Happy said:I have been using a computer for 20+ years, and have never
really
heard an answer to the question-- it is better to turn it off
overnight, or when you're not using it, or leave it on?
Malvern said:I basically agree with Ken. I have been at this for some seven
years--old Windows 95- and back then it was advised to turn
things
off to permit "housecleaning".
Well, I've tried both leaving computers on, and turning them off.
Basically, the advice is true for different reasons on both sides
of the argument!
If a system is left on for extended periods off time, it will tend
to be more reliable hardware wise while it's on. On the other hand
turning such a system off, then back on has a high failure rate
than those turned off every day. This sounds a bit odd, but what
really happens is that components eventually degrade, and degrade
faster (in most cases) when hot. So, the system left on tends not
to fail until it is turned off and back on.
The system that is turned off and back on more often has a better
chance of not failing when it's turned back on.
In either case, the likely failures are more or less the same.
(Power supply, hard drive etc.) The one exception occrs with a
computer that does not have a UPS. Shut it down if any power
interruptions or major surges are anticipated. Befoer a storm,
it may be a good idea to disconnect it from power and phone
lines.
My first minicomputer 1973
First home computer 1979 (Apple ][+)
First PC 1990
First Laptop 1997
w_tom said:First, go to those data sheets. Component life expectancy
is more commonly expressed in terms of hours of operation.
Hours of operation best measures degradation of hardware.
When is seriously degraded hardware most likely to demonstrate
the problem? When turned on. The so gentle event called
startup is often enough to expose a damaged component.
Light bulbs also demonstrate this concept. Many assume that
power on causes light bulb damage. However that is bluntly
disputed by numbers from light bulb manufacturers. Light bulb
life expectancy is determined by voltage and hours of
operation. When does a light bulb fail on turn on? Look for
dark spots inside the glass. Damage previously created by
hours of operation. The gentle transient created at power on
is then enough to cause bulb failure - typically in the last
10 hours of light bulb life expectancy.
This concept also applies to computers. The naive assume
that power on damaged computers only because the computer
failed when powered on. Their observation alone was
sufficient to make it fact? Instead, the computer failed
because some component had too many hours of operation - and
was about to fail anyway. Unlike a light bulb, we cannot see
dark spots where 'filament damage' appears inside
transistors. But failures are most often measured in hours of
operation. Some components may have manufacturer defects;
therefore fail prematurely. But the failure is created by
hours of operation - not by power cycling. Power on simply
exposes a pre-existing problem created by too many hours.
Second, experimental evidence alone is not sufficient to make
conclusions. Experimental evidence is only enough to
speculate. Conclusions require both experimental evidence in
agreement with the underlying theory. Both conditions are
necessary to avoid junk science. Provided are the underlying
concepts that explain why computers fail. If power cycling
causes premature component failure, then the claim is
supported by manufacturer facts or by datasheets from
component manufacturers. Leaving a computer running does not
significantly extend computer life expectancy. Numbers are
required to dispute this statement.
There are two sides to this argument. The first is based on
speculative assumptions and urban myth reasoning. The second
is supported by manufacture numbers, underlying theory, and
even explains why failure is observed at power up. IOW the
first is classic junk science reasoning. The second is what
was taught in junior high school science. A conclusion must
be based both upon underlying concepts AND experimental
evidence. Only the second meets junior high school science
criteria.
Power it down or put it to sleep when done. Don't fall for
myths that power cycling does damage. Notice those myths
never provide numbers to underlying concepts. If power
cycling was so destructive, then also leave all televisions
and radios powered on.
Well, I've tried both leaving computers on, and turning them off.
Basically, the advice is true for different reasons on both sides
of the argument!
If a system is left on for extended periods off time, it will tend
to be more reliable hardware wise while it's on. On the other hand
turning such a system off, then back on has a high failure rate
than those turned off every day. This sounds a bit odd, but what
really happens is that components eventually degrade, and degrade
faster (in most cases) when hot. So, the system left on tends not
to fail until it is turned off and back on.
The system that is turned off and back on more often has a better
chance of not failing when it's turned back on.
In either case, the likely failures are more or less the same.
