Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!zdenko
From: zdenko@csd4.milw.wisc.edu (Zdenko Tomasic)
Newsgroups: comp.sys.next
Subject: Re: Selling through Businessland
Summary: bundled software, multiprocessing advantages
Keywords: NeXT, RISC, MACH, multiprocessing, bundled software
Message-ID: <1749@csd4.milw.wisc.edu>
Date: 30 Mar 89 07:00:06 GMT
References: <1554@neoucom.UUCP> <121@dg.dg.com>
Sender: news@csd4.milw.wisc.edu
Reply-To: zdenko@csd4.milw.wisc.edu (Zdenko Tomasic)
Organization: University of Wisconsin-Milwaukee
Lines: 51

In article <121@dg.dg.com> rec@dg.UUCP (Robert Cousins) writes:
>Actually, since there are now a number of more powerful machines
>available at lower cost, one wonders how successful this tactic
              ^^^^^^^^^
		  True, but what do you get with it?
		  How much in disks, memory for the low price?
		  What about software costs? Is it included?

>will actually be.  The new wave of 88000 based workstations (which
>do an honest 17+ MIPS) are available with Unix, monitor, etc. for
              ^^^^^^^^^^
		  Speed is intriguing, but how does it perform with respect
		  to i/o? Does the rest of the system keep up with CPU?
		  Are there compilers available to take advantage of the new
		  capabilities?
		  Which languages? Other software?

>approximately the same cost.  I know, I managed the DG AViiON AVX 400
               ^^^^^^^^^^^^^
		   With equivalent software and peripherals?

>development here at DG which fits in this class.  Anyone who has ever
>used a machine of this class will never want to move back to the 
>(now) relatively slow '030.
>
>Robert Cousins


I certainly appreciate fast CPUs, but I am afraid that the total
RISC system in the NeXT sense with bundled software will be much
more expensive. I wish someone would prove me wrong!

Also note that NeXT has a great potential, but it has yet to be
realized (afterall it was not officially released yet). It does
not seem to be impossible to drop 88k in the NeXT box without total
system redesign, but one should exploit multicpu capabilities of
MACH first with several '030s (and later 88K) including paralellizing
compilers. The latter is badly needed; just think of Postscript
printing while doing other stuff, not to mention the promised
non-baby 3D color capabilities (Randerman etc.).

In short, NeXT won't be obsolete tomorrow if they keep developing
and start materializing the cube's potential. The new RISC machines
have speed advantage, but also a lack of new software disadvantage.
Perhaps both can somehow merge for users benefit (let's hope
so).
--
___________________________________________________________________
Zdenko Tomasic, UWM, Chem. Dept., P.O. Box 413, Milwaukee, WI 53201
UUCP: uwvax!uwmcsd1!uwmcsd4!zdenko
ARPA: zdenko@csd4.milw.wisc.edu