Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sun-barr!lll-winken!ames!uhccux!waikato.ac.nz!ldo From: ldo@waikato.ac.nz (Lawrence D'Oliveiro, Waikato University) Newsgroups: comp.sys.mac.programmer Subject: Re: Gripes about System 7.0 Message-ID: <1991Jan25.183443.2825@waikato.ac.nz> Date: 25 Jan 91 05:34:43 GMT References: <5611@idunno.Princeton.EDU> <3810@uakari.primate.wisc.edu> <2899@casbah.acns.nwu.edu> <1991Jan24.224108.19413@ux1.cso.uiuc.edu> Organization: University of Waikato, Hamilton, New Zealand Lines: 53 In article <1991Jan24.224108.19413@ux1.cso.uiuc.edu>, dorner@pequod.cso.uiuc.edu (Steve Dorner) has a few things to say about AppleTalk and the Macintosh system: "I think Apple BLEW IT on AppleTalk. "The only interesting idea they had was the plug-and-play nature of the thing." No, there are *two* interesting ideas, as far as I'm concerned: dynamic node address assignment (the single major reason for the plug-and-play nature), and transaction-based protocols. See pages I-14 to I-15 of Inside AppleTalk, 2nd Ed. "The protocol itself is horrendously myopic; it just won't scale to large internetworks. That's why they've patched in Phase 2." AppleTalk is a great local-area networking system (better with Phase 2), lousy on wide-area networking, agreed. TCP/IP is a fairly good, very well understood wide-area networking system, a real headache in a local-area network. What network protocols do you know of that work well in both situations? Are they available today? "Even the hardware was bad; nobody in their right mind uses real LocalTalk, because the cable is expensive and the connectors fall out. PhoneNet is cheaper, more reliable, and has fewer distance/topology limitations." Ding! You've just proven that AppleTalk is properly designed after all! If it wasn't, you wouldn't have such a choice of different physical media. Even the ease with which you can replace LocalTalk cabling with PhoneNet tells you something about the design of that particular data-link layer... "... Macintosh is a GUI without an operating system." Rubbish. Note: talking about the design of an operating system is meaningless unless you keep in mind the needs of actual applications. Macintosh is not an operating system like traditional ones running on time-shared machines. But then, the applications that run under it are not like the traditional ones developed for time-shared systems. Yes, Macintosh needs preemptive multi-tasking. But I would assert that the claimed benefits of memory protection and virtual memory are *overrated*. Lawrence D'Oliveiro fone: +64-71-562-889 Computer Services Dept fax: +64-71-384-066 University of Waikato electric mail: ldo@waikato.ac.nz Hamilton, New Zealand 37^ 47' 26" S, 175^ 19' 7" E, GMT+13:00