jessea@dynasys.UUCP (Jesse W. Asher) (02/17/90)
I was wondering if MS Windows gives you a true multi-tasking environment?
I don't know much about MS Windows and I was wondering if there were any
advantages/disadvantages to running MS Windows versus running a unix
environment with virtual terminals. Thanx in advance.
--
Jesse W. Asher - Dynasys - (901)382-1705
6196-1 Macon Rd., Suite 200, Memphis, TN 38134
UUCP: {uunet,fedeva,rayo,ddmtg1}!dynasys!jesseapapa@pollux.usc.edu (Marco Papa) (02/18/90)
In article <51@dynasys.UUCP> jessea@dynasys.UUCP (Jesse W. Asher) writes: >I was wondering if MS Windows gives you a true multi-tasking environment? >I don't know much about MS Windows and I was wondering if there were any >advantages/disadvantages to running MS Windows versus running a unix >environment with virtual terminals. Thanx in advance. If for "true" multi-tasking you mean "preemptive" multi-tasking (i.e. multitasking based on a quantum of time) the answer is NO. MS-Windows (like Mac's Finder, for example) implements "non-preemptive" multi-tasking: task switching is performed only when particular system calls are executed. This means that a program can take over the system and disallow everybody else from getting cycles. This is in contrast with systems like UNIX and AmigaDOS which implement preemptive scheduling, in which task switching is controlled both by systems calls AND a timer. -- Marco -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= "Xerox sues somebody for copying?" -- David Letterman -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
abukarb@prism.cs.orst.edu (Bassam A. Abu-Karroum) (02/18/90)
In article <51@dynasys.UUCP> jessea@dynasys.UUCP (Jesse W. Asher) writes: >I was wondering if MS Windows gives you a true multi-tasking environment? >I don't know much about MS Windows and I was wondering if there were any >advantages/disadvantages to running MS Windows versus running a unix >environment with virtual terminals. Thanx in advance. > If you mean Windows/286 then the answer is no. Windows/286 uses memory swapping/banking to simulate multitasking. What this means is that once an application is in the background it is suspended until it is brought back into the foreground as the active application. Besides Windows/286 cannot regain control of the CPU if the application is busy. However, the case is different with Windows/386. Windows/386 provides you with true multitasking environment and it is preemptive for DOS applications. A foreground and a background application can multitask, if you choose to. Windows/386 offers you a preemptive environment for DOS applications. You may terminate a DOS application if it takes control of the CPU. Unless the problem is serious then you need to reboot the machine, but usually you can terminate that application and continue working without any problems. Windows/386 differs from UNIX by that it does not allow time-slicing control over multitasking process. Hope this clear things up. Bassam
amf@ecs.soton.ac.uk (Andrew Fountain) (02/19/90)
abukarb@prism.cs.orst.edu (Bassam A. Abu-Karroum) writes: >In article <51@dynasys.UUCP> jessea@dynasys.UUCP (Jesse W. Asher) writes: >>I was wondering if MS Windows gives you a true multi-tasking environment? >>I don't know much about MS Windows and I was wondering if there were any >>advantages/disadvantages to running MS Windows versus running a unix >>environment with virtual terminals. Thanx in advance. >> >If you mean Windows/286 then the answer is no. Windows/286 uses memory >swapping/banking to simulate multitasking. What this means is that once >an application is in the background it is suspended until it is brought >back into the foreground as the active application. Besides Windows/286 >cannot regain control of the CPU if the application is busy. The above is only true for non-windows tasks. Windows uses non-preemptive or cooperative multitasking whereas unix and os/2 use pre-emptive (time-sliced) multitasking. This means that a well behaved Windows program will regularly give control back to the system, and reasonably successful multitasking can be done. (How do you think the spooler program works.) The memory management is not directly linked to multitasking, although it makes it faster when lots of programs are loaded. Petzold's book gives a good explanation of this. amf@ecs.soton.ac.uk Andrew Fountain Tel: +44 703 592831 Dept of Electronics and Computer Science Fax: +44 703 593045 University of Southampton Telex: 47661 SOTONU G Southampton SO9 5NH England