Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/5/84; site ittral.UUCP
Path: utzoo!watmath!clyde!bonnie!akgua!whuxlm!harpo!decvax!ittvax!ittral!hall
From: hall@ittral.UUCP (Doug Hall)
Newsgroups: net.micro
Subject: Re: demise of 8086 family?
Message-ID: <169@ittral.UUCP>
Date: Fri, 21-Jun-85 23:47:28 EDT
Article-I.D.: ittral.169
Posted: Fri Jun 21 23:47:28 1985
Date-Received: Thu, 27-Jun-85 05:15:35 EDT
References: <120@SCINEWS.UUCP> <156@ittral.UUCP> <159@SCINEWS.UUCP>
Reply-To: hall@ittral.UUCP (Doug Hall)
Distribution: net
Organization: ITT Telecom B&CC Eng. Group, Raleigh, NC
Lines: 73
Summary: 

In article <159@SCINEWS.UUCP> jimi@SCINEWS.UUCP (Jim Ingram) writes:
>This paragraph leads one to believe the original posting's reference to
>software was to 8086 family software. How Mr. Hall misconstrued this I
>don't know. The original reference was to the large amount of S/370
>software out in the world. Even an ardent  IBM-hater must admit that
>$3.2 billion 1984 software revenues indicates acceptance in the business world.
> 

If anyone has misconstrued anything here, it is Mr. Ingram, who not
only missed my point, but the point of the original posting as well.
Certainly IBM has sold a lot of S/370 software, but that doesn't
guarantee the acceptance or success of the 370 chip set in the
PC market. If this chip set succeeds, it will be in a completely
different market than the current 8088/6 market. Somebody has to prove
to Joe User that an S/370 system offers more bang/buck than his IBM PC,
and they won't do it by waving a "370" flag in front of him. He
doesn't care what's in the box, as long as he can get nice integrated
applications for it, do his spreadsheets, write his reports, and make
his boss happy. Of that $3.2 billion in 370 software revenues, how
much was for specific applications like this?
>
>A 10MB disk is inadequate for serious business applications. An 8086
>family CPU is inadequate for large serious applications due to its 
>retarded architecture (64KB segments). Everything about the IBM PC and
>XT shouts "toy." A $3000 IBM PC or clone, despite what "power users"
>and "spreadsheet jockeys" think, is not a business machine without a
>lot of bags hanging off it. Single user systems may make some sense,
>but single tasking is brain-dead (yes I know CMS is single-tasking).
>
Again, Mr. Ingram misses the point. The original posting predicted
that this wonderful new 370-on-a-chip would mean the demise of Intel
micros, namely the 8088/6. The problem is that "serious business
applications" means two different things when you are comparing 370
systems and 8086 systems. The "spreadsheet jock" may find the 10 Mb
hard disk completely adequate, and he certainly views his applications
as serious. As far as "serious business applications" on a 370 scale
are concerned, nobody ever said that 10 MB was enough.

I won't argue about the retarded architecture of the 8086. A recent
poll I took on the net showed that over 90% preferred anything but
Intel, but many went on to admit that they still use the stuff because
of the software available. And *that* was the intent of my posting.
You can't win just by building a better mousetrap. I'll take my 68010
based Valid workstation over an IBM PC any day, but you'd never get a
manager to buy them for his "spreadsheet jocks."
>
>Assume there are a couple of million 3270 terminals installed, all connected
>to IBM mainframes, running applications under CICS, IMS, etc. It seems 
>plausible that all the MIS managers of IBM shops would buy machines that 
>beat the IBM 327x terminals on bang/buck (absurdly easy) and don't pose
>the data comm and architectural problems inherent in the PC. These same 
>managers control DP activities in companies with hundreds of thousands of 
>PCs. They could kill two birds with one stone with a box like this, and 
>they're asking for them now. 
>
While the 327x terminals may be attached to mainframes running
applications under CICS, IMS, etc, the PC's scattered all over the
building may not be. Why do you conclude that you can just substitute
a 370 system for an IBM PC and everyone will be happy? And even if
this did happen in all the "IBM shops" you'd still have a market for
the PC, right? So the S/370-on-a-chip still hasn't killed the Intel
(read PC) market, which was *my point to begin with*. I never said it
wouldn't succeed. But it won't drive everyone else into the dust,
either, no matter how much you'd like for it to.

By the way, I wasn't referring to the XT/370 in my original posting.
I know the XT/370 is a totally different beast. And I'm not pro-Intel or
pro-IBM, even though I used some of their trademarks in this article.
That should do for a disclaimer. 

Doug Hall
ITT Telecom, Raleigh NC
ittvax!ittral!hall