E. W. Dijkstra VS. John McCarthy. A rebuttal to Paul Graham's web writings.

From: Mike Cox (mikecoxlinux_at_yahoo.com)
Date: 08/13/04 

Date: Thu, 12 Aug 2004 23:04:16 -0700

The arguments that Paul Graham makes on his site in relation to LISP's
preceived superiourity to languages such as C/C++ and Java are incorrect.
In fact it *is* Java, C/C++ and even C# that are superiour to LISP.

This debate could actually be deconstructed in darwinian terms. The faster
(in terms of execution on a computer), the philosophy of design (quick and
gets the job done) usually win in nature and the market place. Look at the
competition between vi and emacs. Vi has a greater market share because it
is quick and easy, and supports the philosophy of being quick and getting
the job done.

The elitists want to make everything grand. Emacs has everything including
the kitchen sink and a lisp interpreter. Vi is simply a text editor,
nothing more. Little wonder that Vi won. I know some companies that ask
prespective employees which editor they use, and if they say "emacs", they
don't get hired because they will likely be prima donnas, where everything
has to be perfect before they can work.

John McCarthy built a programming language using Lambda Calculus as the
theory, while E. W. Dijkstra used Predicate Calculus to build Algol. In
terms of the progress to humanity that computers have made, most of it came
from the Algol, C, C++, and Java lineage. While Lambda Calculus is
difficult to understand, even McCarthy admitted he misunderstood it,
Predicate Calculus is clean efficient and gets the job done. That is why
Algol, C/C++ and Java won in the computing world. No other outcome was
possible, in nature, or computing.

Quantcast