real-time system - a word processor

Hi

Clarification on "real-time system":

A word processing system, if it is very slow to respond, say it takes
15mis or longer after typing each key for it to print to the document,
would you consider it to be a "real-time system" ?

Thanks

2005 wrote:
> Hi
> 
> Clarification on "real-time system":
> 
> A word processing system, if it is very slow to respond, say it takes
> 15mis or longer after typing each key for it to print to the document,
> would you consider it to be a "real-time system" ?
> 
> Thanks
> 

Next homework question?

-- 

Tauno Voipio
tauno voipio (at) iki fi

2005 skrev:
> Hi
>
> Clarification on "real-time system":
>
> A word processing system, if it is very slow to respond, say it takes
> 15mis or longer after typing each key for it to print to the document,
> would you consider it to be a "real-time system" ?

What do you mean by "15mis"? 15 minutes?

As for "real time" or not, the system is "real time" if all
operations are completed "sufficiently fast". The hard part
is, of course, to determine exactly what "sufficiently fast"
means...

Rune

> > Clarification on "real-time system":
> As for "real time" or not, the system is "real time" if all
> operations are completed "sufficiently fast". The hard part
> is, of course, to determine exactly what "sufficiently fast"
> means...

An excellent answer!

We all picked up on it quickly that 2005's questions gives away his age
and experience.  His perspective is limited by experience (or lack of
it) and so the language he uses and his expecations from it are
similarly limited.  We were all inexperienced like 2005 is at one point
in our lives.

I'm not lecturing but just noting that language will bite you in the
ass every time you fail to use it accurately.  2005 needs to understand
that every word he writes is significant and weighty.  It's my pet
peeve with my colleagues who say one thing and mean another.  That's
why I'm writing this.

JJS

On 2006-12-01, Rune Allnor <allnor@tele.ntnu.no> wrote:

>> A word processing system, if it is very slow to respond, say it takes
>> 15mis or longer after typing each key for it to print to the document,
>> would you consider it to be a "real-time system" ?
>
> What do you mean by "15mis"? 15 minutes?
>
> As for "real time" or not, the system is "real time" if all
> operations are completed "sufficiently fast". The hard part
> is, of course, to determine exactly what "sufficiently fast"
> means...

I would say that a real time system is one where there are
_requirements_ that things have to be completed within certain
time limits.  Even if they aren't completed sufficiently fast
it's still a real-time system, it's just one that's broken.

-- 
Grant Edwards                   grante             Yow!  A dwarf is passing
                                  at               out somewhere in Detroit!
                               visi.com            

johnspeth@yahoo.com skrev:
> > > Clarification on "real-time system":
> > As for "real time" or not, the system is "real time" if all
> > operations are completed "sufficiently fast". The hard part
> > is, of course, to determine exactly what "sufficiently fast"
> > means...
>
> An excellent answer!

If you say so...

The answer was inspired by a request to contribute to a marine
operations project a few years ago. The customer was about to
deploy this piece of equipment at the sea floor, with very tight
bounds on positions. Consequently, they needed to monitor the
positions of the gear as it was deployed, keeping track of position
beakons as they were lowered down the water column.

In our brief the objective was to keep the beacon positions updated
in "real time". A general procedure was outlined where all sorts
of bells and whistles were included; orthogonal codes, signalling
robust with respect to Doppler, you name it (I can't help thinking
the spec was a cut'n paste of the GPS position system with some
more constraints added...).

As I looked into this, I never really figured out why "Doppler
resistance" would be necessary. The laying vessel would
travel at 1 knot at most, inducing insignificant Doppler.
As for orthogonal code requirements, the deployment velocity
would never exceed 2 m/s, meaning that the position could
be triangulated to within 5 m in five seconds or so, by
coordinating reference beacons. Even if such coordination
should not be possible, using different frequency bands
for different beacons ought to be possible. Real time?
Marine operations work on a time scale a few orders
slower than most other operations. One position update
every minute would be plenty; one would not be able
to react faster, and it would take at least five minutes
from corrective actions were initiated onboard the vessel to
effects would be observable in the path of the gear to
be installed.

Maybe not surprising, then, that the project was cancelled...

> I'm not lecturing but just noting that language will bite you in the
> ass every time you fail to use it accurately.  2005 needs to understand
> that every word he writes is significant and weighty.  It's my pet
> peeve with my colleagues who say one thing and mean another.

Heh, don't get me started along those sorts of lines... ;)

Rune

2005 wrote:

> A word processing system, if it is very slow to respond, say it takes
> 15mis or longer after typing each key for it to print to the document,
> would you consider it to be a "real-time system" ?

I used to know someone who supposedly (I never actually saw it) could
type faster than the PDP-10 interrupt routine could process characters
from the input port.  In that case, it would not be a real time system.

-- glen

glen herrmannsfeldt wrote:
> 2005 wrote:
> 
>> A word processing system, if it is very slow to respond, say it
>> takes 15mis or longer after typing each key for it to print to
>> the document, would you consider it to be a "real-time system"?
> 
> I used to know someone who supposedly (I never actually saw it)
> could type faster than the PDP-10 interrupt routine could process
> characters from the input port.  In that case, it would not be a
> real time system.

I can recall a system (Tymshare IIRC), which I believe was PDP10
based, and which echoed back char by char, on which there was a
roughly 5 minute delay in echoing back anything.  The system was
unusable, and we rapidly dropped it.  Our terminal was a Teletype
at 110 baud.  This was about 1971 or 72.

-- 
Chuck F (cbfalconer at maineline dot net)
   Available for consulting/temporary embedded and systems.
   <http://cbfalconer.home.att.net>

glen herrmannsfeldt wrote:
>
> I used to know someone who supposedly (I never actually saw it) could
> type faster than the PDP-10 interrupt routine could process characters
> from the input port.  In that case, it would not be a real time system.
>

I think you could still call it a real time system, it is just the
operator that typed beyond specification.


Regards

Josep

I remember an example of a real-time system where times specified in days;
the payroll. The data has to be at the bank at a certain date, or everybody
gets their salary too late. (and will be very upset)

Wim