The Universe of Discourse : My 1992 view of the problems of computer programming in 1992
The Universe of Discourse
Mark Dominus (陶敏修)
mjd@pobox.com
About me
RSS<br>Atom
12 recent entries
My 1992 view of the problems of computer programming in 1992<br>Egyptian fraction multiplication<br>Update: Here I am at the Sagrada Família<br>Egyptian fractions for 2/105<br>Did Ahmes find the best expansions for 2/n?<br>Programmers will document for Claude, but not for each other<br>How are John Waters movies like James Bond movies?<br>Documentation is a message in a bottle<br>Bo Diddley<br>Language models imply world models<br>John Haugeland on the failure of micro-worlds<br>Crooked politicians love crab cakes!
Archive:
2026:<br>JFMAMJ<br>2025:<br>JFMAMJ<br>JASOND<br>2024:<br>JFMAMJ<br>JASOND<br>2023:<br>JFMAMJ<br>JASOND<br>2022:<br>JFMAMJ<br>JASOND<br>2021:<br>JFMAMJ<br>JASOND<br>2020:<br>JFMAMJ<br>JASOND<br>2019:<br>JFMAMJ<br>JASOND<br>2018:<br>JFMAMJ<br>JASOND<br>2017:<br>JFMAMJ<br>JASOND<br>2016:<br>JFMAMJ<br>JASOND<br>2015:<br>JFMAMJ JASOND<br>2014:<br>JFMAMJ JASOND<br>2013:<br>JFMAMJ JASOND<br>2012:<br>JFMAMJ<br>JASOND<br>2011:<br>JFMAMJ<br>JASOND<br>2010:<br>JFMAMJ<br>JASOND<br>2009:<br>JFMAMJ<br>JASOND<br>2008:<br>JFMAMJ<br>JASOND<br>2007:<br>JFMAMJ<br>JASOND<br>2006:<br>JFMAMJ<br>JASOND<br>2005: OND
Subtopics:
Mathematics250<br>Programming102<br>Language95<br>Miscellaneous75<br>Book50<br>Tech49<br>Etymology36<br>Haskell33<br>Oops30<br>Unix27<br>Cosmic Call25<br>Math SE25<br>Law23<br>Physics21<br>Perl17<br>Biology16<br>Brain15<br>Calendar15<br>Food15
-->
-->
Technorati Profile<br>-->
Comments disabled
Thu, 18 Jun 2026
My 1992 view of the problems of computer programming in 1992
While cleaning out my office today, I found this, which I wrote in 1992:
In the middle 1970's, the IBM corporation did (and perhaps<br>still does) most of their in-house programming in a computer language<br>called FORTRAN. They had a pretty good FORTRAN compiler, called the<br>FORTRAN G compiler. It was fast at translating FORTRAN into machine<br>instructions, and the machine instructions it produced implemented the<br>desired behavior fairly efficiently. Nevertheless, IBM decided to write<br>a new compiler.
This was very daring in the middle 1970's, because compilers were<br>quite complicated programs, and are difficult to write, and it was<br>surprising that IBM was willing to invest the vast resources that a new<br>compiler would require when an adequate compiler was still available.<br>IBM spent millions of dollars and hundreds of programmer-years, and<br>produced the FORTRAN H compiler, which was fast, efficient, and full of<br>nice features. It was an excellent compiler and is still the one that<br>they use.
Here is the first punch line: Compiler programs are no longer<br>difficult to write. The past fifteen years have seen an enormous<br>increase in our understanding of compiler technology and how to write a<br>compiler. Compilers are so easy to write now that third-year<br>undergraduate computer science majors are expected to be able to turn<br>out passable compilers in one semester.
Now a question: Since we're obviously thousands of times better at<br>producing compilers than we were fifteen years ago, so much so that a<br>single undergraduate can write a passable one in four months, why hasn't<br>IBM invested millions of dollars and hundreds of programmer-years to<br>produce a super FORTRAN I compiler that's thousands of times better than<br>the FORTRAN H compiler?
The answer is that compiler program quality is no longer the limiting<br>factor on our ability to write computer programs. The problems that<br>programmers face no longer have to do with how good the compiler is.<br>Instead, they are problems of method and language. We don't really<br>know how to program yet, or how to manage our programs. We don't<br>really know what we want to say or how to say it. We don't have good<br>computer languages for expressing what we want to computer to do. We<br>don't know how to think about programming. In short, the reason IBM<br>doesn't bother with a super FORTRAN I compiler, is that no matter how<br>good it was, it would still be FORTRAN.
Computer programming is still a black art. It's less than fifty years<br>old, and nobody is very good at it yet. We can make better tools than<br>we know how to use.
[Other articles in category /prog]<br>permanent link