Return to BSD News archive
Newsgroups: comp.os.linux.misc,comp.os.linux.networking,comp.os.linux.setup,comp.unix.bsd.bsdi.misc,comp.unix.bsd.misc,comp.os.ms-windows.nt.advocacy,comp.os.os2.advocacy Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.ecn.uoknor.edu!feed1.news.erols.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!cam-news-hub1.bbnplanet.com!uunet!in2.uu.net!204.147.226.2!quack!quack.kfu.com!nsayer From: nsayer@quack.kfu.com (Nick Sayer) Subject: Re: Betting on Unix Message-ID: <nLVF2tL@quack.kfu.com> Sender: news@quack.kfu.com (0000-News(0000)) Organization: The Duck Pond public unix, +1 408 249 9630, log in as guest. References: <5d3sr2$44n@nntp1.best.com> <E50rGo.K3n@nonexistent.com> Date: Mon, 3 Feb 1997 17:35:21 UTC Lines: 57 Xref: euryale.cc.adfa.oz.au comp.os.linux.misc:155851 comp.os.linux.networking:67072 comp.os.linux.setup:95218 comp.unix.bsd.bsdi.misc:5845 comp.unix.bsd.misc:2209 comp.os.ms-windows.nt.advocacy:51712 comp.os.os2.advocacy:265276 Those predictions are too focused. I think more broad predictions are called for. It is not difficult to predict that PCI is going to be the bus of choice. I see the PPC CHiRP platform being the next industry platform of choice. Look at the past: The success of a platform has in large part been directly proportional to the number of manufacturers that have made the hardware (yes, Apple is the one glaring exception to this). A countless number of vendors make x86 machines, a perhaps not so countless, but non-trivial number of manufacturers make sparc machines. By contrast, one manufacturer makes Alpha machines, one manufacturer makes PA-risc machines. Many manufacturers are going to make CHiRP machines, and each of them is going to be compatable with any CHiRP OS. Buy an Apple and run Solaris on it. Buy a Sun and run NT on it. Buy a Motorola and run MacOS on it. Buy an IBM and run NetBSD on it. Buy a copy of AIX-PPC and burn it in effigy. CHiRP has the promise to displant the x86 platform from its throne. As for OS, there was an article in one of the magazines that opined that vis-a-vis NT, "The honeymoon is over." I have heard many a nightmarish tale about NT4.0 being unable to run on otherwise perfectly good hardware, and huge mountains of software that runs under win95 but refuses to under NT4. I may be a unix bigot, but if NT is the answer, it must be a stupid question. What would make Unix' future? Scott MacNeily can whine about Microsoft desktops wasting monumental amounts of admin time and energy if he likes, but unix will die without applications. If WABI or Wine can run Office-95, then that's one way. If not, then Sun or someone else needs to make an office productivity suite for Unix. It is in the application arena that an OS lives or dies. If applications are as plentiful and priced equally ($300 for Word and $1500 for Framemaker?!), then at long last the OS wars will be fought on even footing. But Unix' achiles heel is that its compatability is only at source level. Each CPU requires its own compiler, and each OS requires its own libraries. This, truly, is POSIX's next task - Application independence at the shared library level. If I have an x86 ELF binary, it should not matter what OS it was compiled on, I should be able to run it without modification or without fetching the libraries from somewhere. Unix is heading in this direction, but the quicker it gets there, the quicker it will be seen as an equal platform. And the quicker the applications will be ported to it. -- Nick Sayer <nsayer@quack.kfu.com> | "When DEC hits bottom, they're going N6QQQ @ N0ARY.#NORCAL.CA.USA.NOAM | to make an awful big splat." +1 408 249 9630, log in as 'guest' | URL: http://www.kfu.com/~nsayer/ | -- David Hawkins