Browse Prior Art Database

ADAPTABILITY OF THE LINEAR PROGRAMMING CODES

IP.com Disclosure Number: IPCOM000128846D
Original Publication Date: 1956-Aug-01
Included in the Prior Art Database: 2005-Sep-19
Document File: 8 page(s) / 33K

Publishing Venue

Software Patent Institute

Related People

Orchard-Hays, Wm.: AUTHOR [+3]

Abstract

The operation of an elaborate set of computer codes raises problems of its own which can only be appreciated from experience. Nevertheless, certain general principles for designing such a system can be set forth. The activities which are engaged in during the evolution of such a system are not a simple sequence of events but there is feedback from later steps to earlier ones. How easily the resulting changes can be handled is dependent on the organization of the codes and on the assembly program used. An appendix discusses some shortcomings of the latter together with suggested improvements.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 12% of the total text.

Page 1 of 8

THIS DOCUMENT IS AN APPROXIMATE REPRESENTATION OF THE ORIGINAL.

ADAPTABILITY OF THE LINEAR PROGRAMMING CODES

Wm. Orchard-Hays

P-908

1 August 1956

The RAND Corporation 1700 MAIN ST., SANTA MONICA, CALIFORNIA

SUMMARY

The operation of an elaborate set of computer codes raises problems of its own which can only be appreciated from experience. Nevertheless, certain general principles for designing such a system can be set forth. The activities which are engaged in during the evolution of such a system are not a simple sequence of events but there is feedback from later steps to earlier ones. How easily the resulting changes can be handled is dependent on the organization of the codes and on the assembly program used. An appendix discusses some shortcomings of the latter together with suggested improvements.

ADAPTABILITY OF THE LINEAR PROGRAMMING CODES

Wm. Orchard-Hays

The preceding discussion1 illustrated two points: (1) the operation of an elaborate set of computer codes raises problems of its own which have little to do with the nature of the mathematics, and (ii) if sufficient thought is given to what difficulties can arise and what variations on the method will be required -- both by the formulator and by exigencies of computer operations -- then provision can be made to handle the most common "unusual situations" smoothly.

It is difficult to demonstrate these ideas without getting into the messy details, which we have had to do in several lectures. The operation and use of a large set of codes tends to become an art in itself with considerations which override the very reasons for its existence. One purpose in presenting as many details of our 704 program as we have, has been to furnish an existing -- and we believe a good -example of a system in actual use with all its attendant paraphernalia . Nevertheless, there are some general principles that can be stated with regard to the design of such a system. First of all, we note that computers create a lot of work, simply by their nature, which does not exist in hand calculations. Hence the first principle is:

1 P-909

Rand Corporation Page 1 Aug 01, 1956

Page 2 of 8

ADAPTABILITY OF THE LINEAR PROGRAMMING CODES

(1) Get the computer to do as much of its own work as possible. This notion is not as obvious as, once stated, it might sound. It has taken four or five years for computer people to come to a firm conclusion in this regard. On punched-card equipment, including the CPC, every effort was made to cut down the work for the machines. But with the high- speed arithmetic and large storage of today's machines, it is a necessity for the computer to do its own assembling of code and data. There is often yet, however, a tendency for a programmer to rely heavily on proven library subroutines and to try to build a code to suit his needs from these. This mode of operation certainly has its place but, for a large class of problems, a compiling routine can be programmed which does the putting...