Why do we have to keep rebooting things?

Everything from satellite receivers to computers seems to need
unplugging from the mains so often and plugging back in. How come? Why
don't they make these things so that when they're incapable of doing
their job they just automatically reboot? Couldn't satellite receivers
(for instance) just run some sort of self-check routine in the
background and if it fails do a reboot? (and come back on the same
channel of course?).

Bill

In article de898962-6535-4431-9faa-
,
wrote:
Everything from satellite receivers to computers seems to need
unplugging from the mains so often and plugging back in. How come? Why
don't they make these things so that when they're incapable of doing
their job they just automatically reboot? Couldn't satellite receivers
(for instance) just run some sort of self-check routine in the
background and if it fails do a reboot? (and come back on the same
channel of course?).

It would need software to detect the problem and make the decision to do
this - but it's usually software that is the cause of the problem in the
first place. Catch 22.

There's no substitute for a brain, and a finger on a button.

Or a big hammer.

Rod.
--
Virtual Access V6.3 free usenet/email software from
http://sourceforge.net/projects/virtual-access/

" wrote in
message
...
Everything from satellite receivers to computers seems to need
unplugging from the mains so often and plugging back in. How
come? Why
don't they make these things so that when they're incapable of
doing
their job they just automatically reboot? Couldn't satellite
receivers
(for instance) just run some sort of self-check routine in the
background and if it fails do a reboot? (and come back on the
same
channel of course?).

Bill



You can actually make Windoze do that - but if it then detects a
problem during any subsequent boot-up it will keep restarting
itself ad infinitum.


--
Woody

harrogate three at ntlworld dot com

" wrote in message
...
Everything from satellite receivers to computers seems to need
unplugging from the mains so often and plugging back in. How come? Why
don't they make these things so that when they're incapable of doing
their job they just automatically reboot? Couldn't satellite receivers
(for instance) just run some sort of self-check routine in the
background and if it fails do a reboot? (and come back on the same
channel of course?).

Bill

You need to talk to some telecoms people. Telecoms require around 6-9s
reliability (i.e. dial-tone 99.9999% of the time - work out how many seconds
a year you can be "down").

Actually this is incredibly difficult using modern software because as
others have said, there is just so much of it and it can go wrong in so may
ways. For example consider the following code...

if (it's monday)
{
if (coronation street is on)
{
record it (1)
}
else
{
wait... (2)
}
else if (eastenders is on)
{
record it (3)
}
else
{
do something else (4)
}

There are 4 different blocks of "code" to test. Now software has 1000s of
such blocks and unlike in this case, they are often not independent. Think
perhaps of some code which does something if Coronation Street is on, or
East Enders, or both. You might think that these would be independent but
the programs are on different MUXs so two tuning operations not one, and
what if you're watching a third Mux? Oh, and you only have one hard disk to
record onto so the two recordings have to share - you get the idea!

So for something like a PVR, if things do go wrong it's far better to leave
the smartest device in the room (normally), i.e. the owner to decide what to
do next rather than take a guess which might be totally the wrong thing to
do.

For example if my PVR locks up, as it has done occasionally, I never reboot
until after the "thing it should be recording" has finished. Sometimes it
makes no difference but sometimes it actually is recording even though it's
not responding to commands.

Of course with telecoms there isn't that luxury but then we would be talking
of powerful boxes costing a large amount of money and the cost of specialist
hardware and the like can be justified.

So, all down to bangs per buck. If all you want is something that does all
that a video recorder can then fine, but PVRs do far, far, more and for far
less cost these days - but you get what you pay for.

Paul DS

wrote:

Couldn't satellite receivers
(for instance) just run some sort of self-check routine in the
background and if it fails do a reboot? (and come back on the same
channel of course?).

My sony IDTV used to do just that, bloody thing; might only have
happened once every few months, but always at the wrong moment.

Roderick Stewart wrote:

It would need software to detect the problem and make the decision to do
this - but it's usually software that is the cause of the problem in the
first place. Catch 22.

Sometimes the watchdog approach does work, so long as it is taken by the
developer as a sign that there is a problem rather than that it is the cure.

