Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!uakari.primate.wisc.edu!unmvax!ariel.unm.edu!hydra.unm.edu!burd
From: burd@hydra.unm.edu (Stephen Burd ASM)
Newsgroups: comp.infosystems
Subject: Re: UNIX applications in Information Systems.
Message-ID: <1990Oct2.222322.24372@ariel.unm.edu>
Date: 2 Oct 90 22:23:22 GMT
References: <1990Sep28.020409.16535@murdoch.acc.Virginia.EDU> <1990Sep28.171417.29070@ariel.unm.edu> <1990Oct2.092412.7363@tukki.jyu.fi>
Sender: news@ariel.unm.edu (USENET News System)
Organization: University of New Mexico, Albuquerque
Lines: 53

In article <1990Oct2.092412.7363@tukki.jyu.fi> sakkinen@jytko.jyu.fi (Markku Sakkinen) writes:
>In article <1990Sep28.171417.29070@ariel.unm.edu> burd@hydra.unm.edu (Stephen Burd ASM) writes:
>> ...
>>The primary advantage of UNIX in any software development effort lies
>>in its inherent flexibility for integrating various pieces of software.

<... system description deleted ...>

>>Neither VMS nor any IBM operating system would have been capable of tying all
>>of this software together to form an integrated on-line system.
>> ...
>
>Any solid arguments for "would not have been capable"?

Detailed arguments have been emailed to the poster.  Others that are
interested are referred to the summary below, the references that 
follow, and/or invited to correspond with me directly.  Unless there 
is substantial interest by comp.infosystems readers I'll restrict 
this discussion to email.

To summarize, building a system that integrates many components 
in a transparent fashion (to the user) and that operates in an 
on-line/real-time environment requires extensive operating system
support for multiple process execution and for communication between
processes.  These capabilities are an integral part of the UNIX operating
system and access to them is provided through interfaces to the
various programming languages (e.g., C and FORTRAN).  Such capabilities 
are not nearly so robust or extensive in the VMS environment and are 
virtually nonexistent in IBM environments other than AIX (a UNIX variant).
The problem is far less critical if the constraints of user transparency 
and on-line/real-time performance are relaxed.  Perhaps "would not have 
been capable" was too strong a statement, but I doubt that it's far off 
the mark for the system to which I was referring.


(1)	Burd, Stephen D., and Suleiman K. Kassicieh, "Logic Based
	Decision Support for Computer Capacity Planning", Information &
	Management, volume 13 (October 1987), pp.125-133.

(2)	Burd, Stephen D., and Suleiman K. Kassicieh, "Decision Support
	for Supercomputer Acquisitions", to appear in Operations Research,
	early 1991.

(3)	Bullers, William I., Stephen D. Burd, and Suleiman K. Kassicieh,
	"Implementatilon Issues for Intelligent Decision Support Systems",
	Proceedings of the IEEE Workshop on Languages for Automation, New
	Orleans, Louisiana, November 7-9, 1984.

--
FROM:      Stephen D. Burd			AT&T:      (505)-277-6418
USNAIL:    Anderson School of Management
           University of New Mexico		INTERNET:  burd@hydra.unm.edu
           Albuquerque, NM  87131