(Power supply, hard drive etc.) The one exception occrs with a
computer that does not have a UPS. Shut it down if any power
interruptions or major surges are anticipated. Befoer a storm,
it may be a good idea to disconnect it from power and phone
lines.
My first minicomputer 1973
First home computer 1979 (Apple ][+)
First PC 1990
First Laptop 1997
w_tom said:First, go to those data sheets. Component life expectancy
is more commonly expressed in terms of hours of operation.
Hours of operation best measures degradation of hardware.
When is seriously degraded hardware most likely to demonstrate
the problem? When turned on. The so gentle event called
startup is often enough to expose a damaged component.
Light bulbs also demonstrate this concept. Many assume that
power on causes light bulb damage. However that is bluntly
disputed by numbers from light bulb manufacturers. Light bulb
life expectancy is determined by voltage and hours of
operation. When does a light bulb fail on turn on? Look for
dark spots inside the glass. Damage previously created by
hours of operation. The gentle transient created at power on
is then enough to cause bulb failure - typically in the last
10 hours of light bulb life expectancy.
This concept also applies to computers. The naive assume
that power on damaged computers only because the computer
failed when powered on. Their observation alone was
sufficient to make it fact? Instead, the computer failed
because some component had too many hours of operation - and
was about to fail anyway. Unlike a light bulb, we cannot see
dark spots where 'filament damage' appears inside
transistors. But failures are most often measured in hours of
operation. Some components may have manufacturer defects;
therefore fail prematurely. But the failure is created by
hours of operation - not by power cycling. Power on simply
exposes a pre-existing problem created by too many hours.
Second, experimental evidence alone is not sufficient to make
conclusions. Experimental evidence is only enough to
speculate. Conclusions require both experimental evidence in
agreement with the underlying theory. Both conditions are
necessary to avoid junk science. Provided are the underlying
concepts that explain why computers fail. If power cycling
causes premature component failure, then the claim is
supported by manufacturer facts or by datasheets from
component manufacturers. Leaving a computer running does not
significantly extend computer life expectancy. Numbers are
required to dispute this statement.
There are two sides to this argument. The first is based on
speculative assumptions and urban myth reasoning. The second
is supported by manufacture numbers, underlying theory, and
even explains why failure is observed at power up. IOW the
first is classic junk science reasoning. The second is what
was taught in junior high school science. A conclusion must
be based both upon underlying concepts AND experimental
evidence. Only the second meets junior high school science
criteria.
Power it down or put it to sleep when done. Don't fall for
myths that power cycling does damage. Notice those myths
never provide numbers to underlying concepts. If power
cycling was so destructive, then also leave all televisions
and radios powered on.
Well, I've tried both leaving computers on, and turning them off.
Basically, the advice is true for different reasons on both sides
of the argument!
If a system is left on for extended periods off time, it will tend
to be more reliable hardware wise while it's on. On the other hand
turning such a system off, then back on has a high failure rate
than those turned off every day. This sounds a bit odd, but what
really happens is that components eventually degrade, and degrade
faster (in most cases) when hot. So, the system left on tends not
to fail until it is turned off and back on.
The system that is turned off and back on more often has a better
chance of not failing when it's turned back on.
In either case, the likely failures are more or less the same.
(Power supply, hard drive etc.) The one exception occrs with a
computer that does not have a UPS. Shut it down if any power
interruptions or major surges are anticipated. Befoer a storm,
it may be a good idea to disconnect it from power and phone
lines.
My first minicomputer 1973
First home computer 1979 (Apple ][+)
First PC 1990
First Laptop 1997
Humans tend to believe the first person's lies and then denyYou are a complete idiot.
Hi,
Interesting responses from everyone. My main computer at home has
been powered on for 4 years plus. Only time it has been turned
off was during maintenance, like cleaning the cooling fans and
blowing out dust. I haven`t had any problems with that system.
As to another point, what is the average life expectancy of a
personal computer now a days. I know people who replace thier
computers every few years, others do upgrades of components
when new ones come out. For most PC`s, they rarely ever live out
their minimum or maximum life expectency. So most of the statisics
are worthless, since computers and parts tend to be replaced,
upgraded, or updated before they usually reach their mininum life
potential.
