minich@d.cs.okstate.edu (Robert Minich) (07/14/90)
by russotto@eng.umd.edu (Matthew T. Russotto): | pbiron@weber.ucsd.edu (Paul Biron) writes: |> |>It's even more of a misnomer than you think. |>When you switch between applications with Multi-Finder, the |>application you WERE using goes to sleep. I may be mistaken |>here, I don't get to use Multi-Finder very often because none |>of the Macs that I work on have enough memory. (If I am I'm |>sure someone will point it out :-) | | You ARE mistaken, as anyone who has downloaded a file, decoded a Stuffit file, | and used Microsoft Word all at the same time knows. [blah blah blah blah] Actually, a program has to tell MultiFinder to give it background time. (Setting the appropriate bit(s) in the SIZE resource.) Most apps have nothing useful to do in the background most of the time so they don't ask for background time and thus are actually asleep. Also, apps written before MF don't get any background time. Have you really decoded a stuffit file and done something else t the time? I have found that StuffIt (1.5.1) is a cpu hog like nobody's business and pretty much locks up the machine. Is there something I missed? -- | _ /| | Robert Minich |Q: Why is the food so lousy, and | \'o.O' | Oklahoma State University |the service so bad? Time traveler: | =(___)= | minich@d.cs.okstate.edu |A:The waiters know in advance what | U | - Bill sez "Ackphtth" |kind of tip they'll be getting.
ngg@bridge2.ESD.3Com.COM (Norman Goodger) (07/31/90)
In article <1990Jul14.043050.8494@d.cs.okstate.edu> minich@d.cs.okstate.edu (Robert Minich) writes: > Have you really decoded a stuffit file and done something else t the >time? I have found that StuffIt (1.5.1) is a cpu hog like nobody's >business and pretty much locks up the machine. Is there something I >missed? > Stuffit has a preference setting for allowing it to work in the background under MultiFinder, which really slows it down (for obvious reasons) If you do not have this setting turned on, Stuffit will become the CPU hog that you claim it to be. With all the discussion of Cooperative and pre-emptive multi- tasking, I may have missed it, but rarely is it mentioned that Cooperative multi-tasking appears on the surface to allow for much better performance than pre-emptive will allow. I suspect that if MultiFinder or future systems finally give us pre-emptive multi-tasking, performance is going to suffer right along with it, unless we have 50mhz+ CPU's for the masses.. -- Norm Goodger SysOp - MacInfo BBS @415-795-8862 3Com Corp. Co-SysOp FreeSoft RT - GEnie. Enterprise Systems Division (I disclaim anything and everything) UUCP: {3comvax,auspex,sun}!bridge2!ngg Internet: ngg@bridge2.ESD.3Com.COM
minich@d.cs.okstate.edu (Robert Minich) (07/31/90)
I write: | Have you really decoded a stuffit file and done something else t the | time? I have found that StuffIt (1.5.1) is a cpu hog like nobody's | business and pretty much locks up the machine. Is there something I | missed? ngg@bridge2.ESD.3Com.COM (Norman Goodger): | Stuffit has a preference setting for allowing it to work in | the background under MultiFinder, which really slows it down | (for obvious reasons) If you do not have this setting turned | on, Stuffit will become the CPU hog that you claim it to be. Actually, I _did_ have the checkbox set, but I found that StuffIt would give enough ccles in a 100K file to, say, update two or three windows in the finder. This, however, is on a 2048KE. Since then, I've gotten an SE/30 which seems to handle the load much better, although I still wasn't completely satisified. | With all the discussion of Cooperative and pre-emptive multi- | tasking, I may have missed it, but rarely is it mentioned | that Cooperative multi-tasking appears on the surface to | allow for much better performance than pre-emptive will allow. | I suspect that if MultiFinder or future systems finally give | us pre-emptive multi-tasking, performance is going to suffer | right along with it, unless we have 50mhz+ CPU's for the | masses.. Actually, with all the spoofing MultiFinder has to do, a well written scheduler that heavily favors the foreground task would most likely do a pretty good job at making MultiFinder look bad from a technical side, although the difference would most likely be less than perceptible with well written cooperative apps. Preemption is just one of those things that can make a programmer's life easier and cleaner. More important for the Mac, IMHO, is memory protection. Wouldn't it be nice to know when some app steps out of bounds, potentially twiddling someone else's bits? Perhaps Apple will make some sweeping changes with System 8, but after the delays we have seen for Sys 7 even though some features were dropped (my most wanted being the new printing architecture), I don't think I'll wait for it. :-) Maybe someday I'll be able to afford enough machine for A/UX and get a good bit from both ends of the spectrum at once. -- |_ /| | Robert Minich | |\'o.O' | Oklahoma State University| There are no heroes -- |=(___)= | minich@a.cs.okstate.edu | We all wear gray hats. | U | - Ackphtth |