Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!usc!zaphod.mps.ohio-state.edu!mips!daver!zorch!xanthian From: xanthian@zorch.SF-Bay.ORG (Kent Paul Dolan) Newsgroups: comp.sys.amiga Subject: Re: Multitasking at home (Was Reality check: ....) Message-ID: <1990Dec28.000727.24394@zorch.SF-Bay.ORG> Date: 28 Dec 90 00:07:27 GMT References: <186ab528.ARN08a6@easy.hiam> <186abe5f.ARN08a9@easy.hiam> <1990Dec27.164630.10864@ncsuvx.ncsu.edu> Organization: SF-Bay Public-Access Unix Lines: 73 Kevin put lots of disclaimers at the bottom of his article to get him off the hook on this, but I still want to answer one part. kdarling@hobbes.ncsu.edu (Kevin Darling) writes: > Dwight Hubbard writes: >> Kevin Darling writes: >>> Change the program a bit: write one which simply loops with no I/O at all, >>> and then boost the priority on it under Amiga Exec to be highest. >>> What happens to multitasking then? >> Same thing as would happen under Unix or OS/2 it dies. >First, hope you had a good holiday! Second, I thought the answers people >gave to the original "true" question were very good, and pointed out that >preemptive multitasking gives a real, tho sometimes subtle, advantage. >My impulse for posting that obviously leading question was to note >that there _are_ layers of subtlety involved when it comes to defining >"true multitasking"... especially in what it means to an end user. Right, and "true" and "false" are directions on a continuum, not really well defined endpoints. Lots of times it helps in keeping a discussion like this civil to remember that we are using rather fuzzy terminology at best. >For instance, the stated example shows that a higher priority task >under Exec with no I/O becomes exactly like a similar Mac process: it >must give up control explicitly... something some of us would not consider >"true" multitasking. Or at least, not as user-nice. Under OS9, all >processes also "age", and eventually get a chance to run. Even OS/2 has >a settable timeout to prevent one process from locking out the others. Probably so; aging cpu bound processes to get lower priorities and i/o bound ones to get higher priorities, and old processes to get higher priorities is a fairly common refinement for multiple user multi-tasking OSs. However, I would absolutely _hate_ priority aging on the Amiga, or any single user machine, except perhaps as an option I have to specifically _en_able. When I up the priority of my terminal session, since it is almost 80% of the big chunks of time either waiting for keystrokes, or waiting for new text from the host site to show up for the screen, my extra three priority "points" have little effect; other processes get almost all the cpu cycles. _But_, when I turn on the kermit session for which I default the terminal task priority to 3 in my startup script, I _desperately_ need that priority to stay high, even, no, especially if the task runs at high priority for hours. The last thing I want is to start losing packets long into a multimegabyte single file download, something I do an average of weekly at least. On a multi-user system, priority aging is appropriate; it helps prevent one user from unfairly slowing down the system for everyone else, and it keeps jobs from sitting in queue "forever". I think it is entirely inappropriate to a single user system, where the user has, and wants to have, complete control over priorities. In your example, the user has just done something stupid by setting the priority of a compute bound job high; that is the exact opposite of normal practice, and the bad results should be blamed on operator error, not on inferior multitasking algorithms. Just a comment from out in left field, lest anyone start clamoring for aged priorities without thinking it through. You're right in the part I omitted, Kevin, I would NOT want the OS/9 method for the Amiga; OS/9 is, I understand, fairly often used for multiuser systems. I hope no one _ever_ creates a deliberately multiuser Amiga; that would be a huge step backwards, even though hacks to make BBSs work are fine anyway; I just don't want to see the machine go multiuser in focus; the world of computing is moving away from that idea, for me, correctly. Kent, the man from xanth.