I know lots of companies that never turn off the computers and
those computers are 3 to 5 years old. They occasionally have a
peice of hardware go bad, but that is expected. Same rule applies
to pretty much anything, it could last years or break down in 10
minutes, one can never tell when something will go bad.
Jeff
"David Candy" <.> wrote in message
Cycling does cause wear from thermal expansion and contraction.
--
----------------------------------------------------------
http://www.uscricket.com
w_tom said:First, go to those data sheets. Component life expectancy
is more commonly expressed in terms of hours of operation.
Hours of operation best measures degradation of hardware.
When is seriously degraded hardware most likely to demonstrate
the problem? When turned on. The so gentle event called
startup is often enough to expose a damaged component.
Light bulbs also demonstrate this concept. Many assume that
power on causes light bulb damage. However that is bluntly
disputed by numbers from light bulb manufacturers. Light bulb
life expectancy is determined by voltage and hours of
operation. When does a light bulb fail on turn on? Look for
dark spots inside the glass. Damage previously created by
hours of operation. The gentle transient created at power on
is then enough to cause bulb failure - typically in the last
10 hours of light bulb life expectancy.
This concept also applies to computers. The naive assume
that power on damaged computers only because the computer
failed when powered on. Their observation alone was
sufficient to make it fact? Instead, the computer failed
because some component had too many hours of operation - and
was about to fail anyway. Unlike a light bulb, we cannot see
dark spots where 'filament damage' appears inside
transistors. But failures are most often measured in hours of
operation. Some components may have manufacturer defects;
therefore fail prematurely. But the failure is created by
hours of operation - not by power cycling. Power on simply
exposes a pre-existing problem created by too many hours.
Second, experimental evidence alone is not sufficient to make
conclusions. Experimental evidence is only enough to
speculate. Conclusions require both experimental evidence in
agreement with the underlying theory. Both conditions are
necessary to avoid junk science. Provided are the underlying
concepts that explain why computers fail. If power cycling
causes premature component failure, then the claim is
supported by manufacturer facts or by datasheets from
component manufacturers. Leaving a computer running does not
significantly extend computer life expectancy. Numbers are
required to dispute this statement.
There are two sides to this argument. The first is based on
speculative assumptions and urban myth reasoning. The second
is supported by manufacture numbers, underlying theory, and
even explains why failure is observed at power up. IOW the
first is classic junk science reasoning. The second is what
was taught in junior high school science. A conclusion must
be based both upon underlying concepts AND experimental
evidence. Only the second meets junior high school science
criteria.
Power it down or put it to sleep when done. Don't fall for
myths that power cycling does damage. Notice those myths
never provide numbers to underlying concepts. If power
cycling was so destructive, then also leave all televisions
and radios powered on.
Well, I've tried both leaving computers on, and turning them off.
Basically, the advice is true for different reasons on both sides
of the argument!
If a system is left on for extended periods off time, it will tend
to be more reliable hardware wise while it's on. On the other hand
turning such a system off, then back on has a high failure rate
than those turned off every day. This sounds a bit odd, but what
really happens is that components eventually degrade, and degrade
faster (in most cases) when hot. So, the system left on tends not
to fail until it is turned off and back on.
The system that is turned off and back on more often has a better
chance of not failing when it's turned back on.
In either case, the likely failures are more or less the same.
(Power supply, hard drive etc.) The one exception occrs with a
computer that does not have a UPS. Shut it down if any power
interruptions or major surges are anticipated. Befoer a storm,
it may be a good idea to disconnect it from power and phone
lines.
My first minicomputer 1973
First home computer 1979 (Apple ][+)
First PC 1990
First Laptop 1997
w_tom said:Again, the damning fact that "leave it on" advocates must
avoid. Why do they turn off their TVs and radios? Notice they
fear to respond to that question.
....blah, blah, blahKen Blake said:In
I am not an "always leave it on" advocate, but for the moment I'll pretend
that I am and answer your question:
anyone said:Ken are you telling us you don't have backups.