Saturday, May 14, 2005

Paradigm contest

Through his divine love, God bestowed mortality upon us. Otherwise, we would be doomed to an eternal existence under the mentality of Adam. At youth we are curious and eagerly seek explanations. Soon, we erect the China Wall with bricks of dogma. The only hope for a different future is then, our perishing so the new generation can start over.

I would not be surprised to run into an article showing how to write a device driver in Lisp. For some, the theoretical proof of equivalence of paradigms is not the issue. They insist that the paradigm of their choice is all we need.

Constraining expressiveness results in lack of it. Unrestricted flexibility is not the remedy, either. We need to classify our target domains and investigate the linguistic abstractions enabling us to correctly state problems in those domains. The product of our endeavor must be a coherent language, not a bag of languages.

Some problems are intrinsic to the language itself. The constructs of structured programming dealt with this type of problem. Otherwise, the nature of problems inspires the language, not the contrary.

The contemporary class of problems are concerned with large distributed software concurrently executing on diverse platforms. We need a form of expressiveness to ensure some degree of correctness directly supported by the compiler. Ultimately, it is the language that determines the outcome, not the comments.

The abstract language Z++ has been designed with above considerations.

Labels: , ,