Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!cs.utexas.edu!uunet!crdgw1!sixhub!davidsen From: davidsen@sixhub.UUCP (Wm E. Davidsen Jr) Newsgroups: comp.sys.ibm.pc Subject: Re: Xerox sues Apple!!! Message-ID: <332@sixhub.UUCP> Date: 28 Dec 89 03:37:17 GMT References: <6767@tank.uchicago.edu> <1989Dec17.112127.27333@me.toronto.edu> <624@bogart.UUCP> <3368@rti.UUCP> Reply-To: davidsen@sixhub.UUCP (bill davidsen) Organization: *IX Public Access UNIX, Schenectady NY Lines: 67 In article <3368@rti.UUCP> bcw@rti.UUCP (Bruce Wright) writes: | I think anyone who looks at the two processor families objectively will | have to agree that the 68000 family is a better architecture. That's what starts flame wars, the implicit assumption that anyone who doesn't agree with you is not objective. | It's | certainly not perfect (maybe it should be more RISC-like, maybe it | shouldn't have the two different types of registers, etc), the NS32000 is an example of one set of registers. And they are really interchangable. | but it is | certainly better than the 80x86 family (8 equally inconvenient registers), At one time I posted a routine to take the integer square root of the stack pointer, as a demo that the registers are more general than is commonly thought. I make no claims that this is useful for anything. | though 80386 is a significant improvement (still a somewhat warped | architecture, though). The 68k has always had a linear address space, | which can be a great help compared to the 80x86's 64k segments The 386 and 486 have 2 GB segments (yes GB not MB). | (I'm not | sure that segmented architectures are always bad, but the segments need | to be at least 16M before they are large enough that you aren't always | tripping over them ... even then, some applications may be awkward on a | segmented machine). Some applications are difficult on a linear machine without memory mapping, since having separate text and data space requires some kind of control, such as segments or memory management. | Arguments about processor speed tend to be arguments | about levels of semiconductor technology, not arguments about processor | architecture. There's a big difference (although there can be some | differences in how efficiently a particular architecture can be | implemented). Generally true, although I'm sure someone will take exception. The hidden advantage of the Intel instruction set is that many of the instructions are single byte, thus reducing memory size (minor benefit) and allowing a greater number of instructions to be fetched per unit time given a limited memory bandwidth. This makes a difference in performance at the low end, with slow memory. I am NOT proposing any Intel or other CPU as "better" in some way than another, just adding a few observations. Since I was a user of the 68000 and 68010 I will let someone who is a hardware heavy explain all the things which were done wrong in the first cut of the 68000. Siffice it to say there were a LOT of fixes in the 68010. Not that there weren't bugs in the 8086, etc, or the early NS16000 line (the 32000 used to be called the 16000, to confuse the innocent). I'm sure someone can remember the things which were said about the Z8000, too, but that was never a large production chip. -- bill davidsen - sysop *IX BBS and Public Access UNIX davidsen@sixhub.uucp ...!uunet!crdgw1!sixhub!davidsen "Getting old is bad, but it beats the hell out of the alternative" -anon