Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!elroy.jpl.nasa.gov!hacgate!ashtate!tomr
From: tomr@ashtate (Tom Rombouts)
Newsgroups: comp.software-eng
Subject: The De-Sovietization of the Software Industry
Summary: Summary of Sept 90 "BYTE" Mag, pg. 252
Message-ID: <1183@ashton.UUCP>
Date: 7 Sep 90 00:46:23 GMT
Reply-To: tomr@ashton.UUCP (Tom Rombouts)
Organization: Ashton-Tate, Torrance, CA
Lines: 34

Buried amid BYTE's excellent 15th anniversay issue (page 252) is
a short piece by Lee Felsenstein that I have decied to excerpt without
permission for the benefit of this group:

"....Now I define Sovietization...as something run for the benefit
of management.  And I think there's been a process under way that results
in incremental improvement in the product performance as it should matter
to the customer, and really retrograde development in terms of the resources
that the product needs.  Is as if the efficiency of the personal computer
continues to decline.......I then asked "What do high-level languages get
you?....He said "...you can use cheaper programmers."  And so that has been
the development of the software industry as I have observed it.

The same programs, pretty much, [are] being written with ever higher
and higher levels of languages, being created with structures - industrial
structures, in effect - of programmers, in which everyone is filling in
boxes inside of boxes.  And the people [who] should know what [the software
is] doing, up at the top, play musical chairs.  Nobody, therefore, knows
what the product is supposed to be.  Whatever comes out is years late,
full of bugs, and the next new and improved version will fix some of the
bugs....The last thing on the list is a product that is efficient, a
product that does things that no one has ever done before, and a product
that is useful and usable by the users without requiring that they vastly
increase their hardware."   [ end of excerpts ]

Although this piece is by a hardware designer discussing micro computers,
his criticisms would seem to apply to many, many software projects.
All the fine theory and research that is discussed in this group is
great, but it still takes a great deal of realpolitik and praxis
(to borrow some Soviet phrases) to achieve changes in real, bottom-line,
quarterly-results oriented world.


Tom Rombouts  Torrance Techie  tomr@ashtate.A-T.com  V:(213)538-7108