On Tue, 28 Sep 2010 05:55:03 +0100, Roderick Stewart wrote:
In article de898962-6535-4431-9faa-
,
wrote:
Everything from satellite receivers to computers seems to need
unplugging from the mains so often and plugging back in. How come? Why
don't they make these things so that when they're incapable of doing
their job they just automatically reboot? Couldn't satellite receivers
(for instance) just run some sort of self-check routine in the
background and if it fails do a reboot? (and come back on the same
channel of course?).

It would need software to detect the problem and make the decision to do
this - but it's usually software that is the cause of the problem in the
first place. Catch 22.

What you're asking is commonplace in embedded software. it's known as a
watchdog timer and is implemented in hardware. Briefly, every few
seconds a hardware reset is sent to the processor _unless_ in the
preceeding time it has been cancelled by code in the main programme
that says "it's OK, I'm still running normally" - at which point the
countdown to the next hardware reset starts again.

So far as consumer electronics goes, the reason it fails is because
it's not very well written. Often due to time constraints on the
developers (or lack of a decent specification, or last-minute changes)
or cheapness in the hardware or plain ineptitude - though often the
root cause is a combination of all of these.

Just be gald that stuff which really counts, such as the engine
management unit in your car is written to professional standards
and is rigorously tested (well, usually) to a much higher standard.

--
http://www.thisreallyismyhost.99k.or...9011619151.php

wrote:
Everything from satellite receivers to computers seems to need
unplugging from the mains so often and plugging back in. How come? Why
don't they make these things so that when they're incapable of doing
their job they just automatically reboot?

I read a wonderful piece of problem analysis/engineering design.

A team of programmers were working on an engine management system.
Now a car engine is a lousy place to put a computer; it's full
of vibration, noise and high frequency, high voltage electricity.

No matter how many fixes they added to the EMS they simply couldn't
make it run for more than a few minutes before it crashed.

And the cost of the EMS was spirally with all the "bomb proofing"
they were adding.

The eventual solution?

Solve a different problem.

They put a lot of effort into making the EMS boot up nearly
instantaneously, and rebooted it (from a simple timer)
every couple of seconds!

BugBear

On Tue, 28 Sep 2010 08:49:59 +0100, Paul D Smith
wrote:

Actually this is incredibly difficult using modern software because as
others have said, there is just so much of it and it can go wrong in so may
ways. For example consider the following code...

if (it's monday)
{
if (coronation street is on)
{
record it (1)
}
else
{
wait... (2)
}
else if (eastenders is on)
{
record it (3)
}
else
{
do something else (4)
}

Oh the irony. Syntax error: mismatched braces. Computer went bang.
This post was inadequately tested before the Send operation was initiated.

So Clive Sinclairs claim that a nuclear power station could run on a zx81 is
a little irrisponsible then.
Brian

--
Brian Gaff -
Note:- In order to reduce spam, any email without 'Brian Gaff'
in the display name may be lost.
Blind user, so no pictures please!
"pete" wrote in message
...
On Tue, 28 Sep 2010 05:55:03 +0100, Roderick Stewart wrote:
In article de898962-6535-4431-9faa-
,
wrote:
Everything from satellite receivers to computers seems to need
unplugging from the mains so often and plugging back in. How come? Why
don't they make these things so that when they're incapable of doing
their job they just automatically reboot? Couldn't satellite receivers
(for instance) just run some sort of self-check routine in the
background and if it fails do a reboot? (and come back on the same
channel of course?).

It would need software to detect the problem and make the decision to do
this - but it's usually software that is the cause of the problem in the
first place. Catch 22.

What you're asking is commonplace in embedded software. it's known as a
watchdog timer and is implemented in hardware. Briefly, every few
seconds a hardware reset is sent to the processor _unless_ in the
preceeding time it has been cancelled by code in the main programme
that says "it's OK, I'm still running normally" - at which point the
countdown to the next hardware reset starts again.

So far as consumer electronics goes, the reason it fails is because
it's not very well written. Often due to time constraints on the
developers (or lack of a decent specification, or last-minute changes)
or cheapness in the hardware or plain ineptitude - though often the
root cause is a combination of all of these.

Just be gald that stuff which really counts, such as the engine
management unit in your car is written to professional standards
and is rigorously tested (well, usually) to a much higher standard.

--
http://www.thisreallyismyhost.99k.or...9011619151.php