my previous post. Since this discussion is no longer serving any other purpose than wanting to be right, I will leave you with a few simple defintions:

Quote

Off: out of operation or effective existence:

Turn the lights off.

 

Quote

Off:

a :  to a state of discontinuance or suspension <shut off an engine>

 

 

Quote

Off: a machine or piece of electrical equipment that is off is not switched on or is not being used

Edited June 16, 2016 by Camacha

[Plusck]

3 hours ago, Camacha said:

Since this discussion is no longer serving any other purpose than wanting to be right, I will leave you with a few simple defintions:

Fair enough... but one parting shot deserves another:

Next time I have an opinion about my own computer, I'll run it by you to check it isn't false, makeshift, contrived, mere trickery or selectively strict.