Browse Prior Art Database

Early Days of FORTRAN: Afterword

IP.com Disclosure Number: IPCOM000129432D
Original Publication Date: 1984-Jan-01
Included in the Prior Art Database: 2005-Oct-06
Document File: 2 page(s) / 15K

Publishing Venue

Software Patent Institute

Related People

JOHN BACKUS: AUTHOR [+2]

Abstract

I would like to discuss what my friends on the FORTRAN project really accomplished. Fran Allen's talk has made my task simple, because it brings out a little-known fact that the computer community has never quite recognized: that 25 years ago, my friends did this incredible thing of producing a compiler that did a better job of optimizing programs in virtually all respects than any compiler in the next 20 years (until 1978). In fact, Fran goes so far as to say -- and I think she is someone who really understands these things [Top of Page 27] better than most of us -- that most of today's production compilers do not optimize as well as the FORTRAN I compiler. It is unusual for any technical achievement to remain the best in its field for even 5 years. What circuit design or program has not been outperformed by a successor within 3 or 4 years? It is an incredible achievement that 25 years ago these people designed and produced a compiler that has remained the best overall optimizer for not 5 years, not 10 years, but 20 years.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

THIS DOCUMENT IS AN APPROXIMATE REPRESENTATION OF THE ORIGINAL.

Copyright ©; 1984 by the American Federation of Information Processing Societies, Inc. Used with permission.

Early Days of FORTRAN: Afterword

JOHN BACKUS

I would like to discuss what my friends on the FORTRAN project really accomplished. Fran Allen's talk has made my task simple, because it brings out a little-known fact that the computer community has never quite recognized: that 25 years ago, my friends did this incredible thing of producing a compiler that did a better job of optimizing programs in virtually all respects than any compiler in the next 20 years (until 1978). In fact, Fran goes so far as to say -- and I think she is someone who really understands these things better than most of us -- that most of today's production compilers do not optimize as well as the FORTRAN I compiler. It is unusual for any technical achievement to remain the best in its field for even 5 years. What circuit design or program has not been outperformed by a successor within 3 or 4 years? It is an incredible achievement that 25 years ago these people designed and produced a compiler that has remained the best overall optimizer for not 5 years, not 10 years, but 20 years.

By the way, when I say that somebody "wrote a section of the compiler," it is important to remember that what I really mean is that they invented it -- they developed all the groundbreaking techniques used in it. It is a great understatement to say only, "Somebody wrote a section."

(Image Omitted: David Sayre)

(Image Omitted: Harlan Herrick)

(Image Omitted: Peter Sheridan)

Peter Sheridan wrote Section 1, which parsed algebraic expressions, translated them into code, and optimized that code. Harlan Herrick invented the DO-statement and wrote the part of Section 1 that put all the source information that wasn't used in algebraic expressions into tables that were needed by later sections. Herrick should also be credited with introducing the key words GO To! Roy Nutt designed most of the input/output language and wrote the part of Section 1 that translated I/O statements into Do-loops....