ed@alt.dah.sub.org (Ed Braaten) (11/21/90)
m1phm02@fed.frb.gov (Patrick H. McAllister) writes: > ... Someone posted a followup about how much they got >"free" off the MIT distribution -- having played with some of that >stuff, it seems to me that it is only free if you are not the person >paying for the hours of time required to port it (unless you happen to >be running the same type of machine as the author -- in spite of the >"portability" of UNIX software, somehow I find myself having to do >some porting on about half of the stuff I get). I find that once you've gotten used to having source code at your disposal it is hard to go back to shrink-wrapped software. However, I wouldn't try to push the idea of "do-it-yourself" on the average computer user... ;-) >Back to the original question, about advanced technology; my >candidate, assuming it actually works, is the way in which the >scheduler gives priority to the task with which the user is currently >interacting. (I say "assuming . . ." because I haven't used OS/2 >myself, but only read about it). On my Sun, I find it extremely >frustrating to start typing away into EMACS and not even have my >keystrokes echoed for several seconds, I assume because EMACS has >been paged out since I have't typed anything for a little while. Hmm. Isn't that why they say EMACS = "Eating Memory And Constantly Swapping"... Too bad you haven't experienced swapping under OS/2 - you would appreciate your Sun more... :-/ > Has >anyone ever put a scheduler into UNIX that is designed to maximize >interactive responsiveness for a single user, rather than to maximize >total system throughput? Why change the scheduler? Have you tried /bin/ed? On my System V/386 the binary is a mere 33K and is *blazingly* fast! Your fellow users will love you for using it! ;-) --------------------------------------------------------------------------- Ed Braaten | "... Man looks at the outward appearance, Work: ed@imuse.de.intel.com | but the Lord looks at the heart." Home: ed@alt.dah.sub.org | 1 Samuel 16:7b ---------------------------------------------------------------------------