-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

>>
>> You cannot tell in general, that's a basic result from CS. But you can
>> measure previous runs and do predictions based on that, in some cases
>> at least. I hope I'm not answering a homework assignment...

Nope , this has nothing to do with homework , just i was wondering how
does an OS implement such an algorithm in real world.

>>
>>      -Otto
>>
> In general you cannot predict, however there are many (long) jobs with 
> very predictable times to completion: sorts, merges, most anything that
> processes thousands of records in one batch operation.
> (and ties up various resources for the duration --- thein is the gotcha)
> I would not trust counting instructions, loops, subroutine calls as
> being usefully predictive of execution time.

Yeah , looping time depends the complexity of that loop , i've learned
that ,
We use a O(n) to present such complexity of a program.

> 
> The fun thing about scheduling algorithms is that any one of them
> is usually theoretically capable of giving the worst possible overall 
> performance.


That's why there's so many runtime exceptions in cplusplus , can't avoid
all of it. It's unpredictable anyway.


- -- 
Best Regards,
Aaron Lewis - PGP: 0x4A6D32A0
FingerPrint EA63 26B2 6C52 72EA A4A5 EB6B BDFE 35B0 4A6D 32A0
irc: A4r0n on freenode
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkvLoNgACgkQvf41sEptMqDkJgCfXSYyJHBBzyt4QeFmKu8v/Ra7
aAUAn3jdYLwCvUfeyjA0BjsEchphInqC
=zInF
-----END PGP SIGNATURE-----

Reply via email to